# Sort the letters of string in alphabetical order

Write a function/program/subroutine/procedure to sort the characters of a string in lexicographical order.

Sort the letters of string in alphabetical order is a draft programming task. It is not yet considered ready to be promoted as a complete task, for reasons that should be found in its talk page.

Sorting Algorithm
This is a sorting algorithm.   It may be applied to a set of data in order to sort it.     For comparing various sorts, see compare sorts.   For other sorting algorithms,   see sorting algorithms,   or:

O(n logn) sorts

O(n log2n) sorts
Shell Sort

A character for this purpose should be whatever is natural for your language.

Show the results here on this page.   White-space may be optionally removed.

The case   (uppercase and lowercase)   of any letters should be preserved.

Write the function even if your language has a built-in function for it.

## 11l

```F sorted_str(s)
DefaultDict[Char, Int] d
L(c) s
d[c]++
V r = ‘’
L(k) sorted(d.keys())
r ‘’= k * d[k]
R r

print(sorted_str(‘The quick brown fox jumps over the lazy dog, apparently’))
print(sorted_str(‘Is this misspelling of alphabetical as alphabitical a joke ?’))```
Output:
```         ,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
?Iaaaaaaaabbcceeefghhhiiiiiijkllllllmnoopppsssssttt
```

## Action!

```INCLUDE "D2:SORT.ACT" ;from the Action! Tool Kit

PROC Test(CHAR ARRAY s)
PrintF("Original:%E""%S""%E%E",s)
SortB(s+1,s(0),0)
PrintF("Sorted:%E""%S""%E%E",s)
RETURN

PROC Main()
Put(125) PutE() ;clear the screen
Test("The quick brown fox jumps over the lazy dog, apparently")
Test("Now is the time for all good men to come to the aid of their country.")
RETURN```
Output:
```Original:
"The quick brown fox jumps over the lazy dog, apparently"

Sorted:
"         ,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz"

Original:
"Now is the time for all good men to come to the aid of their country."

Sorted:
"               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy"
```

```with Ada.Text_Io;

procedure Sort_Letters is

function Compare (Left, Right : Character) return Boolean
is (Left < Right);

procedure Sort is new
Element_Type => Character,
Array_Type   => String,
"<"          => Compare);

B : String := "When Roman engineers built a bridge, they had to stand under it while the first legion marched across. If programmers today worked under similar ground rules, they might well find themselves getting much more interested in Ada!";

begin
Put_Line (B); Sort (B); Put_Line (B);
end Sort_Letters;
```
Output:
```When Roman engineers built a bridge, they had to stand under it while the first legion marched across. If programmers today worked under similar ground rules, they might well find themselves getting much more interested in Ada!

## ALGOL 68

As with the Wren, Go and probably other samples, this defines a bubble sort to sort the text. Non-alphabetic characters are retained.

```BEGIN
# returns s with the characters sorted into lexicographic order          #
OP   LSORT = ( STRING s )STRING:
BEGIN
[ 1 : UPB s[ @ 1 ] ]CHAR c := s[ @ 1 ];
FOR u FROM UPB c - 1 BY -1 TO LWB c
WHILE
BOOL sorted := TRUE;
FOR p FROM LWB c BY 1 TO u DO
IF c[ p ] > c[ p + 1 ] THEN
CHAR t := c[ p ];
c[ p     ] := c[ p + 1 ];
c[ p + 1 ] := t;
sorted := FALSE
FI
OD;
NOT sorted
DO SKIP OD;
c
END; # SORT #
print( ( LSORT "The quick brown fox jumps over the lazy dog, apparently", newline ) )
print( ( LSORT "Now is the time for all good men to come to the aid of their country.", newline ) )
END```
Output:
```         ,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
```

## ALGOL W

Sorts the letters, leaving the non-letters unsorted.

```begin
% returns s with the letters sorted                           %
%         as Algol W doesn't have variable length strings,    %
%         the number of characters to sort must be specified  %
%         in len                                              %
string(256) procedure lSort( string(256) value s; integer value len ) ;
begin
string(256) letters, sortedS;
integer lLen, lPos, u;
sortedS := s;
% get the letters %
letters := "";
lLen    := -1;
for i := 0 until len - 1 do begin
string(1) c;
c := s( i // 1 );
if ( c >= "a" and c <= "z" ) or ( c >= "A" and c <= "Z" ) then begin
lLen := lLen + 1;
letters( lLen // 1 ) := c
end if_we_have_a_letter
end for_i ;
% bubble sort the letters %
u := lLen;
while begin
logical sorted;
sorted := true;
u      := u - 1;
for p := 0 until u do begin
string(1) c, d;
c := letters( p     // 1 );
d := letters( p + 1 // 1 );
if c > d then begin
letters( p     // 1 ) := d;
letters( p + 1 // 1 ) := c;
sorted                := false
end if_c_gt_d
end for_p ;
not sorted
end do begin end;
% put the sorted letters into the result string %
lPos := -1;
for i := 0 until len - 1 do begin
string(1) c;
c := s( i // 1 );
if ( c >= "a" and c <= "z" ) or ( c >= "A" and c <= "Z" ) then begin
lpos := lPos + 1;
sortedS( i // 1 ) := letters( lPos // 1 )
end if_we_have_a_letter
end for_i ;
sortedS
end lSort ;
% prints the first len characters of s %
procedure writeOnString( string(256) value s; integer value len ) ;
for i := 0 until len - 1 do writeon( s_w := 0, s( i // 1 ) );
% tests the lSort procedure %
procedure testSort( string(256) value s; integer value len ) ;
begin
writeon( s_w := 0, "    [" );writeOnString(        s, len );        writeon( "]" );write();
writeon( s_w := 0, " -> [" );writeOnString( lSort( s, len ), len ); writeon( "]" );write()
end testSort ;

testSort( "The quick brown fox jumps over the lazy dog, apparently", 55 );
testSort( "Now is the time for all good men to come to the aid of their country.", 69 );
testSort( "Stop! In the name of Wuv!", 25 )
end.```
Output:
```    [The quick brown fox jumps over the lazy dog, apparently]
-> [Taa abcde eeefg hhi jkllm nnoo oop ppqr rrs, ttuuvwxyyz]
[Now is the time for all good men to come to the aid of their country.]
-> [Naa cc dde eeee eff ghh hiii ill mm mnno oo ooo ooo rr rsttt ttttuwy.]
[Stop! In the name of Wuv!]
-> [ISWa! ee fhm nnoo pt tuv!]```

