Sorting algorithms/Strand sort

This page uses content from Wikipedia. The original article was at Strand sort. The list of authors can be seen in the page history. As with Rosetta Code, the text of Wikipedia is available under the GNU FDL. (See links for details on variance)

Task

Implement the Strand sort.

This is a way of sorting numbers by extracting shorter sequences of already sorted numbers from an unsorted list.

11l

Translation of: Python

F merge_list(&a, &b)
   [Int] out
   L !a.empty & !b.empty
      I a[0] < b[0]
         out.append(a.pop(0))
      E
         out.append(b.pop(0))
   out [+]= a
   out [+]= b
   R out

F strand(&a)
   V i = 0
   V s = [a.pop(0)]
   L i < a.len
      I a[i] > s.last
         s.append(a.pop(i))
      E
         i++
   R s

F strand_sort(&a)
   V out = strand(&a)
   L !a.empty
      out = merge_list(&out, &strand(&a))
   R out

print(strand_sort(&[1, 6, 3, 2, 1, 7, 5, 3]))

Output:

[1, 1, 2, 3, 3, 5, 6, 7]

AppleScript

Strand sort seems to be essentially a merge sort with a particular way of setting up the initial blocks.

-- Sort items l thru r of theList in place, ascending.
on strandSort(theList, l, r)
    -- Deal with negative and/or transposed range index parameters.
    set listLength to (count theList)
    if (l < 0) then set l to listLength + l + 1
    if (r < 0) then set r to listLength + r + 1
    if (l > r) then set {l, r} to {r, l}
    if ((l < 1) or (r > listLength)) then error "strandSort(): range index parameter(s) outside list range."
    
    script o
        property dest : theList -- Original list.
        property src : my dest's items l thru r -- The items in the sort range.
        property ranges : {}
    end script
    
    -- Individually list-wrap the items in o's src to avoid having to
    -- hard-code their actual class in the line marked ** below.
    repeat with i from 1 to (r - l + 1)
        set o's src's item i to {o's src's item i}
    end repeat
    -- Extract "strands" of existing order from the sort range items
    -- and write the resulting runs over the range in the original list.
    set i to l
    repeat until (i > r)
        set j to i
        set jv to o's src's beginning's beginning -- The value in src's first sublist.
        set o's dest's item j to jv -- Store it in the next original-list slot
        set o's src's item 1 to missing value -- Replace the sublist with a placeholder.
        -- Do the same with any later values that are sequentially greater or equal.
        repeat with k from 2 to (count o's src)
            set kv to o's src's item k's beginning
            if (kv < jv) then
            else
                set j to j + 1
                set o's dest's item j to kv
                set jv to kv
                set o's src's item k to missing value
            end if
        end repeat
        set o's ranges's end to {i, j} -- Note this strand's range in the list.
        set o's src to o's src's lists -- Lose src's zapped sublists.  **
        set i to j + 1
    end repeat
    set strandCount to (count o's ranges)
    if (strandCount = 1) then return -- The input list was already in order.
    
    -- Work out how many passes the iterative merge will take and from this whether
    -- the auxiliary list has to be the source or the destination during the first pass.
    -- The destination in the final pass has to be the original list.
    set passCount to 0
    repeat while (2 ^ passCount < strandCount)
        set passCount to passCount + 1
    end repeat
    if (passCount mod 2 = 0) then
        set o's src to o's dest
        set o's dest to o's dest's items
    else
        set o's src to o's dest's items
    end if
    
    -- Merge the strands.
    repeat passCount times
        set k to l -- Destination index.
        repeat with rr from 2 to strandCount by 2 -- Per pair of ranges.
            set {{i, ix}, {j, jx}} to o's ranges's items (rr - 1) thru rr
            set o's ranges's item (rr - 1) to {i, jx}
            set o's ranges's item rr to missing value
            
            set iv to o's src's item i
            set jv to o's src's item j
            repeat until (k > jx)
                if (iv > jv) then
                    set o's dest's item k to jv
                    if (j < jx) then
                        set j to j + 1
                        set jv to o's src's item j
                    else
                        repeat with i from i to ix
                            set k to k + 1
                            set o's dest's item k to o's src's item i
                        end repeat
                    end if
                else
                    set o's dest's item k to iv
                    if (i < ix) then
                        set i to i + 1
                        set iv to o's src's item i
                    else
                        repeat with k from j to jx
                            set o's dest's item k to o's src's item k
                        end repeat
                    end if
                end if
                set k to k + 1
            end repeat
        end repeat
        if (rr < strandCount) then -- Odd range at the end of this pass?
            set {i, ix} to o's ranges's end
            repeat with k from i to ix
                set o's dest's item k to o's src's item k
            end repeat
        end if
        
        set o's ranges to o's ranges's lists
        set strandCount to (strandCount + 1) div 2
        set {o's src, o's dest} to {o's dest, o's src}
    end repeat
    
    return -- nothing.
end strandSort

local lst
set lst to {5, 1, 4, 37, 2, 0, 9, 6, -44, 3, 8, 7}
strandSort(lst, 1, -1)
return lst

Output:

{-44, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 37}

AutoHotkey

Works with: AutoHotkey_L

string =
(
-2 0 -2 5 5 3 -1 -3 5 5 0 2 -4 4 2
)
string2 := string
Loop
{
	loop, parse, string, %A_space%
	{
		list := 1 = A_index ? A_loopfield : list
		StringSplit, k, list, %A_space%

		if ( k%k0% <= A_loopfield ) && ( l != "" ) && ( A_index != 1 )
			list := list . " " . A_loopfield 

		if ( k%k0% > A_loopfield )
			list := A_loopfield . " " . list , index++
		l := A_loopfield
	}		 
		if ( index = 0 )
		{
			MsgBox % "unsorted:" string2 "`n    Sorted:" list
			exitapp
		}
		string := list, list = "", index := 0
	}
esc::ExitApp

outout

unsorted:-2 0 -2 5 5 3 -1 -3 5 5 0 2 -4 4 2
  Sorted:-4 -3 -2 -2 -1 0 0 2 2 3 4 5 5 5 5

C

Strand sort using singly linked list. C99, compiled with gcc -std=c99

#include <stdio.h>

typedef struct node_t *node, node_t;
struct node_t { int v; node next; };
typedef struct { node head, tail; } slist;

void push(slist *l, node e) {
	if (!l->head) l->head = e;
	if (l->tail)  l->tail->next = e;
	l->tail = e;
}

node removehead(slist *l) {
	node e = l->head;
	if (e) {
		l->head = e->next;
		e->next = 0;
	}
	return e;
}

void join(slist *a, slist *b) {
	push(a, b->head);
	a->tail = b->tail;
}

void merge(slist *a, slist *b) {
	slist r = {0};
	while (a->head && b->head)
		push(&r, removehead(a->head->v <= b->head->v ? a : b));

	join(&r, a->head ? a : b);
	*a = r;
	b->head = b->tail = 0;
}

void sort(int *ar, int len)
{
	node_t all[len];

