Loop over multiple arrays simultaneously: Difference between revisions
add fermat |
Added solution for Action! |
||
Line 224: | Line 224: | ||
(print-lists '("a" "b" "c") '(A B C) '(1 2 3))</lang> |
(print-lists '("a" "b" "c") '(A B C) '(1 2 3))</lang> |
||
=={{header|Action!}}== |
|||
<lang Action!>PROC Main() |
|||
CHAR ARRAY a="abc",b="ABC" |
|||
BYTE ARRAY c=[1 2 3] |
|||
BYTE i |
|||
FOR i=0 TO 2 |
|||
DO |
|||
PrintF("%C%C%B%E",a(i+1),b(i+1),c(i)) |
|||
OD |
|||
RETURN</lang> |
|||
{{out}} |
|||
[https://gitlab.com/amarok8bit/action-rosetta-code/-/raw/master/images/Loop_over_multiple_arrays_simultaneously.png Screenshot from Atari 8-bit computer] |
|||
<pre> |
|||
aA1 |
|||
bB2 |
|||
cC3 |
|||
</pre> |
|||
=={{header|Ada}}== |
=={{header|Ada}}== |
Revision as of 22:29, 25 November 2021
You are encouraged to solve this task according to the task description, using any language you may know.
- Task
Loop over multiple arrays (or lists or tuples or whatever they're called in your language) and display the i th element of each.
Use your language's "for each" loop if it has one, otherwise iterate through the collection in order with some other loop.
For this example, loop over the arrays:
(a,b,c) (A,B,C) (1,2,3)
to produce the output:
aA1 bB2 cC3
If possible, also describe what happens when the arrays are of different lengths.
- Related tasks
- Loop over multiple arrays simultaneously
- Loops/Break
- Loops/Continue
- Loops/Do-while
- Loops/Downward for
- Loops/For
- Loops/For with a specified step
- Loops/Foreach
- Loops/Increment loop index within loop body
- Loops/Infinite
- Loops/N plus one half
- Loops/Nested
- Loops/While
- Loops/with multiple ranges
- Loops/Wrong ranges
11l
<lang 11l>L(x, y, z) zip(‘abc’, ‘ABC’, ‘123’)
print(x‘’y‘’z)</lang>
- Output:
aA1 bB2 cC3
360 Assembly
<lang 360asm>* Loop over multiple arrays simultaneously 09/03/2017 LOOPSIM CSECT
USING LOOPSIM,R12 base register LR R12,R15 LA R6,1 i=1 LA R7,3 counter=3
LOOP LR R1,R6 i
SLA R1,1 *2 LH R2,R-2(R1) r(i) XDECO R2,PG edit r(i) LA R1,S-1(R6) @s(i) MVC PG+3(1),0(R1) output s(i) LA R1,Q-1(R6) @q(i) MVC PG+7(1),0(R1) output q(i) XPRNT PG,80 print s(i),q(i),r(i) LA R6,1(R6) i++ BCT R7,LOOP decrement and loop BR R14 exit
S DC C'a',C'b',C'c' Q DC C'A',C'B',C'C' R DC H'1',H'2',H'3' PG DC CL80' ' buffer
YREGS END LOOPSIM</lang>
- Output:
a A 1 b B 2 c C 3
8080 Assembly
The 8080 has no indexing mechanism at all, so generally one would iterate over arrays by incrementing the pointers in-place rather than do it this way, but it can (just about) be done.
The 8080 has 7 eight-bit registers (A, B, C, D, E, H, L), six of which can form
three 16-bit pairs (BC, DE, HL). Of those, HL is special: only it can be used
for math, and it can be used as a pointer (the 8-bit pseudo-register M refers
to the byte in memory at [HL]
). Furthermore, the contents of DE
and HL can be swapped, so a secondary pointer can be kept in DE and easily accessed.
However, this cannot be done with BC.
Therefore, this code keeps the index in BC, and the list of arrays in DE. Array access is done by loading the array pointers into HL one by one, calculating the address by adding BC to it, then loading the appropriate value.
This code simply assumes that the arrays are all the same size (Alen
),
and if they are not, it will simply read from the wrong addresses.
<lang 8080asm> org 100h lxi b,0 ; Let (B)C be the array index outer: lxi d,As ; Use DE to walk the array-of-arrays inner: xchg ; Swap DE and HL (array-of-array pointer into HL) mov e,m ; Load low byte of array pointer into E inx h mov d,m ; Load high byte of array pointer into D inx h xchg ; Array base in HL, array-of-array pointer in DE mov a,h ; Is HL 0? ora l jz azero ; If so, we are done. dad b ; Otherwise, add index to array base mov a,m ; Get current item (BC'th item of HL) call chout ; Output jmp inner ; Next array azero: mvi a,13 ; Print newline call chout mvi a,10 call chout inr c ; Increment index (we're only using the low byte) mvi a,Alen ; Is it equal to the length? cmp c jnz outer ; If not, get next item from all the arrays. ret ;;; Print character in A, saving all registers. ;;; This code uses CP/M to do it. chout: push psw ; CP/M destroys all registers push b ; Push them all to the stack push d push h mvi c,2 ; 2 = print character syscall mov e,a call 5 pop h ; Restore registers pop d pop b pop psw ret ;;; Arrays A1: db 'a','b','c' A2: db 'A','B','C' A3: db '1','2','3' Alen: equ $-A3 ;;; Zero-terminated array-of-arrays As: dw A1,A2,A3,0</lang>
- Output:
aA1 bB2 cC3
8086 Assembly
The 8086 processor has two index registers si
and di
,
and an address register bx
. (There is also the base pointer bp
,
which is used to point to the stack segment, and is not used here.)