## APL

Works with: Dyalog APL
```sort ← ⊂∘⍋⌷⊣
```
Output:
```      sort 'Now is the time for all good men to come to the aid of their country.'
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## AutoHotkey

```sortLetters(str, RemoveSpace := 1){
oChar := []
for i, v in StrSplit(str)
if (v <> " ") && RemoveSpace
oChar[Asc(v), i] := v
else if !RemoveSpace
oChar[Asc(v), i] := v

for ascii, obj in oChar
for i, letter in obj
result .= letter
return result
}
```
Examples:
```str1 := "The quick brown fox jumps over the lazy dog, apparently"
str2 := "Now is the time for all good men to come to the aid of their country."

MsgBox, 262144, , % result := str1 " ->`n" sortLetters(str1)
. "`n`n" str2 " ->`n" sortLetters(str2, 0)
```
Output:
```The quick brown fox jumps over the lazy dog, apparently ->
,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz

Now is the time for all good men to come to the aid of their country. ->
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## AWK

```# syntax GAWK -f SORT_THE_LETTERS_OF_STRING_IN_ALPHABETICAL_ORDER.AWK
BEGIN {
str = "Now is the time for all good men to come to the aid of their country."
printf("old: %s\n",str)
printf("new: %s\n",sortstr(str))
exit(0)
}
function sortstr(str,  i,j) {
for (i=2; i<=length(str); i++) {
for (j=i; j>1 && substr(str,j-1,1) > substr(str,j,1); j--) {
#             < left          > < these are swapped             > < right       >
str = substr(str,1,j-2) substr(str,j,1) substr(str,j-1,1) substr(str,j+1)
}
}
return(str)
}
```
Output:
```old: Now is the time for all good men to come to the aid of their country.
new:                .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
```

## BCPL

```get "libhdr"

let sortchars(str) be
\$(  let count = vec 255 and loc = ?
for i=0 to 255 do i!count := 0
for i=1 to str%0 do (str%i)!count := (str%i)!count + 1
loc := 1
for i=0 to 255 until i!count = 0
\$(  str%loc := i
loc := loc + 1
i!count := i!count - 1
\$)
\$)

let start() be
\$(  let string =
"Now is the time for all good men to come to the aid of their country."
writef("%S*N", string)
sortchars(string)
writef("%S*N", string)
\$)```
Output:
```Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## C

```#include <stdio.h>

/* Sort a character string in place */
void strsort(char *s) {
unsigned int n[256] = {0};
unsigned char i = 0;
char *t = s;
while (*s) ++n[(unsigned char) *s++];
while (++i) while (n[i]--) *t++ = (char) i;
}

int main() {
char s[] = "Now is the time for all good men "
"to come to the aid of their country.";
puts(s);
strsort(s);
puts(s);
return 0;
}
```
Output:
```Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## C++

```#include <algorithm>
#include <iostream>

int main() {
std::string s = "Now is the time for all good men "
"to come to the aid of our country.";

std::cout << s << std::endl;
std::sort(s.begin(), s.end());
std::cout << s << std::endl;
return 0;
}
```
Output:
```Now is the time for all good men to come to the aid of our country.
.Naaccddeeeeeffghhiiillmmmnnoooooooooorrrsttttttuuwy```

## C#

Dubbing the following sorting method as "Slacksort". This "Slacksort" method can easily be adapted for reverse sorting, or removing other characters besides space. Not recommended for larger strings though.

```using System; using static System.Console;
class Program {
static void Main(string[] args) {
var nl = "\n";
var omit_spaces = true;
var str = "forever ring programming language";
Write( "working..." + nl );
Write( "Sort the letters of string in alphabitical order:" + nl );
Write( "Input: " + str + nl );
Write( "Output: " );
for (var ch = omit_spaces ? 33 : 0; ch < 256; ch++)
foreach (var itm in str)
if (ch == itm) Console.Write(itm);
Write( nl + "done..." );
}
}
```

Note: this is a bit of a tribute to the original task description and initial Ring entry, so the typographical errors have intentionally not been corrected.

Output:
```working...
Sort the letters of string in alphabitical order:
Input: forever ring programming language
Output: aaaeeefgggggiilmmnnnooprrrrruv
done...
```