	// array to list
	for (int i = 0; i < len; i++)
		all[i].v = ar[i], all[i].next = i < len - 1 ? all + i + 1 : 0;

	slist list = {all, all + len - 1}, rem, strand = {0},  res = {0};

	for (node e = 0; list.head; list = rem) {
		rem.head = rem.tail = 0;
		while ((e = removehead(&list)))
			push((!strand.head || e->v >= strand.tail->v) ? &strand : &rem, e);

		merge(&res, &strand);
	}

	// list to array
	for (int i = 0; res.head; i++, res.head = res.head->next)
		ar[i] = res.head->v;
}

void show(const char *title, int *x, int len)
{
	printf("%s ", title);
	for (int i = 0; i < len; i++)
		printf("%3d ", x[i]);
	putchar('\n');
}

int main(void)
{
	int x[] = {-2,0,-2,5,5,3,-1,-3,5,5,0,2,-4,4,2};
#	define SIZE sizeof(x)/sizeof(int)

	show("before sort:", x, SIZE);
	sort(x, sizeof(x)/sizeof(int));
	show("after sort: ", x, SIZE);

	return 0;
}

outout

before sort:  -2   0  -2   5   5   3  -1  -3   5   5   0   2  -4   4   2
after sort:   -4  -3  -2  -2  -1   0   0   2   2   3   4   5   5   5   5

C++

#include <list>

template <typename T>
std::list<T> strandSort(std::list<T> lst) {
  if (lst.size() <= 1)
    return lst;
  std::list<T> result;
  std::list<T> sorted;
  while (!lst.empty()) {
    sorted.push_back(lst.front());
    lst.pop_front();
    for (typename std::list<T>::iterator it = lst.begin(); it != lst.end(); ) {
      if (sorted.back() <= *it) {
        sorted.push_back(*it);
        it = lst.erase(it);
      } else
        it++;
    }
    result.merge(sorted);
  }
  return result;
}

Clojure

(ns rosettacode.strand-sort)

(defn merge-join
  "Produces a globally sorted seq from two sorted seqables"
  [[a & la :as all] [b & lb :as bll]]
  (cond (nil? a) bll
        (nil? b) all
        (< a b) (cons a (lazy-seq (merge-join la bll)))
        true    (cons b (lazy-seq (merge-join all lb)))))

(defn unbraid
  "Separates a sorted list from a sequence"
  [u]
  (when (seq u)
    (loop [[x & xs] u
           u []
           s []
           e x]
      (if (nil? x)
        [s u]
        (if (>= x e)
          (recur xs u (conj s x) x)
          (recur xs (conj u x) s e))))))

(defn strand-sort
  "http://en.wikipedia.org/wiki/Strand_sort"
  [s]
  (loop [[s u] (unbraid s)
         m nil]
    (if s
      (recur (unbraid u) (merge-join m s))
      m)))

(strand-sort [1, 6, 3, 2, 1, 7, 5, 3])
;;=> (1 1 2 3 3 5 6 7)

CMake

Only for lists of integers.

# strand_sort(<output variable> [<value>...]) sorts a list of integers.
function(strand_sort var)
  # Strand sort moves elements from _ARGN_ to _answer_.
  set(answer)                   # answer: a sorted list
  while(DEFINED ARGN)
    # Split _ARGN_ into two lists, _accept_ and _reject_.
    set(accept)                 # accept: elements in sorted order
    set(reject)                 # reject: all other elements
    set(p)
    foreach(e ${ARGN})
      if(DEFINED p AND p GREATER ${e})
        list(APPEND reject ${e})
      else()
        list(APPEND accept ${e})
        set(p ${e})
      endif()
    endforeach(e)

    # Prepare to merge _accept_ into _answer_. First, convert both lists
    # into arrays, for better indexing: set(e ${answer${i}}) is faster
    # than list(GET answer ${i} e).
    set(la 0)
    foreach(e ${answer})
      math(EXPR la "${la} + 1")
      set(answer${la} ${e})
    endforeach(e)
    set(lb 0)
    foreach(e ${accept})
      math(EXPR lb "${lb} + 1")
      set(accept${lb} ${e})
    endforeach(e)

    # Merge _accept_ into _answer_.
    set(answer)
    set(ia 1)
    set(ib 1)
    while(NOT ia GREATER ${la})         # Iterate elements of _answer_.
      set(ea ${answer${ia}})
      while(NOT ib GREATER ${lb})       # Take elements from _accept_,
        set(eb ${accept${ib}})          #   while they are less than
        if(eb LESS ${ea})               #   next element of _answer_.
          list(APPEND answer ${eb})
          math(EXPR ib "${ib} + 1")
        else()
          break()
        endif()
      endwhile()
      list(APPEND answer ${ea})         # Take next from _answer_.
      math(EXPR ia "${ia} + 1")
    endwhile()
    while(NOT ib GREATER ${lb})         # Take rest of _accept_.
      list(APPEND answer ${accept${ib}})
      math(EXPR ib "${ib} + 1")
    endwhile()

    # This _reject_ becomes next _ARGN_. If _reject_ is empty, then
    # set(ARGN) undefines _ARGN_, breaking the loop.
    set(ARGN ${reject})
  endwhile(DEFINED ARGN)

  set("${var}" ${answer} PARENT_SCOPE)
endfunction(strand_sort)

strand_sort(result 11 55 55 44 11 33 33 44 22 22)
message(STATUS "${result}")  # -- 11;11;22;22;33;33;44;44;55;55

Common Lisp

(defun strand-sort (l cmp)
  (if l
    (let* ((l (reverse l))    
	   (o (list (car l))) n)
      (loop for i in (cdr l) do
	    (push i (if (funcall cmp (car o) i) n o)))
      (merge 'list o (strand-sort n cmp) #'<))))

(let ((r (loop repeat 15 collect (random 10))))
  (print r)
  (print (strand-sort r #'<)))

output

(5 8 6 0 6 8 4 7 0 7 1 5 3 3 6) 
(0 0 1 3 3 4 5 5 6 6 6 7 7 8 8)

D

Using doubly linked lists

import std.stdio, std.container;

DList!T strandSort(T)(DList!T list) {
    static DList!T merge(DList!T left, DList!T right) {
        DList!T result;
        while (!left.empty && !right.empty) {
            if (left.front <= right.front) {
                result.insertBack(left.front);
                left.removeFront();
            } else {
                result.insertBack(right.front);
                right.removeFront();
            }
        }
        result.insertBack(left[]);
        result.insertBack(right[]);
        return result;
    }

    DList!T result, sorted, leftover;

    while (!list.empty) {
        leftover.clear();
        sorted.clear();
        sorted.insertBack(list.front);
        list.removeFront();
        foreach (item; list) {
            if (sorted.back <= item)
                sorted.insertBack(item);
            else
                leftover.insertBack(item);
        }
        result = merge(sorted, result);
        list = leftover;
    }

    return result;
}

void main() {
    auto lst = DList!int([-2,0,-2,5,5,3,-1,-3,5,5,0,2,-4,4,2]);
    foreach (e; strandSort(lst))
        write(e, " ");
}

Output:

-4 -3 -2 -2 -1 0 0 2 2 3 4 5 5 5 5

Faster version using slices

import std.stdio, std.array;

T[] strandSort(T)(const(T)[] list) pure nothrow {
    static T[] merge(const(T)[] left, const(T)[] right) pure nothrow {
        T[] res;
        while (!left.empty && !right.empty) {
            if (left.front <= right.front) {
                res ~= left.front;
                left.popFront;
            } else {
                res ~= right.front;
                right.popFront;
            }
        }
        return res ~ left ~ right;
    }

    T[] result;
    while (!list.empty) {
        auto sorted = list[0 .. 1];
        list.popFront;
        typeof(sorted) leftover;
        foreach (const item; list)
            (sorted.back <= item ? sorted : leftover) ~= item;
        result = merge(sorted, result);
        list = leftover;
    }

    return result;
}

void main() {
    const arr = [-2, 0, -2, 5, 5, 3, -1, -3, 5, 5, 0, 2, -4, 4, 2];
    arr.strandSort.writeln;
}