When addressing memory, the 8086 can automatically add up: 1) one of bx
or
bp
, plus 2) one of si
or di
, plus 3)
a direct address.
This code uses si
to keep track of the current index, and loads the
base addresses of the arrays into bx
one by one.
<lang asm> cpu 8086 bits 16 org 100h section .text mov ah,2 ; Tell MS-DOS to print characters xor si,si ; Clear first index register (holds _i_) outer: mov di,As ; Put array-of-arrays in second index register mov cx,Aslen ; Put length in counter register inner: mov bx,[di] ; Load array pointer into BX (address) register mov dl,[bx+si] ; Get SI'th element from array int 21h ; Print character inc di ; Go to next array (pointers are 2 bytes wide) inc di loop inner ; For each array mov dl,13 ; Print newline int 21h mov dl,10 int 21h inc si ; Increment index register cmp si,Alen ; If it is still lower than the array length jb outer ; Print the next items ret section .data ;;; Arrays A1: db 'a','b','c' A2: db 'A','B','C' A3: db '1','2','3' Alen: equ $-A3 ; Length of arrays (elements are bytes) ;;; Array of arrays As: dw A1,A2,A3 Aslen: equ ($-As)/2 ; Length of array of arrays (in words)</lang>
- Output:
aA1 bB2 cC3
ACL2
<lang Lisp>(defun print-lists (xs ys zs)
(if (or (endp xs) (endp ys) (endp zs)) nil (progn$ (cw (first xs)) (cw "~x0~x1~%" (first ys) (first zs)) (print-lists (rest xs) (rest ys) (rest zs)))))
(print-lists '("a" "b" "c") '(A B C) '(1 2 3))</lang>
Action!
<lang Action!>PROC Main()
CHAR ARRAY a="abc",b="ABC" BYTE ARRAY c=[1 2 3] BYTE i
FOR i=0 TO 2 DO PrintF("%C%C%B%E",a(i+1),b(i+1),c(i)) OD
RETURN</lang>
- Output:
Screenshot from Atari 8-bit computer
aA1 bB2 cC3
Ada
<lang Ada>with Ada.Text_IO; use Ada.Text_IO;
procedure Array_Loop_Test is
type Array_Index is range 1..3; A1 : array (Array_Index) of Character := "abc"; A2 : array (Array_Index) of Character := "ABC"; A3 : array (Array_Index) of Integer := (1, 2, 3);
begin
for Index in Array_Index'Range loop Put_Line (A1 (Index) & A2 (Index) & Integer'Image (A3
(Index))(2));
end loop;
end Array_Loop_Test;</lang>
ALGOL 68
<lang algol68>[]UNION(CHAR,INT) x=("a","b","c"), y=("A","B","C"), z=(1,2,3); FOR i TO UPB x DO
printf(($ggd$, x[i], y[i], z[i], $l$))
OD</lang>
- Output:
aA1 bB2 cC3
ALGOL W
<lang algolw>begin
% declare the three arrays % string(1) array a, b ( 1 :: 3 ); integer array c ( 1 :: 3 ); % initialise the arrays - have to do this element by element in Algol W % a(1) := "a"; a(2) := "b"; a(3) := "c"; b(1) := "A"; b(2) := "B"; b(3) := "C"; c(1) := 1; c(2) := 2; c(3) := 3; % loop over the arrays % for i := 1 until 3 do write( i_w := 1, s_w := 0, a(i), b(i), c(i) );
end. </lang>
If the arrays are not the same length, a subscript range error would occur when a non-existant element was accessed.
APL
In APL, one would not use an explicit loop for this. Rather, there is a built-in function to turn a vector of vectors
into a matrix, which is ↑
. The matrix can be transposed (⍉
), and then turned back into a nested
vector (↓
). The elements could be processed linearly afterwards.
If the input vectors are not all the same size, the shorter vectors will be padded with empty values (spaces for character vectors, zeroes for numeric vectors) to match the longest vector.
<lang APL>f ← ↓∘⍉∘↑</lang>
- Output:
f 'abc' 'ABC' '123' ┌───┬───┬───┐ │aA1│bB2│cC3│ └───┴───┴───┘
AppleScript
(Functional ES 5 zipListsWith version)
If we have a generic Applescript map function, we can use it to write a generic zipListsWith, which applies a given function over lists derived from the nth members of an arbitrary list of (equal-length) lists. (Where lists are of uneven length, items beyond the maximum shared length are ignored).