## CLU

```% Unicode is explicitly not supported, the standard says
% that "every implementation must provide at least 128,
% but no more than 512, characters".
% That means we can do it in O(N) using a counting sort.

sort_string = proc (s: string) returns (string)
char_count: array[int] := array[int]\$fill(0,512,0)
for c: char in string\$chars(s) do
i: int := char\$c2i(c)
char_count[i] := char_count[i]+1
end

sorted_chars: array[char] := array[char]\$predict(1,string\$size(s))
for i: int in array[int]\$indexes(char_count) do
for j: int in int\$from_to(1,char_count[i]) do
end
end

return(string\$ac2s(sorted_chars))
end sort_string

start_up = proc ()
po: stream := stream\$primary_output()
str: string := "Now is the time for all good men to come to the aid of their country."

stream\$putl(po, str)
stream\$putl(po, sort_string(str))
end start_up```
Output:
```Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## Comal

```0010 PROC strsort(REF s\$) CLOSED
0020   DIM count#(0:255)
0030   FOR i#:=1 TO LEN(s\$) DO count#(ORD(s\$(i#))):+1
0040   i#:=1
0050   FOR j#:=0 TO 255 DO
0060     IF count#(j#)>0 THEN
0070       s\$(i#:i#+count#(j#)-1):=CHR\$(j#)*count#(j#)
0080       i#:+count#(j#)
0090     ENDIF
0100   ENDFOR j#
0110 ENDPROC strsort
0120 //
0130 test\$:="Now is the time for all good men "
0140 test\$:+"to come to the aid of their country."
0150 PRINT test\$
0160 strsort(test\$)
0170 PRINT test\$
0180 END```
Output:
```Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## Common Lisp

```(sort "Now is the time for all good men to come to the aid of their country." #'char<=)
```
Output:
```"               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy"
```

## Draco

```/* Sort a string in place, using a counting sort. */
proc nonrec stringsort(*char str) void:
[256] word counts;
byte i;
word j;
char c;
channel input text chi;
channel output text cho;

/* zero array */
for i from 0 upto 255 do counts[i] := 0 od;

/* count all characters */
open(chi, str);
while read(chi; c) do counts[c] := counts[c] + 1 od;
close(chi);

/* write the characters back in order */
open(cho, str);
for i from 0 upto 255 do
for j from 1 upto counts[i] do
write(cho; pretend(i, char))
od
od;
close(cho)
corp

/* Test */
proc nonrec main() void:
*char s;
s := "Now is the time for all good men to come to the aid of their country.";

writeln(s);
stringsort(s);
writeln(s)
corp```
Output:
```Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## Eiffel

```class
SORT_STRING_LETTERS

feature -- Basic Ops

sort_string (s: STRING): STRING
-- Perform `sort_string' on `s' such that
--	each letter is in ascending alphabetical order.
note
deviation: "[
This Eiffel example deviates from the task requirement for this
Rosetta Code task in that we reuse Eiffel Base library code for
the {SORTED_TWO_WAY_LIST [G]}. We do this for two very good reasons:

1. Reuse is king. Never code what is already coded and tested.
2. The library code is readily available for examination
(i.e. the library code is not hidden and unaccessible).

Based on #1 and #2 above, examine the code in: {SORTED_TWO_WAY_LIST}.make_from_iterable
Specifically, look at the `extend' routine and the routines it calls (i.e. `search_after',
`put_left', and `back'). These routines will tell you the story of how sorting can
be coded in Eiffel. There is no need to rehash that code here.
]"
local
l_list: SORTED_TWO_WAY_LIST [CHARACTER]
do
create l_list.make_from_iterable (s)			-- Add & Auto-sort string by chars
create Result.make_empty				-- Create the Result STRING
⟳ c:l_list ¦ Result.append_character (c) ⟲		-- Populate it with the sorted chars
ensure
no_spaces: ∀ c:Result ¦ c /= ' '			-- All spaces removed.
has_all: ∀ c:Result ¦ s.has (c)				-- All non-space chars present.
valid_count: Result.count =
(s.count - s.occurrences (' '))		-- Every character (even repeating).
end

end
```

Notice the use of Design-by-Contract in the "ensure" at the end of `sort_string'. At testing runtime, we want the routine itself to ensure that the resulting string has no space character and that every character that we passed in the `s' argument is represented in the result string. We even go so far as to ensure that repeating characters are all represented. We could go further, but we felt these contracts were sufficient to get the point of Design-by-Contract across to you as the reader.

Also note the unicode characters of ⟳ .. ⟲ and ∀. These are called across-loops and we have used the "symbolic form" of them. The first one simply says, "Across all items in l_list as c, perform the code between the ¦ and closing ⟲. In the case of the ∀, this is a loop stating that every member of the Result STRING must hold true for the Boolean expression. For example, in the "no_spaces" contract, we are saying that each character "c" must not be equal to a space (e.g. ' ').

And the Test Code to operate it.

```class
RC_SORT_STRING_LETTERS_TEST_SET

inherit
TEST_SET_SUPPORT

feature -- Test routines

rc_sort_string_letters_test
-- Test {SORT_STRING_LETTERS}.
note
testing:
"covers/{SORT_STRING_LETTERS}",
"execution/isolated",
"execution/serial"
do
assert_strings_equal ("sorted", now_is_string, item.sort_string ("Now is the time for all good men to come to the aid of their country."))
end

feature {NONE} -- Test Support

now_is_string: STRING = "[
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
]"

item: SORT_STRING_LETTERS
-- An `item' for testing.
once
create Result
end

end
```

## F#

```// Alphabetic sort. Nigel Galloway: July 27th., 2021
let fG n g=let g=g|>Seq.countBy id|>Map.ofSeq in [for n in n->if Map.containsKey n g then [|for g in 1..g.[n]->n|]|>System.String else ""]|>String.concat ""
let English=fG ['a';'A';'b';'B';'c';'C';'d';'D';'e';'E';'f';'F';'g';'G';'h';'H';'i';'I';'j';'J';'k';'K';'l';'L';'m';'M';'n';'N';'o';'O';'p';'P';'q';'Q';'r';'R';'s';'S';'t';'T';'u';'U';'v';'V';'w';'W';'x';'X';'y';'Y';'z';'Z']
let Turkish=fG ['a';'A';'b';'B';'c';'C';'ç';'Ç';'d';'D';'e';'E';'f';'F';'g';'G';'ğ';'Ğ';'h';'H';'ı';'I';'i';'İ';'j';'J';'k';'K';'l';'L';'m';'M';'n';'N';'o';'O';'ö';'Ö';'p';'P';'r';'R';'s';'S';'ş';'Ş';'t';'T';'u';'U';'ü';'Ü';'v';'V';'y';'Y';'z';'Z'];
let main args=use n=new System.IO.StreamWriter(System.IO.File.Create("out.txt"))
n.WriteLine(English "baNAnaBAnaNA")
n.WriteLine(Turkish (String.filter((<>)' ') "Meseleyi anlamağa başladı"))
```
Output:
```aaaAAAbBnnNN
aaaaaabdeeeğıilllmMnsşy
```