Output:

[-4, -3, -2, -2, -1, 0, 0, 2, 2, 3, 4, 5, 5, 5, 5]

EasyLang

Translation of: 11l

proc merge . a[] b[] .
   a = 1 ; b = 1
   while a <= len a[] and b <= len b[]
      if a[a] < b[b]
         r[] &= a[a]
         a += 1
      else
         r[] &= b[b]
         b += 1
      .
   .
   while a <= len a[]
      r[] &= a[a]
      a += 1
   .
   while b <= len b[]
      r[] &= b[b]
      b += 1
   .
   swap a[] r[]
.
proc strand . a[] s[] .
   s[] = [ a[1] ]
   for i = 2 to len a[]
      if a[i] > s[$]
         s[] &= a[i]
      else
         an[] &= a[i]
      .
   .
   swap a[] an[]
.
proc strandsort . a[] .
   strand a[] out[]
   while len a[] > 0
      strand a[] b[]
      merge out[] b[]
   .
   swap a[] out[]
.
a[] = [ 1 6 3 2 1 7 5 3 ]
strandsort a[]
print a[]

Output:

[ 1 1 2 3 3 5 6 7 ]

Elixir

Translation of: Ruby

defmodule Sort do
  def strand_sort(args), do: strand_sort(args, [])
  
  defp strand_sort([], result), do: result
  defp strand_sort(a, result) do
    {_, sublist, b} = Enum.reduce(a, {hd(a),[],[]}, fn val,{v,l1,l2} ->
                        if v <= val, do: {val, [val | l1], l2},
                                   else: {v,   l1, [val | l2]}
                      end)
    strand_sort(b, :lists.merge(Enum.reverse(sublist), result))
  end
end

IO.inspect Sort.strand_sort [7, 17, 6, 20, 20, 12, 1, 1, 9]

Output:

[1, 1, 6, 7, 9, 12, 17, 20, 20]

Euphoria

function merge(sequence left, sequence right)
    sequence result
    result = {}
    while length(left) > 0 and length(right) > 0 do
        if left[$] <= right[1] then
            exit
        elsif right[$] <= left[1] then
            return result & right & left
        elsif left[1] < right[1] then
            result = append(result,left[1])
            left = left[2..$]
        else
            result = append(result,right[1])
            right = right[2..$]
        end if
    end while
    return result & left & right
end function

function strand_sort(sequence s)
    integer j
    sequence result
    result = {}
    while length(s) > 0 do
        j = length(s)
        for i = 1 to length(s)-1 do
            if s[i] > s[i+1] then
                j = i
                exit
            end if
        end for
        
        result = merge(result,s[1..j])
        s = s[j+1..$]
    end while
    return result
end function

constant s = rand(repeat(1000,10))
puts(1,"Before: ")
? s
puts(1,"After:  ")
? strand_sort(s)

Output:

Before: {551,746,940,903,51,18,346,417,340,502}
After:  {18,51,340,346,417,502,551,746,903,940}

F#

// Strand sort. Nigel Galloway: August 18th., 2023
let     fN g=let mutable n=g in fun g->if n>g then false else n<-g; true
let     fI n=let fN=fN(List.head n) in List.partition fN n
let rec fG n g=[match n,g with [],g|g,[]->yield! g
                              |n::gn,i::ng when n<i->yield n; yield! fG gn g
                              |n,g::ng->yield g; yield! fG n ng]
let rec fL n g=match n with []->g |_->let n,i=fI n in fL i (n::g)
let sort n=fL n []|>List.fold(fun n g->fG n g)[]
printfn "%A" (sort ["one";"two";"three";"four"]);;
printfn "%A" (sort [2;3;1;5;11;7;5])

Output:

["four"; "one"; "three"; "two"]
[1; 2; 3; 5; 5; 7; 11]

Go

package main

import "fmt"

type link struct {
    int
    next *link
}

func linkInts(s []int) *link {
    if len(s) == 0 {
        return nil
    }
    return &link{s[0], linkInts(s[1:])}
}

func (l *link) String() string {
    if l == nil {
        return "nil"
    }
    r := fmt.Sprintf("[%d", l.int)
    for l = l.next; l != nil; l = l.next {
        r = fmt.Sprintf("%s %d", r, l.int)
    }
    return r + "]"
}

func main() {
    a := linkInts([]int{170, 45, 75, -90, -802, 24, 2, 66})
    fmt.Println("before:", a)
    b := strandSort(a)
    fmt.Println("after: ", b)
}

func strandSort(a *link) (result *link) {
    for a != nil {
        // build sublist
        sublist := a
        a = a.next
        sTail := sublist
        for p, pPrev := a, a; p != nil; p = p.next {
            if p.int > sTail.int {
                // append to sublist
                sTail.next = p
                sTail = p
                // remove from a
                if p == a {
                    a = p.next
                } else {
                    pPrev.next = p.next
                }
            } else {
                pPrev = p
            }
        }
        sTail.next = nil // terminate sublist
        if result == nil {
            result = sublist
            continue
        }
        // merge
        var m, rr *link
        if sublist.int < result.int {
            m = sublist
            sublist = m.next
            rr = result
        } else {
            m = result
            rr = m.next
        }
        result = m
        for {
            if sublist == nil {
                m.next = rr
                break
            }
            if rr == nil {
                m.next = sublist
                break
            }
            if sublist.int < rr.int {
                m.next = sublist
                m = sublist
                sublist = m.next
            } else {
                m.next = rr
                m = rr
                rr = m.next
            }
        }
    }
    return
}

Output:

before: [170 45 75 -90 -802 24 2 66]
after:  [-802 -90 2 24 45 66 75 170]

Haskell

-- Same merge as in Merge Sort
merge :: (Ord a) => [a] -> [a] -> [a]
merge [] ys = ys
merge xs [] = xs
merge (x : xs) (y : ys)
	| x <= y = x : merge xs (y : ys)
	| otherwise = y : merge (x : xs) ys

strandSort :: (Ord a) => [a] -> [a]
strandSort [] = []
strandSort (x : xs) = merge strand (strandSort rest) where
	(strand, rest) = extractStrand x xs
	extractStrand x [] = ([x], [])
	extractStrand x (x1 : xs)
		| x <= x1 = let (strand, rest) = extractStrand x1 xs in (x : strand, rest)
		| otherwise = let (strand, rest) = extractStrand x xs in (strand, x1 : rest)

J

Generally, this task should be accomplished in J using /:~. Here we take an approach that's more comparable with the other examples on this page.

Using merge defined at Sorting algorithms/Merge sort#J:

strandSort=: (#~ merge $:^:(0<#)@(#~ -.)) (= >./\)

Example use:

   strandSort 3 1 5 4 2
1 2 3 4 5

Note: the order in which this J implementation processes the strands differs from the pseudocode currently at the wikipedia page on strand sort and matches the haskell implementation currently at the wikipedia page.

Also note that the individual strands can be seen by using ; instead of merge.