<lang AppleScript>-- ZIP LISTS WITH FUNCTION ---------------------------------------------------
-- zipListsWith :: ([a] -> b) -> a -> b on zipListsWith(f, xss)
set n to length of xss script on |λ|(_, i) script on |λ|(xs) item i of xs end |λ| end script if i ≤ n then apply(f, (map(result, xss))) else {} end if end |λ| end script if n > 0 then map(result, item 1 of xss) else [] end if
end zipListsWith
-- TEST ( zip lists with concat ) -------------------------------------------
on run
intercalate(linefeed, ¬ zipListsWith(concat, ¬ [["a", "b", "c"], ["A", "B", "C"], [1, 2, 3]]))
end run
-- GENERIC FUNCTIONS ---------------------------------------------------------
-- apply (a -> b) -> a -> b on apply(f, a)
mReturn(f)'s |λ|(a)
end apply
-- concat :: a -> [a] | [String] -> String on concat(xs)
if length of xs > 0 and class of (item 1 of xs) is string then set acc to "" else set acc to {} end if repeat with i from 1 to length of xs set acc to acc & item i of xs end repeat acc
end concat
-- intercalate :: Text -> [Text] -> Text on intercalate(strText, lstText)
set {dlm, my text item delimiters} to {my text item delimiters, strText} set strJoined to lstText as text set my text item delimiters to dlm return strJoined
end intercalate
-- map :: (a -> b) -> [a] -> [b] on map(f, xs)
tell mReturn(f) set lng to length of xs set lst to {} repeat with i from 1 to lng set end of lst to |λ|(item i of xs, i, xs) end repeat return lst end tell
end map
-- Lift 2nd class handler function into 1st class script wrapper -- mReturn :: Handler -> Script on mReturn(f)
if class of f is script then f else script property |λ| : f end script end if
end mReturn</lang>
- Output:
aA1 bB2 cC3
But a transpose function might be simpler:
<lang Applescript>-- CONCAT MAPPED OVER A TRANSPOSITION ---------------------------------------- on run
unlines(map(concat, transpose([["a", "b", "c"], ["A", "B", "C"], [1, 2, 3]])))
end run
-- GENERIC FUNCTIONS ---------------------------------------------------------
-- concat :: a -> [a] | [String] -> String on concat(xs)
if length of xs > 0 and class of (item 1 of xs) is string then set acc to "" else set acc to {} end if repeat with i from 1 to length of xs set acc to acc & item i of xs end repeat acc
end concat
-- intercalate :: String -> [String] -> String on intercalate(s, xs)
set {dlm, my text item delimiters} to {my text item delimiters, s} set str to xs as text set my text item delimiters to dlm return str
end intercalate
-- map :: (a -> b) -> [a] -> [b] on map(f, xs)
tell mReturn(f) set lng to length of xs set lst to {} repeat with i from 1 to lng set end of lst to |λ|(item i of xs, i, xs) end repeat return lst end tell
end map
-- Lift 2nd class handler function into 1st class script wrapper -- mReturn :: Handler -> Script on mReturn(f)
if class of f is script then f else script property |λ| : f end script end if
end mReturn
-- transpose :: a -> a on transpose(xss)
script column on |λ|(_, iCol) script row on |λ|(xs) item iCol of xs end |λ| end script map(row, xss) end |λ| end script map(column, item 1 of xss)
end transpose
-- unlines :: [String] -> String on unlines(xs)
intercalate(linefeed, xs)
end unlines</lang>
- Output:
aA1 bB2 cC3
Arturo
<lang rebol>parts: ["abc" "ABC" [1 2 3]]
loop 0..2 'x ->
print ~"|parts\0\[x]||parts\1\[x]||parts\2\[x]|"</lang>
- Output:
aA1 bB2 cC3
AutoHotkey
Pseudo-arrays
StringSplit creates a pseudo-array <lang autohotkey>List1 = a,b,c List2 = A,B,C List3 = 1,2,3 MsgBox, % LoopMultiArrays()
List1 = a,b,c,d,e List2 = A,B,C,D List3 = 1,2,3 MsgBox, % LoopMultiArrays()
- ---------------------------------------------------------------------------
LoopMultiArrays()
{ ; print the ith element of each
- ---------------------------------------------------------------------------
local Result StringSplit, List1_, List1, `, StringSplit, List2_, List2, `, StringSplit, List3_, List3, `, Loop, % List1_0 Result .= List1_%A_Index% List2_%A_Index% List3_%A_Index% "`n" Return, Result
}</lang> An array that is too short on creation will return empty strings when trying to retrieve further elements. The 2nd Message box shows:
aA1 bB2 cC3 dD e
Real arrays
In AutoHotkey_L, we can use true arrays (Objects) and the For loop. <lang AHK>List1 := ["a", "b", "c"] List2 := ["A", "B", "C"] List3 := [ 1 , 2 , 3 ] MsgBox, % LoopMultiArrays()
List1 := ["a", "b", "c", "d", "e"] List2 := ["A", "B", "C", "D"] List3 := [1,2,3] MsgBox, % LoopMultiArrays()
LoopMultiArrays() {
local Result For key, value in List1 Result .= value . List2[key] . List3[key] "`n" Return, Result
}</lang> The output from this script is identical to the first one.