## FreeBASIC

```function value( s as string ) as integer
'maps a=A=0, b=B=1, etc
return asc(ucase(s))-65
end function

sub sortstr( ins as string, outs as string )
dim as string c
dim as integer cv, i, j
outs = ""
for i = 1 to len(ins)
c = mid(ins,i,1)
cv = value(c)
if cv > 25 or cv < 0 then continue for   'this isn't a letter so don't output it
for j = 1 to len(outs)
if value(mid(outs,j,1))>cv then exit for
next j
outs = left(outs,j-1) + c + right(outs,len(outs)-j+1)
next i
end sub

dim as string example = "Once upon a midnight dreary as I pondered weak and weary over many a dumb and buggy line of irritating code"
dim as string sorted

sortstr( example, sorted )
print sorted
```
Output:
`aaaaaaaaaabbccddddddddeeeeeeeeefgggghiiIiiiiklmmmnnnnnnnnnOooooopprrrrrrrstttuuuvwwyyyy`

## FutureBasic

```local fn AlphabetizeString( string as CFStringRef, deleteSpaces as BOOL ) as CFStringRef
NSUInteger  i, count
CFStringRef tempStr, alphaStr = NULL
CFArrayRef  array

if deleteSpaces then tempStr = fn StringByReplacingOccurrencesOfString( string, @" ", @"" ) else tempStr = string
count = fn StringLength( tempStr )

CFMutableArrayRef mutArr = fn MutableArrayWithCapacity(count)
for i = 0 to count -1
CFStringRef chr = fn StringWithFormat( @"%C", fn StringCharacterAtIndex( tempStr, i ) )
MutableArrayInsertObjectAtIndex( mutArr, chr, i )
next
array = fn ArraySortedArrayUsingSelector( mutArr, @"localizedCaseInsensitiveCompare:" )
alphaStr = fn ArrayComponentsJoinedByString( array, @"" )
end fn = alphaStr

CFStringRef testStr

testStr = @"The quick brown fox jumped over the lazy dog's back."
print testStr
print "String alphabetized with spaces included:"
print fn AlphabetizeString( testStr, NO )
print
testStr = @"Now is the time for all good men to come to the aid of their country."
print testStr
print "String alphabetized with spaces deleted:"
print fn AlphabetizeString( testStr, YES )

HandleEvents```
Output:
```The quick brown fox jumped over the lazy dog's back.
String alphabetized with spaces included:
.'aabbccddeeeefghhijkklmnoooopqrrsTtuuvwxyz

Now is the time for all good men to come to the aid of their country.
String alphabetized with spaces deleted:
.aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy
```

## Go

As in the case of the Wren entry, we write a function to bubble sort the characters of a string since this method is not, of course, used in Go's standard 'sort' package.

```package main

import (
"fmt"
"strings"
)

func bubbleSort(s string, trim bool) string { // allow optional removal of whitespace
chars := []rune(s)
n := len(chars)
for {
n2 := 0
for i := 1; i < n; i++ {
if chars[i-1] > chars[i] {
tmp := chars[i]
chars[i] = chars[i-1]
chars[i-1] = tmp
n2 = i
}
}
n = n2
if n == 0 {
break
}
}
s = string(chars)
if trim {
s = strings.TrimLeft(s, " \t\r\n")
}
return s
}

func main() {
ss := []string{
"forever go programming language",
"Now is the time for all good men to come to the aid of their country.",
}
trims := []bool{true, false}
for i, s := range ss {
res := bubbleSort(s, trims[i])
fmt.Printf("Unsorted->%s\n", s)
fmt.Printf("Sorted  ->%s\n\n", res)
}
}
```
Output:
```Unsorted->forever go programming language
Sorted  ->aaaeeefgggggilmmnnoooprrrruv

Unsorted->Now is the time for all good men to come to the aid of their country.
Sorted  ->               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
```

```import Data.List (sort)

main :: IO ()
main =
print \$
sort
"Is this misspelling of alphabetical as alphabitical a joke ?"
```
Output:
`"         ?Iaaaaaaaabbcceeefghhhiiiiiijkllllllmnoopppsssssttt"`

Or, sketching a rough re-phrase of the question:

```import Data.List (partition)

main :: IO ()
main =
print \$
qSort
"Is this misspelling of alphabetical as alphabitical a joke ?"

qSort :: (Ord a) => [a] -> [a]
qSort [] = []
qSort (x : xs) = qSort below <> (x : qSort above)
where
(below, above) = partition (<= x) xs
```
Output:
`"         ?Iaaaaaaaabbcceeefghhhiiiiiijkllllllmnoopppsssssttt"`

Or, just constructing a sorted string from the character counts:

```import qualified Data.Map.Strict as M

----------------- MAP OF CHARACTER COUNTS ----------------

charCounts :: String -> M.Map Char Int
charCounts =
foldr (flip (M.insertWith (+)) 1) M.empty