   ((#~ ; $:^:(0<#)@(#~ -.)) (= >./\)) 3 1 5 4 2
┌───┬───┬─┬┐
│3 5│1 4│2││
└───┴───┴─┴┘
   ((#~ ; $:^:(0<#)@(#~ -.)) (= >./\)) 3 3 1 2 4 3 5 6
┌─────────┬─────┬┐
│3 3 4 5 6│1 2 3││
└─────────┴─────┴┘

Java

Works with: Java version 1.6+

import java.util.Arrays;
import java.util.LinkedList;

public class Strand{
	// note: the input list is destroyed
	public static <E extends Comparable<? super E>> 
	LinkedList<E> strandSort(LinkedList<E> list){
		if(list.size() <= 1) return list;

		LinkedList<E> result = new LinkedList<E>();
		while(list.size() > 0){
			LinkedList<E> sorted = new LinkedList<E>();
			sorted.add(list.removeFirst()); //same as remove() or remove(0)
			for(Iterator<E> it = list.iterator(); it.hasNext(); ){
				E elem = it.next();
				if(sorted.peekLast().compareTo(elem) <= 0){
					sorted.addLast(elem); //same as add(elem) or add(0, elem)
					it.remove();
				}
			}
			result = merge(sorted, result);
		}
		return result;
	}

	private static <E extends Comparable<? super E>>
	LinkedList<E> merge(LinkedList<E> left, LinkedList<E> right){
		LinkedList<E> result = new LinkedList<E>();
		while(!left.isEmpty() && !right.isEmpty()){
			//change the direction of this comparison to change the direction of the sort
			if(left.peek().compareTo(right.peek()) <= 0)
				result.add(left.remove());
			else
				result.add(right.remove());
		}
		result.addAll(left);
		result.addAll(right);
		return result;
	}
	
	public static void main(String[] args){
		System.out.println(strandSort(new LinkedList<Integer>(Arrays.asList(3,1,2,4,5))));
		System.out.println(strandSort(new LinkedList<Integer>(Arrays.asList(3,3,1,2,4,5))));
		System.out.println(strandSort(new LinkedList<Integer>(Arrays.asList(3,3,1,2,4,3,5,6))));
	}
}

Output:

[1, 2, 3, 4, 5]
[1, 2, 3, 3, 4, 5]
[1, 2, 3, 3, 3, 4, 5, 6]

jq

Most of the implementation is the "merge" function for merging two arrays. Notice that the helper function, strand, is defined here as an inner function.

# merge input array with array x by comparing the heads of the arrays 
# in turn; # if both arrays are sorted, the result will be sorted:
def merge(x):
  length as $length
  | (x|length) as $xl
  | if $length == 0 then x
    elif $xl == 0 then .
    else 
      . as $in
      | reduce range(0; $xl + $length) as $z
         # state [ix, xix, ans]
         ( [0, 0, []];
           if .[0] < $length and 
              ((.[1] < $xl and $in[.[0]] <= x[.[1]]) or .[1] == $xl)
           then [(.[0] + 1), .[1], (.[2] + [$in[.[0]]]) ]
           else [.[0], (.[1] + 1), (.[2] + [x[.[1]]]) ]
           end
         ) | .[2]
    end ;

def strand_sort:
  # The inner function emits [strand, remainder]
  def strand:
    if length <= 1 then .
    else 
      reduce .[] as $x 
      # state: [strand, remainder]
      ([ [], [] ];
       if ((.[0]|length) == 0) or .[0][-1] <= $x
       then [ (.[0] + [$x]), .[1] ]
       else [ .[0], (.[1] + [$x]) ]
       end )
    end ;

  if length <= 1 then .
  else strand as $s
    | ($s[0] | merge( $s[1] | strand_sort))
  end ;

Example:

[1,3,5,2,4,6] | strand_sort

Julia

Translation of: Python

function mergelist(a, b)
    out = Vector{Int}()
    while !isempty(a) && !isempty(b)
        if a[1] < b[1]
            push!(out, popfirst!(a))
        else
            push!(out, popfirst!(b))
        end
    end
    append!(out, a)
    append!(out, b)
    out
end
 
function strand(a)
    i, s = 1, [popfirst!(a)]
    while i < length(a) + 1
        if a[i] > s[end]
            append!(s, splice!(a, i))
        else
            i += 1
        end
    end
    s
end

strandsort(a) = (out = strand(a); while !isempty(a) out = mergelist(out, strand(a)) end; out)
 
println(strandsort([1, 6, 3, 2, 1, 7, 5, 3]))

Output:


[1, 1, 2, 3, 3, 5, 6, 7]

Kotlin

Translation of: D

// version 1.1.2

fun <T : Comparable<T>> strandSort(l: List<T>): List<T> {
    fun merge(left: MutableList<T>, right: MutableList<T>): MutableList<T> {
        val res = mutableListOf<T>()
        while (!left.isEmpty() && !right.isEmpty()) {
            if (left[0] <= right[0]) {
                res.add(left[0])
                left.removeAt(0)
            }
            else {
                res.add(right[0])
                right.removeAt(0)
            }
        }
        res.addAll(left)
        res.addAll(right)
        return res
    }
    
    var list = l.toMutableList() 
    var result = mutableListOf<T>()
    while (!list.isEmpty()) {
        val sorted = mutableListOf(list[0])
        list.removeAt(0)
        val leftover = mutableListOf<T>()
        for (item in list) {
            if (sorted.last() <= item)
                sorted.add(item)
            else
                leftover.add(item)
        }
        result = merge(sorted, result)
        list = leftover  
    }
    return result
}

fun main(args: Array<String>) {
    val l = listOf(-2, 0, -2, 5, 5, 3, -1, -3, 5, 5, 0, 2, -4, 4, 2)
    println(strandSort(l))
}

Output:

[-4, -3, -2, -2, -1, 0, 0, 2, 2, 3, 4, 5, 5, 5, 5]

Mathematica /Wolfram Language

StrandSort[ input_ ] := Module[ {results = {}, A = input},
While[Length@A > 0, 
 sublist = {A[[1]]}; A = A[[2;;All]];
  For[i = 1, i < Length@A, i++,
   If[ A[[i]] > Last@sublist, AppendTo[sublist, A[[i]]]; A = Delete[A, i];]
  ];
 results = #[[Ordering@#]]&@Join[sublist, results];];
results ]
StrandSort[{2, 3, 7, 5, 1, 4, 7}]

Output:

{1, 2, 3, 4, 5, 7, 7}

MAXScript

fn strandSort arr =
(
	arr = deepcopy arr
	local sub = #()
	local results = #()
	while arr.count > 0 do
	(
		sub = #()
		append sub (amax arr)
		deleteitem arr (for i in 1 to arr.count where arr[i] == amax arr collect i)[1]
		local i = 1
		while i <= arr.count do
		(
			if arr[i] > sub[sub.count] do
			(
				append sub arr[i]
				deleteitem arr i
			)
			i += 1
		)
		results = join sub results
	)
	return results

)

Output:

a = for i in 1 to 20 collect random 1 40
#(19, 26, 14, 31, 11, 33, 2, 14, 32, 28, 12, 38, 2, 37, 27, 18, 31, 24, 39, 28)
strandSort a
#(2, 2, 11, 12, 14, 14, 18, 19, 24, 26, 27, 28, 28, 31, 31, 32, 33, 37, 38, 39)