AWK
<lang awk>BEGIN {
split("a,b,c", a, ","); split("A,B,C", b, ","); split("1,2,3", c, ",");
for(i = 1; i <= length(a); i++) { print a[i] b[i] c[i]; }
}</lang>
Axe
Note that in this example, we use a few bytes from each of L₁, L₂, and L₃ for simplicity. In practice, one would want to arrange the arrays to all fit within L₁ to avoid volatility issues with L₂ and L₃. <lang axe>'a'→{L₁} 'b'→{L₁+1} 'c'→{L₁+2} 'A'→{L₂} 'B'→{L₂+1} 'C'→{L₂+2} 1→{L₃} 2→{L₃+1} 3→{L₃+2} For(I,0,2) Disp {L₁+I}►Char,{L₂+I}►Char,{L₃+I}►Dec,i End</lang>
Babel
There are two ways to do this in Babel. First, you could transpose the lists:
<lang babel>main: { (('a' 'b' 'c')('A' 'B' 'C')('1' '2' '3')) simul_array }
simul_array!:
{ trans { { << } each "\n" << } each }</lang>
The 'trans' operator substitutes nil in the portions of each transposed column wherever a row list was shorter than the longest row list. The '<<' operator prints nothing if the top-of-stack is nil.
A more literal solution to the problem as presented would be to iterate across each list using a user-defined cdrall operator:
<lang babel>main: { (('a' 'b' 'c')('A' 'B' 'C')('1' '2' '3')) simul_array }
simul_array!:
{{ dup { car << } each cdrall } { allnil? not } while }
cdrall!: { { { cdr } each -1 take } nest }
-- only returns true if all elements of a list are nil allnil?!:
{ 1 <-> { car nil? { zap 0 last } { nil } if} each }</lang>
This solution is formally identical to the first and will handle lists of varying lengths by printing inserting nil and printing nothing for the tail ends
of the
short lists.
BaCon
<lang freebasic> DECLARE a1$[] = {"a", "b", "c"} TYPE STRING DECLARE a2$[] = {"A", "B", "C"} TYPE STRING DECLARE a3[] = {1, 2, 3} TYPE int
WHILE (a3[i] <= 3)
PRINT a1$[i], a2$[i], a3[i]
INCR i
WEND
</lang>
BASIC256
<lang BASIC256>dim arr1$(3) : arr1$ = {"a", "b", "c"} dim arr2$(3) : arr2$ = {"A", "B", "C"} dim arr3(3) : arr3 = {1, 2, 3}
for i = 0 to 2 print arr1$[i]; arr2$[i]; arr3[i] next i end</lang>
- Output:
aA1 bB2 cC3
BBC BASIC
<lang bbcbasic> DIM array1$(2), array2$(2), array3%(2)
array1$() = "a", "b", "c" array2$() = "A", "B", "C" array3%() = 1, 2, 3 FOR index% = 0 TO 2 PRINT array1$(index%) ; array2$(index%) ; array3%(index%) NEXT</lang>
Beads
This solution accounts for arrays of varying lengths, and if they are interspersed with undefined characters by replacing them with spaces. <lang Beads>beads 1 program 'Loop over multiple arrays simultaneously' calc main_init const x = ['a', 'b', 'c'] y = ['A', 'B', 'C'] z = [1, 2, 3] const largest = max(tree_hi(x), tree_hi(y), tree_hi(z)) loop reps:largest count:i //where u_cc defines what to use for undefined characters log to_str(x[i], u_cc:' ') & to_str(y[i], u_cc:' ') & to_str(z[i], u_cc:' ')</lang>
- Output:
aA1 bB2 cC3
Befunge
There's no concept of an array data type in Befunge, but you'd typically store your arrays as sequences of data in the Befunge code space. You'd then loop over the range of indices required to access those arrays, and use the loop variable as an offset into each data area. For arrays of differing lengths, you'd need to manually check for an out-of-range index and deal with it appropriately.
<lang befunge>0 >:2g,:3g,:4gv @_^#`2:+1,+55,< abc ABC 123</lang>
C
Given several arrays, especially if they are heterogeneous, the most ordinary way to loop over all of them is to simply use an index variable. Determining when to stop is generally done in some application-specific way.
<lang c>#include <stdio.h>
char a1[] = {'a','b','c'}; char a2[] = {'A','B','C'}; int a3[] = {1,2,3};
int main(void) {
for (int i = 0; i < 3; i++) { printf("%c%c%i\n", a1[i], a2[i], a3[i]); }
}</lang>
(Note: Some compilers may require a flag to accept this modern C code,
such as gcc -std=c99
.)
On the other hand, it is possible to write a more generic higher-order iteration scheme, as demonstrated in this example.
There, a type for arrays with runtime-specified lengths and polymorphic printing is defined, and the iteration continues up to the length of
the shortest array.