--------------------------- TEST -------------------------
main :: IO ()
main =
( print
. (uncurry (flip replicate) =<<)
. M.toList
. charCounts
)
"Was the misspelling of alphabetical as alphabitical a joke ?"
```
Output:
`"         ?Waaaaaaaaabbcceeeefghhhiiiiijkllllllmnoopppssssttt"`

## J

J's builtin is 'grade' and sort is a derived function whose domain includes sequences of characters, so:
```   text0=: 'This is a test'
text1=: 'The sentence "The quick brown fox jumps over the lazy dog" uses every letter in the alphabet.'
/:~ text0
Taehiissstt
/:~ text1
```
However, sorting characters is easy to implement using bucket sort:
```   {{a.#~<:#/.~a.,y}} text0
Taehiissstt
{{a.#~<:#/.~a.,y}} text1
```

(Since there's no comparison between pairs in bucket sort, performance here is O(n) rather than O(n log n).)

## jq

Works with: jq

Works with gojq, the Go implementation of jq

An efficient way to sort an arbitrary JSON string is to use the code points of the constituent characters:

```def sort_by_codepoints:
explode | sort | implode;```

For example:

```"Is this misspelling of alphabetical as alphabitical a joke ?"
| sort_by_codepoints```
produces
`"         ?Iaaaaaaaabbcceeefghhhiiiiiijkllllllmnoopppsssssttt"`
An alternative definition using `sort` on the characters themselves:
```def sort_by_characters:
explode | map([.]|implode) | sort | add;
```
Are these definitions the same?
```def dingbats:
"✁✂✃✄✆✇✈✉✌✍✎✏✐✑✒✓✔✕✖✗✘✙✚✛✜✝✞✟✠✡✢✣✤✥✦✧✩✪✫✬✭✮✯✰✱✲✳✴✵✶✷✸✹✺✻✼✽✾✿❀❁❂❃❄❅❆❇❈❉❊❋❍❏❐❑❒❖❘❙❚❛❜❝❞❡❢❣❤❥❦❧❶❷❸❹❺❻❼❽❾❿➀➁➂➃➄➅➆➇➈➉➊➋➌➍➎➏➐➑➒➓➔➘➙➚➛➜➝";

"Now is the time for all good men to come to the aid of their country.",
"Is this misspelling of alphabetical as alphabitical a joke ?",
dingbats
| (sort_by_codepoints==sort_by_characters)```
produces
```true
true
true
```

## Julia

```function mergesort!(array, lt = <, low = 1, high = length(array), tmp=similar(array, 0))
high <= low && return array
middle = low + div(high - low, 2)
(length(tmp) < middle - low + 1) && resize!(tmp, middle - low + 1)

mergesort!(array, lt, low,  middle, tmp)
mergesort!(array, lt, middle + 1, high, tmp)

i, j = 1, low
while j <= middle
tmp[i] = array[j]
i += 1
j += 1
end

i, k = 1, low
while k < j <= high
if lt(array[j], tmp[i])
array[k] = array[j]
j += 1
else
array[k] = tmp[i]
i += 1
end
k += 1
end

while k < j
array[k] = tmp[i]
k += 1
i += 1
end
return array
end

mergesort(str::String) = String(mergesort!(collect(str)))

function testmergesort(s::String, stripws= true)
println("Unsorted -> ", s)
println("Sorted   -> ", stripws ? strip(mergesort(s)) : mergesort(s))
end

testmergesort("forever julia programming language")
testmergesort("Now is the time for all good men to come to the aid of their country.", false)
```
Output:
```Unsorted -> forever julia programming language
Sorted   -> aaaaeeefggggiijllmmnnooprrrruuv
Unsorted -> Now is the time for all good men to come to the aid of their country.
Sorted   ->                .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
```

### case insensitive quicksort

```function qsort(array)
length(array) < 2 && return array
mid, left, right  = first(array), eltype(array)[], eltype(array)[]
for elem in @view array[begin+1:end]
push!(lowercase(elem) < lowercase(mid) ? left : right, elem)
end
return vcat(qsort(left), mid, qsort(right))
end

qsort(str::String) = str |> collect |> qsort |> String

function testqsort(s::String, stripws= true)
println("Unsorted -> ", s)
println("Sorted   -> ", stripws ? strip(qsort(s)) : qsort(s))
end

testqsort("forever julia programming language")
testqsort("Now is the time for all good men to come to the aid of their country.")
```
Output:
```Unsorted -> forever julia programming language
Sorted   -> aaaaeeefggggiijllmmnnooprrrruuv
Unsorted -> Now is the time for all good men to come to the aid of their country.
Sorted   ->                .aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy
```

### alphabet-only counting version

```const alphabets = "AaBbCcDdEeFfGgHhIiJjKkLlMmNnOoPpQqRrSsTtUuVvWwXxYyZz"

function lettercountingsort(s)
sorted = Char[]
for l in alphabets
append!(sorted, fill(l, count(==(l), s)))
end
return String(sorted)
end

println(lettercountingsort("Now is the time for all good men to come to the aid of their country."))
```
Output:
```aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy
```

## Lua

```fcoll = {} -- forward collation
sl = string.lower -- for case insensitivity
for i=0,255 do fcoll[i]=string.char(i) end -- initially just ASCII (for non-letters)
table.sort(fcoll, function(a,b) return sl(a)==sl(b) and a<b or sl(a)<sl(b) end) -- interleave upper/lower letters
rcoll = {} for i,v in ipairs(fcoll) do rcoll[v]=i end -- reverse collation

function sort(s) -- Latin letters lexicographically, uppercase first, anything else by ASCII
local t={} s:gsub("(%S)", function(c) t[#t+1]=c end) -- use "(.)" as pattern to preserve whitespace
table.sort(t, function(a,b) return rcoll[a]<rcoll[b] end)
return table.concat(t)
end

print(sort("Now is the time for all good men to come to the aid of their country."))
```
Output:
`.aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy`