NetRexx

/* NetRexx */
options replace format comments java crossref savelog symbols binary

import java.util.List

placesList = [String -
    "UK  London",     "US  New York",   "US  Boston",     "US  Washington" -
  , "UK  Washington", "US  Birmingham", "UK  Birmingham", "UK  Boston"     -
]

lists = [ -
    placesList -
  , strandSort(String[] Arrays.copyOf(placesList, placesList.length)) -
]

loop ln = 0 to lists.length - 1
  cl = lists[ln]
  loop ct = 0 to cl.length - 1
    say cl[ct]
    end ct
    say
  end ln

return

method strandSort(A = String[]) public constant binary returns String[]

  rl = String[A.length]
  al = List strandSort(Arrays.asList(A))
  al.toArray(rl)

  return rl

method strandSort(Alst = List) public constant binary returns ArrayList

  A = ArrayList(Alst)
  result = ArrayList()
  loop label A_ while A.size > 0
    sublist = ArrayList()
    sublist.add(A.get(0))
    A.remove(0)
    loop i_ = 0 while i_ < A.size - 1
      if (Comparable A.get(i_)).compareTo(Comparable sublist.get(sublist.size - 1)) > 0 then do
        sublist.add(A.get(i_))
        A.remove(i_)
        end
      end i_
      result = merge(result, sublist)
    end A_

  return result

method merge(left = List, right = List) public constant binary returns ArrayList

  result = ArrayList()
  loop label mx while left.size > 0 & right.size > 0
    if (Comparable left.get(0)).compareTo(Comparable right.get(0)) <= 0 then do
      result.add(left.get(0))
      left.remove(0)
      end
    else do
      result.add(right.get(0))
      right.remove(0)
      end
    end mx
    if left.size > 0 then do
      result.addAll(left)
      end
    if right.size > 0 then do
      result.addAll(right)
      end

  return result

Output

UK  London
US  New York
US  Boston
US  Washington
UK  Washington
US  Birmingham
UK  Birmingham
UK  Boston

UK  Birmingham
UK  Boston
UK  London
UK  Washington
US  Birmingham
US  Boston
US  New York
US  Washington

Nim

proc mergeList[T](a, b: var seq[T]): seq[T] =
  result = @[]
  while a.len > 0 and b.len > 0:
    if a[0] < b[0]:
      result.add a[0]
      a.delete 0
    else:
      result.add b[0]
      b.delete 0
  result.add a
  result.add b

proc strand[T](a: var seq[T]): seq[T] =
  var i = 0
  result = @[a[0]]
  a.delete 0
  while i < a.len:
    if a[i] > result[result.high]:
      result.add a[i]
      a.delete i
    else:
      inc i

proc strandSort[T](a: seq[T]): seq[T] =
  var a = a
  result = a.strand
  while a.len > 0:
    var s = a.strand
    result = mergeList(result, s)

var a = @[1, 6, 3, 2, 1, 7, 5, 3]
echo a.strandSort

Output:

@[1, 1, 2, 3, 3, 5, 6, 7]

OCaml

Translation of: Haskell

let rec strand_sort (cmp : 'a -> 'a -> int) : 'a list -> 'a list = function
   []    -> []
 | x::xs ->
   let rec extract_strand x = function
      [] -> [x], []
    | x1::xs when cmp x x1 <= 0 ->
      let strand, rest = extract_strand x1 xs in x::strand, rest
    | x1::xs ->
      let strand, rest = extract_strand x xs in strand, x1::rest
   in
   let strand, rest = extract_strand x xs in
   List.merge cmp strand (strand_sort cmp rest)

usage

# strand_sort compare [170; 45; 75; -90; -802; 24; 2; 66];;
- : int list = [-802; -90; 2; 24; 45; 66; 75; 170]

PARI/GP

strandSort(v)={
	my(sorted=[],unsorted=v,remaining,working);
	while(#unsorted,
		remaining=working=List();
		listput(working, unsorted[1]);
		for(i=2,#unsorted,
			if(unsorted[i]<working[#working],
				listput(remaining, unsorted[i])
			,
				listput(working, unsorted[i])
			)
		);
		unsorted=Vec(remaining);
		sorted=merge(sorted, Vec(working))
	);
	sorted
};
merge(u,v)={
	my(ret=vector(#u+#v),i=1,j=1);
	for(k=1,#ret,
		if(i<=#u & (j>#v | u[i]<v[j]),
			ret[k]=u[i];
			i++
		,
			ret[k]=v[j];
			j++
		)
	);
	ret
};

Pascal

program StrandSortDemo;
 
type
  TIntArray = array of integer;

function merge(left: TIntArray; right: TIntArray): TIntArray;
  var
    i, j, k: integer;
  begin
    setlength(merge, length(left) + length(right));
    i := low(merge);
    j := low(left);
    k := low(right);
    repeat
      if ((left[j] <= right[k]) and (j <= high(left))) or (k > high(right)) then
      begin
        merge[i] := left[j];
        inc(j);
      end
      else
      begin
        merge[i] := right[k];
        inc(k);
      end;
      inc(i);
    until i > high(merge);
  end;

function StrandSort(s: TIntArray): TIntArray;
  var
    strand: TIntArray;
    i, j: integer;
  begin
    setlength(StrandSort, length(s));
    setlength(strand, length(s));
    i := low(s);
    repeat
      StrandSort[i] := s[i];
      inc(i);
    until (s[i] < s[i-1]);
    setlength(StrandSort, i);
    repeat
      setlength(strand, 1);
      j := low(strand);
      strand[j] := s[i];
      while (s[i+1] > s[i]) and (i < high(s)) do
      begin
        inc(i);
        inc(j);
	setlength(strand, length(strand) + 1);
        Strand[j] := s[i];
      end;
      StrandSort := merge(StrandSort, strand);
      inc(i);
    until (i > high(s));
  end;

var
  data: TIntArray;
  i: integer;

begin
  setlength(data, 8);
  Randomize;
  writeln('The data before sorting:');
  for i := low(data) to high(data) do
  begin
    data[i] := Random(high(data));
    write(data[i]:4);
  end;
  writeln;
  data := StrandSort(data);
  writeln('The data after sorting:');
  for i := low(data) to high(data) do
  begin
    write(data[i]:4);
  end;
  writeln;
end.

Perl

use strict;
use warnings;
use feature 'say';

sub merge {
    my ($x, $y) = @_;
    my @out;
    while (@$x and @$y) {
        my $t = $x->[-1] <=> $y->[-1];
        if    ($t == 1)  { unshift @out, pop @$x }
        elsif ($t == -1) { unshift @out, pop @$y }
        else             { splice @out, 0, 0, pop(@$x), pop(@$y) }
    }
    @$x, @$y, @out
}

sub strand {
    my $x = shift;
    my @out = shift @$x // return;
    for (-@$x .. -1) {
        push @out, splice @$x, $_, 1 if $x->[$_] >= $out[-1];
    }
    @out
}

sub strand_sort {
    my @x = @_;
    my(@out, @strand);
    @out = merge \@out, \@strand while @strand = strand(\@x);
    @out
}

my @a = map (int rand(100), 1 .. 10);
say "Before @a";
@a = strand_sort(@a);
say "After  @a";

Phix

with javascript_semantics

function merge(sequence left, right)
    sequence result = {}
    while length(left)>0 
      and length(right)>0 do
        if left[$]<=right[1] then
            exit
        elsif right[$]<=left[1] then
            return result & right & left
        elsif left[1]<right[1] then
            result = append(result,left[1])
            left = left[2..$]
        else
            result = append(result,right[1])
            right = right[2..$]
        end if
    end while
    return result & left & right
end function
 
function strand_sort(sequence s)
    sequence result = {}
    while length(s)>0 do
        integer j = length(s)
        for i=1 to length(s)-1 do
            if s[i]>s[i+1] then
                j = i
                exit
            end if
        end for
        result = merge(result,s[1..j])
        s = s[j+1..$]
    end while
    return result
end function