C#
<lang csharp>class Program {
static void Main(string[] args) { char[] a = { 'a', 'b', 'c' }; char[] b = { 'A', 'B', 'C' }; int[] c = { 1, 2, 3 }; int min = Math.Min(a.Length, b.Length); min = Math.Min(min, c.Length); for (int i = 0; i < min; i++) Console.WriteLine("{0}{1}{2}", a[i], b[i], c[i]); }
}</lang>
Using Enumerable.Zip (stops when either source runs out of elements):
<lang csharp> int[] numbers = { 1, 2, 3, 4 }; string[] words = { "one", "two", "three" }; Console.WriteLine(numbers.Zip(words, (first, second) => first + " " +
second));
</lang>
Like how a perl programmer would write it (still using Zip):
<lang csharp> Console.WriteLine((new[] { 1, 2, 3, 4 }).Zip(new[] { "a", "b", "c" }, (f, s) => f + " " + s)); </lang>
Custom implementation for arrays of different lengths that pads with spaces after the end of the shorter arrays:
<lang csharp>
public static void Multiloop(char[] A, char[] B, int[] C) { var max = Math.Max(Math.Max(A.Length, B.Length), C.Length); for (int i = 0; i < max; i++) Console.WriteLine($"{(i < A.Length ? A[i] : ' ')}, {(i < B.Length ? B[i] : ' ')}, {(i < C.Length ? C[i] : ' ')}"); }
</lang> usage: <lang csharp>Multiloop(new char[] { 'a', 'b', 'c', 'd' }, new char[] { 'A', 'B', 'C' }, new int[] { 1, 2, 3, 4, 5 });</lang>
C++
With std::vector
s:
<lang cpp>#include <iostream>
- include <vector>
int main(int argc, char* argv[]) {
std::vector<char> ls(3); ls[0] = 'a'; ls[1] = 'b'; ls[2] = 'c'; std::vector<char> us(3); us[0] = 'A'; us[1] = 'B'; us[2] = 'C'; std::vector<int> ns(3); ns[0] = 1; ns[1] = 2; ns[2] = 3;
std::vector<char>::const_iterator lIt = ls.begin(); std::vector<char>::const_iterator uIt = us.begin(); std::vector<int>::const_iterator nIt = ns.begin(); for(; lIt != ls.end() && uIt != us.end() && nIt !=
ns.end();
++lIt, ++uIt, ++nIt) { std::cout << *lIt << *uIt << *nIt << "\n"; }
}</lang>
Using static arrays: <lang cpp>#include <iostream>
int main(int argc, char* argv[]) {
char ls[] = {'a', 'b', 'c'}; char us[] = {'A', 'B', 'C'}; int ns[] = {1, 2, 3};
for(size_t li = 0, ui = 0, ni = 0; li < sizeof(ls) && ui < sizeof(us) && ni
< sizeof(ns) / sizeof(int);
++li, ++ui, ++ni) { std::cout << ls[li] << us[ui] << ns[ni] << "\n"; }
}</lang>
C++11
With std::vector
s:
<lang cpp>#include <iostream>
- include <vector>
int main(int argc, char* argv[]) {
auto lowers = std::vector<char>({'a', 'b', 'c'}); auto uppers = std::vector<char>({'A', 'B', 'C'}); auto nums = std::vector<int>({1, 2, 3});
auto ilow = lowers.cbegin(); auto iup = uppers.cbegin(); auto inum = nums.cbegin();
for(; ilow != lowers.end() and iup != uppers.end() and inum != nums.end() ; ++ilow, ++iup, ++inum) { std::cout << *ilow << *iup << *inum << "\n"; }
}</lang>
Using static arrays: <lang cpp>#include <iostream>
- include <iterator>
int main(int argc, char* argv[]) {
char lowers[] = {'a', 'b', 'c'}; char uppers[] = {'A', 'B', 'C'}; int nums[] = {1, 2, 3};
auto ilow = std::begin(lowers); auto iup = std::begin(uppers); auto inum = std::begin(nums);
for(; ilow != std::end(lowers) and iup != std::end(uppers) and inum != std::end(nums) ; ++ilow, ++iup, ++inum ) { std::cout << *ilow << *iup << *inum << "\n"; }
}</lang>
With std::array
s:
<lang cpp>#include <iostream>
- include <array>
int main(int argc, char* argv[]) {
auto lowers = std::array<char, 3>({'a', 'b', 'c'}); auto uppers = std::array<char, 3>({'A', 'B', 'C'}); auto nums = std::array<int, 3>({1, 2, 3});
auto ilow = lowers.cbegin(); auto iup = uppers.cbegin(); auto inum = nums.cbegin();
for(; ilow != lowers.end() and iup != uppers.end() and inum != nums.end() ; ++ilow, ++iup, ++inum ) { std::cout << *ilow << *iup << *inum << "\n"; }
}</lang>
With std::array
s by indexes:
<lang cpp>#include <iostream>
- include <array>
- include <algorithm>
int main(int argc, char* argv[]) {
auto lowers = std::array<char, 3>({'a', 'b', 'c'}); auto uppers = std::array<char, 3>({'A', 'B', 'C'}); auto nums = std::array<int, 3>({1, 2, 3});
auto const minsize = std::min( lowers.size(), std::min( uppers.size(), nums.size() ) );
for(size_t i = 0; i < minsize; ++i) { std::cout << lowers[i] << uppers[i] << nums[i] << "\n"; }
}</lang>
Chapel
<lang chapel>var a1 = [ "a", "b", "c" ]; var a2 = [ "A", "B", "C" ]; var a3 = [ 1, 2, 3 ];
for (x,y,z) in zip(a1, a2, a3) do
writeln(x,y,z);</lang>
Clojure
<lang clojure>(doseq [s (map #(str %1 %2 %3) "abc" "ABC" "123")]
(println s))</lang>
The sequence stops when the shortest list is exhausted.