Concise version, implicit rather than explicit collation sequence table, adequate for this use, same output:

```function sort(s) -- Latin letters lexicographically, uppercase first, anything else by ASCII
local sl,t=string.lower,{} s:gsub("(%S)", function(c) t[#t+1]=c end) -- use "(.)" as pattern to preserve whitespace
table.sort(t, function(a,b) return sl(a)==sl(b) and a<b or sl(a)<sl(b) end) -- implicitly
return table.concat(t)
end

print(sort("Now is the time for all good men to come to the aid of their country."))
```

## Mathematica/Wolfram Language

```sortstring = Characters /* LexicographicSort /* StringJoin;
sortstring["Now is the time for all good men to come to the aid of their country."]
```
Output:
`".               aaccddeeeeeeffghhhiiiillmmmnnNooooooooorrrstttttttuwy"`

## Nim

```import strutils, tables

func sorted(text: string; omitSpaces = false): string =
let count = text.toCountTable()
for c in '\0'..'\255':
if c == ' ' and omitSpaces: continue

echo sorted("The quick brown fox jumps over the lazy dog, apparently", false)
echo sorted("Now is the time for all good men to come to the aid of their country.", true)
```
Output:
```         ,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy```

## Pascal

Works with: Extended Pascal

In Pascal it is guaranteed that all letters of the Modern English alphabet, 'A' through 'Z' and 'a' through 'z', are in alphabetical order. That means ord('A') < ord('B'). However, uppercase and lowercase letters may be separate (like in ASCII), fully or partially interspersed. The output of this program may differ in that regard.

```program sortTheLettersOfStringInAlphabeticalOrder(input, output);

type
line = string(80);

{
sort characters in a string

\param sample the string to sort
\return \param sample so for all n < m: sample[n] <= sample[m]
}
function alphabeticallySorted(sample: line): line;
var
c: char;
{ `sample.capacity` refers to 80 in this program. }
i: 0..sample.capacity;
{ `… value [otherwise 0]` is an initial state specification. }
tab: array[char] of 0..sample.capacity value [otherwise 0];
begin
{ analyze: how many occurrences of every character? }
for i := 1 to length(sample) do
begin
tab[sample[i]] := tab[sample[i]] + 1
end;

{ process: rebuild string but in alphabetical order }
sample := '';

for c := chr(0) to maxChar do
begin
for i := 1 to tab[c] do
begin
sample := sample + c
end
end;

{ finish: set result variable }
alphabeticallySorted := sample
end;

{ === MAIN =================================================== }
var
s: line;
begin
while not EOF do
begin
writeLn(alphabeticallySorted(s))
end
end.
```

## Perl

```#!/usr/bin/perl -l

use strict; # https://rosettacode.org/wiki/Sort_the_letters_of_string_in_alphabitical_order
use warnings;

my @lines = split /\n/, <<END;
The quick brown fox jumps over the lazy dog, apparently
Now is the time for all good men to come to the aid of their country.
END

for ( @lines, 'dcba', 'sort this string' )
{
print "\n\$_";
print builtinsort(\$_); #     using built in sort
print sortstring(\$_);  # not using built in sort
print inplace(\$_);     # not using built in sort
}

sub builtinsort
{
return join '', sort split //, shift;
}

sub sortstring # IBM card sorters forever !! (distribution sort)
{
my @chars;
\$chars[ord] .= \$_ for split //, shift;
no warnings; # hehehe
return join '', @chars;
}

sub inplace # just swap any adjacent pair not in order until none found
{
local \$_ = shift;
1 while s/(.)(.)(??{\$1 le \$2 && '(*FAIL)'})/\$2\$1/g;
return \$_;
}
```
Output:
```
The quick brown fox jumps over the lazy dog, apparently
,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz

Now is the time for all good men to come to the aid of their country.
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
.Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy

dcba
abcd
abcd
abcd

sort this string
```

## Phix

Not sure this algorithm actually has a name, but it certainly ain't the fastest, though it possibly is just about the shortest...
(If pressed I would dub this "Unoptimised bubble sort without the swapped flag")

```with javascript_semantics
function string_sort(string s)
integer temp
for n=1 to length(s)-1 do
for m=n+1 to length(s) do
if s[n]>s[m] then
temp = s[n]
s[n] = s[m]
s[m] = temp
end if
end for
end for
return s
end function
string s = "Now is the time for all good men to come to the aid of their country."
printf(1,"Original:\"%s\",\n  Sorted:\"%s\"\n Builtin:\"%s\"\n",{s,string_sort(s),sort(s)})
```
Output:
```Original:"Now is the time for all good men to come to the aid of their country.",
Sorted:"               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy"
Builtin:"               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy"
```

### case insensitive

You can make this case insensitive by applying lower() on each internal comparison, whereas with the builtins that is done (more efficiently) by extracting a custom tagsort.
(Just to keep you on your toes I've also replaced the algorithm with a fractionaly saner insertion sort, and just to be awkward I've added the baNAnaBAnaNA case.)