?strand_sort(shuffle(tagset(10)))

Output:

{1,2,3,4,5,6,7,8,9,10}

PHP

Translation of: D

Works with: PHP 5.3.0+

$lst = new SplDoublyLinkedList();
foreach (array(1,20,64,72,48,75,96,55,42,74) as $v)
    $lst->push($v);
foreach (strandSort($lst) as $v)
    echo "$v ";

function strandSort(SplDoublyLinkedList $lst) {
    $result = new SplDoublyLinkedList();
    while (!$lst->isEmpty()) {
        $sorted = new SplDoublyLinkedList();
        $remain = new SplDoublyLinkedList();
        $sorted->push($lst->shift());
        foreach ($lst as $item) {
            if ($sorted->top() <= $item) {
                $sorted->push($item);
            } else {
                $remain->push($item);
            }
        }
        $result = _merge($sorted, $result);
        $lst = $remain;
    }
    return $result;
}

function _merge(SplDoublyLinkedList $left, SplDoublyLinkedList $right) {
    $res = new SplDoublyLinkedList();
    while (!$left->isEmpty() && !$right->isEmpty()) {
        if ($left->bottom() <= $right->bottom()) {
            $res->push($left->shift());
        } else {
            $res->push($right->shift());
        }
    }
    foreach ($left as $v)  $res->push($v);
    foreach ($right as $v) $res->push($v);
    return $res;
}

1 20 42 48 55 64 72 74 75 96

PicoLisp

(de strandSort (Lst)
   (let Res NIL  # Result list
      (while Lst
         (let Sub (circ (car Lst))  # Build sublist as fifo
            (setq
               Lst (filter
                  '((X)
                     (or
                        (> (car Sub) X)
                        (nil (fifo 'Sub X)) ) )
                  (cdr Lst) )
               Res (make
                  (while (or Res Sub)  # Merge
                     (link
                        (if2 Res Sub
                           (if (>= (car Res) (cadr Sub))
                              (fifo 'Sub)
                              (pop 'Res) )
                           (pop 'Res)
                           (fifo 'Sub) ) ) ) ) ) ) )
      Res ) )

Test:

: (strandSort (3 1 5 4 2))
-> (1 2 3 4 5)

: (strandSort (3 abc 1 (d e f) 5 T 4 NIL 2))
-> (NIL 1 2 3 4 5 abc (d e f) T)

PL/I

strand: procedure options (main); /* 27 Oct. 2012 */
   declare A(100) fixed, used(100) bit (1), sorted fixed controlled;
   declare (temp, work) fixed controlled;
   declare (i, j, k, n) fixed binary;

   n = hbound(A, 1);
   used = '1'b;
   A = random()*99;

   put edit (A) (f(3));

   do while (allocation(sorted) < n);
      call fetch (A, work);
      call move  (temp, work);

      call merge(sorted, temp); 
         /* Merges elements in SORTED with elements in TEMP. */
   end;
   /* Transfer the sorted elements to A. */
   do i = 1 to allocation(sorted);
      A(i) = sorted; free sorted;
   end;
   /* Print the sorted values. */
   put skip list ('The sorted values are:');
   put skip edit (A) (f(3));

/* Merges elements of SORTED with elements of TEMP and places  */
/* the result in SORTED. */
/* Elements in SORTED and TEMP are in forward order. */
merge: procedure (sorted, temp);
   declare (sorted, temp) fixed controlled;
   declare work fixed controlled;
   declare (j_ok, k_ok) bit (1);

   do until ((k_ok | j_ok) = '0'b);
      k_ok = allocation(sorted) > 0;
      j_ok = allocation(temp)   > 0;
      if k_ok & j_ok then
         do; 
            if sorted <= temp then 
               do; allocate work; work = sorted; free sorted; end;
            else
               do; allocate work; work = temp; free temp; end;
         end;
      else
         if allocation(temp) = 0 then
             /* temp is empty; copy remainder of sorted into work */
            do while (allocation(sorted) > 0);
               allocate work; work = sorted; free sorted;
            end;
         else
            /* sorted is empty; copy remainder of temp onto work */
            do while (allocation(temp) > 0);
               allocate work; work = temp; free temp;
            end;
   end;

   call move (sorted, work); /* Move the values to SORTED. */

end merge;

/* Collect a thread of ascending values from aray A, and stack them in temp. */
/* Note: the values in temp are in reverse order. */
fetch: procedure (A, temp);
   declare A(*) fixed, temp controlled fixed;
   declare i fixed binary;
   
   do i = 1 to hbound(A,1); 
      if used(i) then
         do; allocate temp; temp = A(i); used(i) = '0'b; go to found; end;
   end;
found:
   do i = i+1 to hbound(A,1);
      if (temp <= A(i)) & used(i) then 
         do; allocate temp; temp = A(i); used(i) = '0'b; end;
   end;
end fetch;

/* Copy the stack at TEMP to the stack at SORTED. */
/* In TEMP, elements are in reverse order;   */
/* in SORTED, elements are in forward order. */
move: procedure (sorted, temp);
   declare (sorted, temp) fixed controlled;

   do while (allocation(sorted) > 0); free sorted; end;
   do while (allocation (temp) > 0);
      allocate sorted; sorted = temp; free temp;
   end;
end move;

end strand;

Generated data:

 43  5 79 16 90 48 29 73 29 19 77 59 49  2 54 35 39 71 25 76 34 48 31 91 28 13 23 70 27 59 96  7 63 82 59 81 28 96 34 43
 81 98 21 47 72 57 45 64 94 51 18 11 65 12 61 97 13 84 95 89 43  8 14 31 58 68 58 39 59 26 72 38 26 85 30 89 42 90 29 11
 14 63 97 60  1 17 45 42 62 29 45 15 69 11 29 25 11 48 92  3

Results:

The sorted values are: 
  1  2  3  5  7  8 11 11 11 11 12 13 13 14 14 15 16 17 18 19 21 23 25 25 26 26 27 28 28 29 29 29 29 29 30 31 31 34 34 35
 38 39 39 42 42 43 43 43 45 45 45 47 48 48 48 49 51 54 57 58 58 59 59 59 59 60 61 62 63 63 64 65 68 69 70 71 72 72 73 76
 77 79 81 81 82 84 85 89 89 90 90 91 92 94 95 96 96 97 97 98

PureBasic

Procedure strandSort(List a())
  Protected NewList subList()
  Protected NewList results()
  
  While ListSize(a()) > 0
    ClearList(subList())
    AddElement(subList())
    FirstElement(a())
    subList() = a()
    DeleteElement(a())
    ForEach a()
      If a() >= subList()
        AddElement(subList())
        subList() = a()
        DeleteElement(a())
      EndIf
    Next
    
    ;merge lists
    FirstElement(subList())
    If Not FirstElement(results())
      ;copy all of sublist() to results()
      MergeLists(subList(), results(), #PB_List_Last)
    Else
      Repeat
        If subList() < results()
          InsertElement(results())
          results() = subList()
          DeleteElement(subList())
          If Not NextElement(subList())
            Break
          EndIf
        ElseIf Not NextElement(results())
          ;add remainder of sublist() to end of results()
          MergeLists(subList(), results(), #PB_List_Last)
          Break 
        EndIf
      ForEver
    EndIf 
    
  Wend 
  CopyList(results(), a())
EndProcedure

Procedure.s listContents(List a())
  Protected output.s
  PushListPosition(a())
  ForEach a()
    output + Str(a()) + ","
  Next
  PopListPosition(a())
  ProcedureReturn Left(output, Len(output) - 1)
EndProcedure