<lang clojure>
(apply map str ["abc" "ABC" "123"])
("aA1" "bB2" "cC3")
</lang>
COBOL
<lang cobol> IDENTIFICATION DIVISION.
PROGRAM-ID. Loop-Over-Multiple-Tables.
DATA DIVISION. WORKING-STORAGE SECTION. 01 A VALUE "abc". 03 A-Vals PIC X OCCURS 3 TIMES.
01 B VALUE "ABC". 03 B-Vals PIC X OCCURS 3 TIMES.
01 C VALUE "123". 03 C-Vals PIC 9 OCCURS 3 TIMES.
01 I PIC 9.
PROCEDURE DIVISION. PERFORM VARYING I FROM 1 BY 1 UNTIL 3 < I DISPLAY A-Vals (I) B-Vals (I) C-Vals (I) END-PERFORM
GOBACK .</lang>
Common Lisp
Using functional application
<lang lisp>(mapc (lambda (&rest args)
(format t "~{~A~}~%" args)) '(|a| |b| |c|) '(a b c) '(1 2 3))</lang>
If lists are different lengths, it stops after the shortest one.
Using LOOP
<lang lisp>
(loop for x in '("a" "b" "c") for y in '(a b c) for z in '(1 2 3) do (format t "~a~a~a~%" x y z))
</lang>
Using DO
<lang lisp> (do ((x '("a" "b" "c") (rest x)) ;
(y '("A" "B" "C" "D") (rest y)) ; (z '(1 2 3 4 6) (rest z))) ; Initialize lists and set to rest on every loop ((or (null x) (null y) (null z))) ; Break condition (format t "~a~a~a~%" (first x) (first y) (first z))) ; On every loop print first elements
</lang>
- Output:
aA1 bB2 cC3
D
<lang d>import std.stdio, std.range;
void main () {
foreach (a, b, c; zip("abc", "ABC", [1, 2, 3])) writeln(a, b, c);
}</lang>
- Output:
aA1 bB2 cC3
zip() allows to specify the stopping policy. On default it stops when the shortest range is exhausted (same as StoppingPolicy.shortest): <lang d>import std.stdio, std.range;
void main () {
auto a1 = [1, 2]; auto a2 = [1, 2, 3]; alias StoppingPolicy sp;
// Stops when the shortest range is exhausted foreach (p; zip(sp.shortest, a1, a2)) writeln(p.tupleof); writeln();
// Stops when the longest range is exhausted foreach (p; zip(sp.longest, a1, a2)) writeln(p.tupleof); writeln();
// Requires that all ranges are equal foreach (p; zip(sp.requireSameLength, a1, a2)) writeln(p.tupleof);
}</lang>
- Output:
11 22 11 22 03 11 22
Followed by an exception with message "Inequal-length ranges passed to Zip".
There is also std.range.lockstep: <lang d>import std.stdio, std.range;
void main() {
auto arr1 = [1, 2, 3, 4, 5]; auto arr2 = [6, 7, 8, 9, 10];
foreach (ref a, ref b; lockstep(arr1, arr2)) a += b;
assert(arr1 == [7, 9, 11, 13, 15]);
// Lockstep also supports iteration with an index variable foreach (index, a, b; lockstep(arr1, arr2)) writefln("Index %s: a = %s, b = %s", index, a, b);
}</lang> Lower level code that stops at the shortest length: <lang d>import std.stdio, std.algorithm;
void main () {
auto s1 = "abc"; auto s2 = "ABC"; auto a1 = [1, 2];
foreach (i; 0 .. min(s1.length, s2.length, a1.length)) writeln(s1[i], s2[i], a1[i]);
}</lang>
- Output:
aA1 bB2
Delphi
<lang Delphi>program LoopOverArrays;
{$APPTYPE CONSOLE}
uses SysUtils;
const
ARRAY1: array [1..3] of string = ('a', 'b', 'c'); ARRAY2: array [1..3] of string = ('A', 'B', 'C'); ARRAY3: array [1..3] of Integer = (1, 2, 3);
var
i: Integer;
begin
for i := 1 to 3 do Writeln(Format('%s%s%d', [ARRAY1[i], ARRAY2[i], ARRAY3[i]]));
Readln;
end.</lang>
DWScript
If the arrays don't have the same bounds, an index out of bound exception will be triggered when attempting to access a non-existing element.