```with javascript_semantics
function string_sort(string s)
for i=2 to length(s) do
integer j = i, sj = s[j]
while j>=2 and lower(sj)<lower(s[j-1]) do
s[j] = s[j-1]
j -= 1
end while
s[j] = sj
end for
return s
end function

procedure test(string s)
string cia = extract(s,custom_sort(lower(s),tagset(length(s))))
printf(1,"Original:\"%s\",\n  Sorted:\"%s\"\n Builtin:\"%s\"\n",{s,string_sort(s),cia})
end procedure
test("Now is the time for all good men to come to the aid of their country.")
test("baNAnaBAnaNA") -- (just to be awkward)
```
Output:
```Original:"Now is the time for all good men to come to the aid of their country.",
Sorted:"               .aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy"
Builtin:"               .aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy"
Original:"baNAnaBAnaNA",
Sorted:"aAaAaAbBNnnN"
Builtin:"aAaAaAbBNnnN"
```

Should you want/prefer the output of baNAnaBAnaNA to be AAAaaaBaNNnn, change the test (leaving the builtin/cia as an exercise) to

```        while j>=2 and (lower(sj)<lower(s[j-1]) or sj=s[j-1]-32) do
```

Or of course for aaaAAAbBnnNN use

```        while j>=2 and (lower(sj)<lower(s[j-1]) or sj-32=s[j-1]) do
```

## Python

```'''Sorted string'''

from functools import reduce

# qSort :: [a] -> [a]
def qSort(xs):
'''Sorted elements of the list xs, where the values
of xs are assumed to be of some orderable type.
'''
if xs:
h = xs[0]
below, above = partition(
lambda v: v <= h
)(xs[1:])

return qSort(below) + [h] + qSort(above)
else:
return []

# ------------------------- TEST -------------------------
def main():
'''A character-sorted version of a test string
'''
print(quoted('"')(
''.join(qSort(list(
"Is this misspelling of alphabetical as alphabitical a joke ?"
)))
))

# ----------------------- GENERIC ------------------------

# partition :: (a -> Bool) -> [a] -> ([a], [a])
def partition(p):
'''The pair of lists of those elements in xs
which respectively do, and don't
satisfy the predicate p.
'''
def go(a, x):
ts, fs = a
return (ts + [x], fs) if p(x) else (ts, fs + [x])
return lambda xs: reduce(go, xs, ([], []))

# quoted :: Char -> String -> String
def quoted(c):
'''A string flanked on both sides
by a specified quote character.
'''
return lambda s: c + s + c

# MAIN ---
if __name__ == '__main__':
main()
```
Output:
`"         ?Iaaaaaaaabbcceeefghhhiiiiiijkllllllmnoopppsssssttt"`

## Raku

### Semi-realistic version

```sub sort_within_string ( \$_ is copy ) {
constant @lexographic_order = sort *.fc, map &chr, 1..255;

return join '', gather for @lexographic_order -> \$l {
my \$count = s:g/\$l//;
take \$l x \$count;
LAST { warn "Original string had non-ASCII chars: {.raku}" if .chars }
}
}
say trim .&sort_within_string for q:to/END/.lines;
The quick brown fox jumps over the lazy dog, apparently
Now is the time for all good men to come to the aid of their country.
END
```
Output:
```,aaabcdeeeefghhijkllmnnoooopppqrrrsTttuuvwxyyz
.aaccddeeeeeeffghhhiiiillmmmNnnooooooooorrrstttttttuwy
```

### Following the actual task title / description

Following a strict interpretation of the actual task title and description.

• Sorts letters. Only letters. Into alphabetical order. Regardless of case. EVERYTHING else is ignored / pushed to the end of the "sorted" string. Not ASCII order. Not EBCDIC order. Only alphabetical order. If it ain't in the alphabet, it ain't sorted.
• Sorts letters of the string two characters at a time as a string. No breaking up the string into a list or array, sorting that then joining back together; or picking characters out of a string to generate a new string. Sorts a string, as a string, in place.

Sorted output is wrapped in double guillemots to make it easier to see where it starts and ends.

```sub moronic-sort (\$string is copy) {
my \$chars = \$string.chars;
loop {
for ^\$chars {
if (\$string.substr(\$_, 1).fc gt \$string.substr(\$_ + 1, 1).fc and \$string.substr(\$_ + 1, 1) ~~ /<:L>/)
or \$string.substr(\$_, 1) ~~ /<:!L>/ {
\$string = \$string.substr(0, \$_) ~ \$string.substr(\$_ , 2).flip ~ \$string.substr(\$_ + 2 min \$chars);
}
}
last if \$++ >= \$chars;
}
\$string
}

sub wrap (\$whatever) { '»»' ~ \$whatever ~ '««' }

# Test sort the exact string as specified in the task title.
say "moronic-sort 'string'\n" ~ wrap moronic-sort 'string';

# Other tests demonstrating the extent of the stupidity of this task.
say "\nLonger test sentence\n" ~
wrap moronic-sort q[This is a moronic sort. It's only concerned with sorting letters, so everything else is pretty much ignored / pushed to the end. It also doesn't much care about letter case, so there is no upper / lower case differentiation.];

say "\nExtended test string:\n" ~ my \$test = (32..126)».chr.pick(*).join;
say wrap moronic-sort \$test;
```
Output:
```moronic-sort 'string'
»»ginrst««

Longer test sentence
»»aaaaaaabccccccccddddddeeeeeeeeeeeeeeeeeeeeeeeeeffggghhhhhhhhiiiIiiiiiIiiiillllllmmmnnnnnnnnnnnnoooooooooooooooopppprrrrrrrrrrrrrrssssssssssssssssTtttttttttttttttttttuuuuuvwwyyy    ,      /   .    . '     ,        /    .   ' ««

Extended test string:
!kjyxAa+,LGh_8?3lXEwW-D]Ku|SY[@VF\.op{=q>MT 1tJ/\$nN(Z*%&9^v57")`PCiOHQe'RUb<gs;6}#cfmrzd42B~0I:

## REXX

For REXX, it is normally faster to convert a string of characters to a one─character array of characters,
sort the array,   and then convert the array back to a (simple) string.

A simple bubble sort is used for this example.