Procedure setupList(List a())
  ClearList(a())
  Protected elementCount, i
  
  elementCount = Random(5) + 10
  For i = 1 To elementCount
    AddElement(a())
    a() = Random(10) - 5
  Next
EndProcedure


If OpenConsole()
  NewList sample()
  Define i
  
  For i = 1 To 3
    setupList(sample())
    PrintN("List " + Str(i) + ":")
    PrintN("  Before:  " + listContents(sample()))
    strandSort(sample())
    PrintN("  After :  " + listContents(sample()))
    PrintN("")
  Next
  
  Print(#CRLF$ + #CRLF$ + "Press ENTER to exit"): Input()
  CloseConsole()
EndIf

Sample output:

List 1:
  Before:  3,-2,-4,4,-1,-3,-2,-2,2,2,0
  After :  -4,-3,-2,-2,-2,-1,0,2,2,3,4

List 2:
  Before:  -4,4,3,-2,3,-2,5,0,-1,0,5,1
  After :  -4,-2,-2,-1,0,0,1,3,3,4,5,5

List 3:
  Before:  -2,0,-2,5,5,3,-1,-3,5,5,0,2,-4,4,2
  After :  -4,-3,-2,-2,-1,0,0,2,2,3,4,5,5,5,5

Python

def merge_list(a, b):
	out = []
	while len(a) and len(b):
		if a[0] < b[0]:
			out.append(a.pop(0))
		else:
			out.append(b.pop(0))
	out += a
	out += b
	return out

def strand(a):
	i, s = 0, [a.pop(0)]
	while i < len(a):
		if a[i] > s[-1]:
			s.append(a.pop(i))
		else:
			i += 1
	return s

def strand_sort(a):
	out = strand(a)
	while len(a):
		out = merge_list(out, strand(a))
	return out

print strand_sort([1, 6, 3, 2, 1, 7, 5, 3])

Output:

[1, 1, 2, 3, 3, 5, 6, 7]

Quackery

  [ [] swap
    1 split witheach
      [ over -1 peek
        over > iff
          [ swap dip join ]
        else join ] ]       is sift       (   [ --> [ [ )

  [ [] temp put
    [ dup  [] != while
      over [] != while
      over 0 peek
      over 0 peek
      > not if dip
        [ behead
          temp take
          swap join
          temp put ]
        again ]
      join
      temp take swap join ] is merge      ( [ [ --> [   )

  [ [] swap
    [ sift
      rot merge swap
      dup [] = until ]
    drop ]                  is strandsort (   [ --> [   )

  [] 25 times
    [ 89 random 10 + join ]
  say "Before: " dup echo cr
  strandsort
  say "After:  " echo cr

Output:

Before: [ 46 66 79 51 21 79 65 46 95 17 92 13 32 11 72 44 83 64 50 88 46 38 57 37 27 ]
After:  [ 11 13 17 21 27 32 37 38 44 46 46 46 50 51 57 64 65 66 72 79 79 83 88 92 95 ]

Racket

#lang racket
(require mzlib/list)
(define (merge xs ys) (merge-sorted-lists xs ys <=))

(define (strand-sort xs)
  (let loop ([xs xs] [ys '[]])
    (cond [(empty? xs) ys]
          [else (define-values (sorted unsorted) (extract-strand xs))
                (loop unsorted (merge sorted ys))])))

(define (extract-strand xs)
  (for/fold ([strand '()] [unsorted '[]]) ([x xs])
    (if (or (empty? strand) (< x (first strand)))
        (values (cons x strand) unsorted)
        (values strand (cons x unsorted)))))        

(strand-sort (build-list 10 (λ(_) (random 15))))

Raku

(formerly Perl 6)

Works with: Rakudo version 2018.04.01

sub infix:<M> (@x-in, @y-in) {
    my @x = | @x-in;
    my @y = | @y-in;
    flat @x, @y,
        reverse gather while @x and @y {
            take do given @x[*-1] cmp @y[*-1] {
                when More { pop @x }
                when Less { pop @y }
                when Same { pop(@x), pop(@y) }
            }
        }
}

sub strand (@x) {
    my $i = 0;
    my $prev = -Inf;
    gather while $i < @x {
        @x[$i] before $prev ?? $i++ !! take $prev = splice(@x, $i, 1)[0];
    }
}

sub strand_sort (@x is copy) {
    my @out;
    @out M= strand(@x) while @x;
    @out;
}

my @a = (^100).roll(10);
say "Before {@a}";
@a = strand_sort(@a);
say "After  {@a}";

@a = <The quick brown fox jumps over the lazy dog>;
say "Before {@a}";
@a = strand_sort(@a);
say "After  {@a}";

Output:

Before 1 20 64 72 48 75 96 55 42 74
After  1 20 42 48 55 64 72 74 75 96
Before The quick brown fox jumps over the lazy dog
After  The brown dog fox jumps lazy over quick the

REXX

This REXX program was written to generate a specified amount of random numbers as well as allowing a pre-pended list of items).

It can handle integers, floating point numbers, exponentiated numbers, and/or character strings.

/*REXX program sorts a random list of words (or numbers)              */
/* using the strand sort algorithm                                    */
Parse Arg size minv maxv old      /* obtain optional arguments from CL*/
if size=='' | size==","  then size=20    /*Not specified? use default.*/
if minv=='' | minv==","  then minv= 0    /*Not specified? use default.*/
if maxv=='' | maxv==","  then maxv=size  /*Not specified? use default.*/
Do i=1 To size
  old=old random(0,maxv-minv)+minv/* append random numbers to the list*/
  End
old=space(old)
Say 'Unsorted list:'
Say old
new=strand_sort(old)  /* sort given list (extended by random numbers) */
Say
Say 'Sorted list:'
Say new
Exit                           /* stick a fork in it,  we're all done */
/*--------------------------------------------------------------------*/
strand_sort: Procedure
  Parse Arg source
  sorted=''
  Do While words(source)\==0
    w=words(source)
    /* Find first word in source that is smaller Than its predecessor */
    Do j=1 To w-1
      If word(source,j)>word(source,j+1) Then
        Leave
      End
    /* Elements source.1 trough source.j are in ascending order       */
    head=subword(source,1,j)
    source=subword(source,j+1)     /* the rest starts with a smaller  */
                                   /* value or is empty (j=w!)        */
    sorted=merge(sorted,head)
    End
  Return sorted
/*--------------------------------------------------------------------*/
merge: Procedure
  Parse Arg a.1,a.2
  p=''
  Do Forever
    w1=words(a.1)
    w2=words(a.2)
    Select
      When w1==0 | w2==0 Then
        Return space(p a.1 a.2)
      When word(a.1,w1)<=word(a.2,1) Then
        Return space(p a.1 a.2)
      When word(a.2,w2)<=word(a.1,1) Then
        Return space(p a.2 a.1)
      Otherwise Do
        nn=1+(word(a.1,1)>=word(a.2,1))
        /* move the smaller first word of a.1 or a.2 to p */
        p=p word(a.nn,1)
        a.nn=subword(a.nn,2)
        End
      End
    End

output when using the input of: 25 -9 30 1000 2000 3000

────────────────────────────────unsorted list────────────────────────────────
1000 2000 3000 9 0 3 -8 17 8 -2 4 0 -3 19 -1 3 1 8 27 14 20 2 -6 23 1 -8 -4 4

─────────────────────────────────sorted list─────────────────────────────────
-8 -8 -6 -4 -3 -2 -1 0 0 1 1 2 3 3 4 4 8 8 9 14 17 19 20 23 27 1000 2000 3000

The REXX program can also sort words as well as numbers.

output when using the input of: 24 -9 100 66 66 8.8 carp Carp

──────────────────────────────────────unsorted list───────────────────────────────────────
66 66 8.8 carp Carp 20 77 88 9 39 -5 10 12 80 87 26 61 87 94 73 27 49 35 95 81 76 40 13 72

───────────────────────────────────────sorted list────────────────────────────────────────
-5 8.8 9 10 12 13 20 26 27 35 39 40 49 61 66 66 72 73 76 77 80 81 87 87 88 94 95 Carp carp

Note that an ASCII computer will sort words differently than an EBCDIC machine.