<lang delphi>const a1 = ['a', 'b', 'c']; const a2 = ['A', 'B', 'C']; const a3 = [1, 2, 3];
var i : Integer; for i := 0 to 2 do
PrintLn(Format('%s%s%d', [a1[i], a2[i], a3[i]]));</lang>
E
E lacks a nice way to do this; this is [http://wiki.erights.org/wiki/Parallel_iteration to be fixed, once we figure out what to do]. However, iteration over an List produces its indexes as keys, so a not entirely awful idiom exists:
<lang e>def a1 := ["a","b","c"] def a2 := ["A","B","C"] def a3 := ["1","2","3"]
for i => v1 in a1 {
println(v1, a2[i], a3[i])
}</lang>
This will obviously fail if a2 or a3 are shorter than a1, and omit items
if a2 or a3 are longer.
Given a parallel iteration utility, we might write this:
<lang e>for [v1, v2, v3] in zip(a1, a2, a3) {
println(v1, v2, v3)
}</lang>
zip
cannot yet be defined for all collections
(other than by iterating over each one and storing the results in a List
first); but we can define it for numeric-indexed collections such as
Lists, as below. Both a definition for any number of collections and two
collections is given; the latter in order to demonstrate the principle
without the clutter resulting from handling a variable number of collections.
<lang e>def zip {
to run(l1, l2) { def zipped { to iterate(f) { for i in int >= 0 { f(i, [l1.fetch(i, fn { return }), l2.fetch(i, fn { return })]) } } } return zipped }
match [`run`, lists] { def zipped { to iterate(f) { for i in int >= 0 { var tuple := [] for l in lists { tuple with= l.fetch(i, fn { return }) } f(i, tuple) } } } zipped }
}</lang>
(This will stop when the end of the shortest collection is reached.)
EchoLisp
<lang scheme>
- looping over different sequences
- infinite stream, string, list and vector
- loop stops as soon a one sequence ends.
- the (iota 6) = ( 0 1 2 3 4 5) sequence will stop first.
(for ((i (in-naturals 1000)) (j "ABCDEFGHIJK") (k (iota 6)) (m #(o p q r s t u v w)))
(writeln i j k m))
1000 "A" 0 o 1001 "B" 1 p 1002 "C" 2 q 1003 "D" 3 r 1004 "E" 4 s 1005 "F" 5 t </lang>
Efene
<lang efene>@public run = fn () {
lists.foreach(fn ((A, B, C)) { io.format("~s~n", A, B, C) },
lists.zip3("abc", "ABC", "123")) } </lang>
If the lists are not all the same length, an error is thrown.
Eiffel
<lang eiffel> example (a_array: READABLE_INDEXABLE [BOUNDED [ANY]]): STRING -- Assemble output for a 2-dim array in `a_array' require non_zero: ∀ nzitem:a_array ¦ nzitem.count > 0 local min_count: INTEGER do ⟳ v_item:a_array ¦ min_count := if min_count = 0 then v_item.count else v_item.count.min (min_count) end ⟲ create Result.make_empty ⟳ j:1 |..| min_count ¦ ⟳ i:a_array ¦ if attached {READABLE_INDEXABLE [ANY]} i as al_i then Result.append_string_general (al_i [j].out) end ⟲ Result.append_string_general ("%N") ⟲ end
input_data: ARRAY [BOUNDED [ANY]] -- Sample `input_data' for `example' (above). do Result := << "abcde", "ABC", <<1, 2, 3, 4>> >> end </lang>
- Output:
aA1
bB2
cC3
Explanation
The `require' Design-by-Contract assertion is a statement of software correctness. It states that all items in `a_array' must have a count > 0 (no empty of type BOUNDED).
If you examine the `input_data', you will see that collection 1 is not just "abc", but is "abcde" (5 character element items in a BOUNDED string). The same is true for the last numeric ARRAY, which has 4 integers. This is done to demonstrate that the `example' code is robust enough to take variants in the inputs in terms of item counts.
The first ⟳ ¦ ⟲ (symbolic across) loop seeks out the count of the smallest (min) collection. In this case, the middle item (#2) has only 3 elements, so this routine will only process the first 3 elements of each collection in the containing array.
Next, we create the output STRING in the `Result'.
Finally, the last ⟳ ¦ ⟲ (symbolic across) loop has a nested loop. The outer loop counts the elements (1-3) and the inner loop goes of the contained collections, adding the j-th element of the i-th collection. This repeats until all of `j' is exhausted for all of `i'.