The particular string used is from a typing drill devised by Charles E. Weller in the early 20th century.

```/*REXX program sorts an array  (of any kind of items)  using the  bubble─sort algorithm.*/
parse arg y                                      /*generate the array elements  (items).*/
if y=''  then y= "Now is the time for all good men to come to the aid of their country."
say 'before sort: ───►'y"◄───"      /*show the  before string of characters*/
call make@    y                                  /*convert a string into an array  (@.) */
call bSort    #                                  /*invoke the bubble sort  with # items.*/
call makeS                                       /*convert an array (@.) into a string. */
say ' after sort: ───►'\$"◄───"      /*show the  before string of characters*/
exit 0                                           /*stick a fork in it,  we're all done. */
/*──────────────────────────────────────────────────────────────────────────────────────*/
bSort: procedure expose @.;  parse arg n         /*N: is the number of @ array elements.*/
do m=n-1  by -1  until ok;        ok= 1 /*keep sorting the  @ array until done.*/
do j=1  for m;  k= j+1;  if @.j<=@.k  then iterate      /*elements in order? */
_= @.j;  @.j= @.k;  @.k= _;     ok= 0 /*swap two elements;  flag as not done.*/
end   /*j*/
end     /*m*/;               return
/*──────────────────────────────────────────────────────────────────────────────────────*/
make@: parse arg z;  #= length(z);    do j=1  for #;  @.j= substr(z, j, 1);  end;   return
makeS: parse arg a;  \$=;              do j=1  for #;  \$= \$  ||  @.j;         end;   return
```
output   when using the default input:
```before sort: ───►Now is the time for all good men to come to the aid of their country.◄───
after sort: ───►               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy◄───
```

## Ring

```see "working..." + nl
see "Sort the letters of string in alphabitical order:" + nl
str = "forever ring programming language"
see "Input: " + str + nl

for n = 1 to len(str)-1
for m = n+1 to len(str)
if ascii(str[n]) > ascii(str[m])
temp = str[n]
str[n] = str[m]
str[m] = temp
ok
next
next

str = substr(str," ","")
see "Output: " + str + nl
see "done..." + nl```
Output:
```working...
Sort the letters of string in alphabitical order:
Input: forever ring programming language
Output: aaaeeefgggggiilmmnnnooprrrrruv
done...
```

## VBScript

VBScript can't index a string so first we convert the string to an array of chars, then use join to get back a string

```sub bubble(arr)
n = UBound(arr)
Do
nn = -1
For j = 0 to n - 1
If arr(j) > arr(j + 1) Then
temp= arr(j):arr(j)=arr(j+1):arr(j+1)=temp
nn = j
End If
Next
n = nn
Loop Until nn = -1
end sub

s="The quick brown fox jumps over the lazy dog, apparently"
redim a(len(s)-1)
for i=1 to len(s)
a(i-1)=mid(s,i,1)
next
bubble a
s1=join(a,"")
wscript.echo s1
```
Output:
```         ,Taaabcdeeeefghhijkllmnnoooopppqrrrsttuuvwxyyz
```

## Wren

Well, we'll write a function for a bubble sort which we don't have in Wren-sort because it's normally much slower than the other methods. However, it's fast enough here.

```var bubbleSort = Fn.new { |s, trim|  // allow optional removal of whitespace
var chars = s.toList
var n = chars.count
while (true) {
var n2 = 0
for (i in 1...n) {
if (chars[i - 1].codePoints[0] > chars[i].codePoints[0]) {
chars.swap(i, i - 1)
n2 = i
}
}
n = n2
if (n == 0) break
}
s = chars.join()
return trim ? s.trim() : s
}

var strs = [
["forever wren programming language", true],
["Now is the time for all good men to come to the aid of their country.", false]
]
for (str in strs) {
System.print(["Unsorted->" + str[0], "Sorted  ->" + bubbleSort.call(str[0], str[1])].join("\n"))
System.print()
}
```
Output:
```Unsorted->forever wren programming language
Sorted  ->aaaeeeefggggilmmnnnooprrrrruvw

Unsorted->Now is the time for all good men to come to the aid of their country.
Sorted  ->               .Naaccddeeeeeeffghhhiiiillmmmnnooooooooorrrstttttttuwy
```

## XPL0

```string  0;              \use zero-terminated strings

func    StrLen(Str);    \Return number of characters in an ASCIIZ string
char    Str;
int     I;
for I:= 0 to -1>>1 do
if Str(I) = 0 then return I;

func Sort(Str);         \Bubble sort string Str
char Str;
int  J, I, T;
[for J:= StrLen(Str)-1 downto 0 do
for I:= 0 to J-1 do
if Str(I) > Str(I+1) then
[T:= Str(I);  Str(I):= Str(I+1);  Str(I+1):= T];
return Str;
];

[Text(0, Sort("The quick brown fox jumps over the lazy dog."));
CrLf(0);
Text(0, Sort("Pack my box with five dozen liquor jugs."));
CrLf(0);
]```
Output:
```        .Tabcdeeefghhijklmnoooopqrrstuuvwxyz
.Pabcdeefghiiijklmnoooqrstuuvwxyz
```

## Yabasic

```// Rosetta Code problem: http://rosettacode.org/wiki/Sort_the_letters_of_string_in_alphabetical_order
// by Galileo, 04/2022

sub Sorted\$(t\$)
local chars(255), c, i, j, r\$

for i = 1 to len(t\$)
c = asc(mid\$(t\$, i, 1))
chars(c) = chars(c) + 1
next

for i = 32 to 126
c = chars(i)
if c then
for j = 1 to c
r\$ = r\$ + chr\$(i)
next
end if
next

return  r\$
end sub

text\$ = "Sort the letters of string in alphabitical order."

print text\$
print Sorted\$(text\$)```
Output:
```Sort the letters of string in alphabitical order.