The order of sorting on an ASCII machine is: numbers, upperCase, lowerCase
The order of sorting on an EBCDIC machine is: lowerCase, upperCase, numbers

Ring

# Project : Sorting algorithms/Strand sort

test = [-2,0,-2,5,5,3,-1,-3,5,5,0,2,-4,4,2]
results = []
resultsend = []
see "before sort:" + nl
showarray(test)
test = strandsort(test)
see "after sort:" + nl
showarray(test)

func strandsort(a) 
        while len(a) > 0
                 sublist = []
                 add(sublist,a[1])
                 del(a,1)
                 for i = 1 to len(a)
                     if a[i] > sublist[len(sublist)] 
                       add(sublist,a[i])
                       del(a,i)
                     ok
                next
                for n = 1 to len(sublist)
                     add(results,sublist[n])
                next 
                for n = 1 to len(results)
                     for m = n + 1 to len(results)  
                          if results[m] < results[n]
                             temp = results[m]
                             results[m] = results[n]
                             results[n] = temp
                          ok 
                     next
                next                      
        end 
        return results

func showarray(vect)
        svect = ""
        for n = 1 to len(vect)
              svect = svect + vect[n] + " "
        next
        svect = left(svect, len(svect) - 1)
        see svect + nl

Output:

before sort:
-2 0 -2 5 5 3 -1 -3 5 5 0 2 -4 4 2
after sort:
-4 -3 -2 -2 -1 0 0 2 2 3 4 5 5 5 5

Ruby

class Array
  def strandsort
    a = dup
    result = []
    until a.empty?
      v = a.first
      sublist, a = a.partition{|val| v=val if v<=val}   # In case of v>val, it becomes nil.
      
      result.each_index do |idx|
        break if sublist.empty?
        result.insert(idx, sublist.shift) if sublist.first < result[idx]
      end
      result += sublist
    end
    result
  end
  
  def strandsort!
    replace(strandsort)
  end
end

p [1, 6, 3, 2, 1, 7, 5, 3].strandsort

Output:

[1, 1, 2, 3, 3, 5, 6, 7]

Sidef

Translation of: Perl

func merge(x, y) {
    var out = [];
    while (x && y) {
        given (x[-1] <=> y[-1]) {
            when ( 1) { out.prepend(x.pop) }
            when (-1) { out.prepend(y.pop) }
            default   { out.prepend(x.pop, y.pop) }
        }
    }
    x + y + out;
}

func strand(x) {
    x || return [];
    var out = [x.shift];
    if (x.len) {
        for i in (-x.len .. -1) {
            if (x[i] >= out[-1]) {
                out.append(x.pop_at(i));
            }
        }
    }
    return out;
}

func strand_sort(x) {
    var out = [];
    while (var strd = strand(x)) {
        out = merge(out, strd);
    }
    return out;
}

var a = 10.of { 100.irand };
say "Before: #{a}";
say "After: #{strand_sort(a)}";

Output:

Before: 24 62 29 95 11 21 46 3 23 20
After: 3 11 20 21 23 24 29 46 62 95

Tcl

proc merge {listVar toMerge} {
    upvar 1 $listVar v
    set i [set j 0]
    set out {}
    while {$i<[llength $v] && $j<[llength $toMerge]} {
	if {[set a [lindex $v $i]] < [set b [lindex $toMerge $j]]} {
	    lappend out $a
	    incr i
	} else {
	    lappend out $b
	    incr j
	}
    }
    # Done the merge, but will be one source with something left
    # This will handle all that by doing a merge of the remnants onto the end
    set v [concat $out [lrange $v $i end] [lrange $toMerge $j end]]
    return
}

proc strandSort A {
    set results {}
    while {[llength $A]} {
	set sublist [lrange $A 0 0]
	# We build a list of items that weren't filtered rather than removing "in place"
	# because this fits better with the way Tcl values work (the underlying data
	# structure is an array, not a linked list).
	set newA {}
	foreach a [lrange $A 1 end] {
	    if {$a > [lindex $sublist end]} {
		lappend sublist $a
	    } else {
		lappend newA $a
	    }
	}
	set A $newA
	merge results $sublist
    }
    return $results
}

puts [strandSort {3 1 5 4 2}]

Ursala

strand_sort "r" =  # parameterized by a relational predicate "r"

@NiX -+
   :-0 ~&B^?a\~&Y@a "r"?abh/~&alh2faltPrXPRC ~&arh2falrtPXPRC,
   ~&r->l ^|rlPlCrrPX/~& @hNCNXtX ~&r->lbx "r"?rllPXh/~&llPrhPlrPCXrtPX ~&rhPllPClrPXrtPX+-

demonstration code:

#cast %nL

x = (strand_sort nat-nleq) <3,1,5,4,2>

output:

<1,2,3,4,5>

Wren

Translation of: Kotlin

var merge = Fn.new { |left, right|
    var res = []
    while (!left.isEmpty && !right.isEmpty) {
        if (left[0] <= right[0]) {
            res.add(left[0])
            left.removeAt(0)
        } else {
            res.add(right[0])
            right.removeAt(0)
        }
    }
    res.addAll(left)
    res.addAll(right)
    return res
}

var strandSort = Fn.new { |a|
    var list = a.toList
    var res = []
    while (!list.isEmpty) {
        var sorted = [list[0]]
        list.removeAt(0)
        var leftover = []
        for (item in list) {
            if (sorted[-1] <= item) {
                sorted.add(item)
            } else {
                leftover.add(item)
            }
        }
        res = merge.call(sorted, res)
        list = leftover
    }
    return res
}

var a = [-2, 0, -2, 5, 5, 3, -1, -3, 5, 5, 0, 2, -4, 4, 2]
System.print("Unsorted: %(a)")
a = strandSort.call(a)
System.print("Sorted  : %(a)")

Output:

Unsorted: [-2, 0, -2, 5, 5, 3, -1, -3, 5, 5, 0, 2, -4, 4, 2]
Sorted  : [-4, -3, -2, -2, -1, 0, 0, 2, 2, 3, 4, 5, 5, 5, 5]

zkl

fcn strandSort(A){ //--> new list, A is cleared, should add A=A.copy()
   sublist:=List.createLong(A.len()); results:=List.createLong(A.len());
   while(A){
      sublist.clear(A.pop(0));
      foreach i in (A.len() - 1){
	 if(A[i]>sublist[-1]) sublist.append(A.pop(i));
      }
      results.merge(sublist);
   }
   results
}

The createLong list method creates a new list with pre-allocated space

strandSort(L(3,1,5,4,2)).println();
strandSort("azbfe".split("")).println();

Output:

L(1,2,3,4,5)
L("a","b","e","f","z")