Ela
<lang ela>open monad io list imperative
xs = zipWith3 (\x y z -> show x ++ show y ++ show z) ['a','b','c']
['A','B','C'] [1,2,3]
print x = do putStrLn x
print_and_calc xs = do
xss <- return xs return $ each print xss
print_and_calc xs ::: IO</lang>
The code above can be written shorter. First there is no need in lists as soon as strings in Ela can be treated as lists. Also instead of explicit labmda one can use partial application and a standard composition operator:
<lang ela>xs = zipWith3 (\x -> (x++) >> (++)) "abc" "ABC"
"123"</lang>
Elena
ELENA 5.0 : <lang elena>import system'routines; import extensions;
public program() {
var a1 := new string[]{"a","b","c"}; var a2 := new string[]{"A","B","C"}; var a3 := new int[]{1,2,3}; for(int i := 0, i < a1.Length, i += 1) { console.printLine(a1[i], a2[i], a3[i]) }; console.readChar()
}</lang>
Using zipBy extension: <lang elena>import system'routines. import extensions.
public program {
var a1 := new string[]{"a","b","c"}; var a2 := new string[]{"A","B","C"}; var a3 := new int[]{1,2,3}; var zipped := a1.zipBy(a2,(first,second => first + second.toString() )) .zipBy(a3, (first,second => first + second.toString() )); zipped.forEach:(e) { console.writeLine:e }; console.readChar();
}</lang>
- Output:
aA1 bB2 cC3
Elixir
string list: <lang elixir>l1 = ["a", "b", "c"] l2 = ["A", "B", "C"] l3 = ["1", "2", "3"] IO.inspect List.zip([l1,l2,l3]) |> Enum.map(fn x-> Tuple.to_list(x) |> Enum.join end)
- => ["aA1", "bB2", "cC3"]</lang>
char_list: <lang elixir>l1 = 'abc' l2 = 'ABC' l3 = '123' IO.inspect List.zip([l1,l2,l3]) |> Enum.map(fn x-> Tuple.to_list(x) end)
- => ['aA1', 'bB2', 'cC3']</lang>
When the length of the list is different: <lang elixir>iex(1)> List.zip(['abc','ABCD','12345']) |> Enum.map(&Tuple.to_list(&1)) ['aA1', 'bB2', 'cC3'] iex(2)> List.zip(['abcde','ABC','12']) |> Enum.map(&Tuple.to_list(&1)) ['aA1', 'bB2']</lang> The zipping finishes as soon as any enumerable completes.
Erlang
Shortest option: <lang erlang>lists:zipwith3(fun(A,B,C)-> io:format("~s~n",A,B,C) end, "abc", "ABC", "123").</lang> However, as every expression in Erlang has to return something, printing text returns 'ok'. A list with as many 'ok's as there are lines printed will thus be created. The technically cleanest way to do things would be with lists:foreach/2, which also guarantees evaluation order: <lang erlang>lists:foreach(fun({A,B,C}) -> io:format("~s~n",A,B,C) end,
lists:zip3("abc", "ABC", "123")).</lang>
If the lists are not all the same length, an error is thrown.
Euphoria
There are many ways to do this. All of them rely on what strings really
are.
If they are all "strings", it's quite easy: <lang Euphoria> sequence a, b, c
a = "abc" b = "ABC" c = "123"
for i = 1 to length(a) do
puts(1, a[i] & b[i] & c[i] & "\n")
end for </lang>
If not, and the other sequence is known to contain only integers:
<lang Euphoria> sequence a, b, c
a = "abc" b = "ABC" c = {1, 2, 3}
for i = 1 to length(a) do
printf(1, "%s%s%g\n", {a[i], b[i], c[i]})
end for </lang>
A general solution for any arbitrary strings of characters or numbers can get a bit complex. This is because of how sequences are stored and printed out. One possible answer is as follows, if you know that only alphanumeric characters are used: <lang Euphoria> for i = 1 to length(a) do
if (a[i] >= '0' and a[i] <= '9') then
a[i] -= '0'
end if if (b[i] >= '0' and b[i] <= '9') then
b[i] -= '0'
end if if (c[i] >= '0' and c[i] <= '9') then
c[i] -= '0'
end if printf(1, "%s%s%s\n", {a[i], b[i], c[i]})
end for </lang> Just as in Java, using single quotes around a character gives you its "char value". In Euphoria, though, it is simply that character's code in ASCII.
With all three of the above solutions, if any of the strings are smaller
than the first, it will return an error.
F#
<lang fsharp>for c1,c2,n in Seq.zip3 ['a';'b';'c'] ['A';'B';'C'] [1;2;3] do
printfn "%c%c%d" c1 c2 n</lang>
When one sequence is exhausted, any remaining elements in the other sequences are ignored.
Factor
<lang factor>"abc" "ABC" "123" [ [ write1 ] tri@ nl ] 3each</lang>
Fantom
This will stop when it reaches the end of the shortest list. <lang fantom> class LoopMultiple {
public static Void main () { List arr1 := ["a", "b", "c"] List arr2 := ["A", "B", "C"] List arr3 := [1, 2, 3] [arr1.size, arr2.size, arr3.size].min.times |Int i| { echo ("${arr1[i]}${arr2[i]}${arr3[i]}") } }
} </lang>
Fermat
<lang fermat>[a] := [('a','b','c')]; [b] := [('A','B','C')]; [c] := [(1,2,3)]; for i=1,3 do !!(a[i]:char,b[i]:char,c[i]:1) od;
- {note the
- char and :1 suffixes. The former}
- {causes the element to be printed as a char}
- {instead of a numerical ASCII code, and the}
- {
- 1 causes the integer to take up exactly one}
- {space, ie. no leading or trailing spaces.}</lang>
- Output:
aA1
bB2
cC3
Forth
<lang forth>create a char a , char b , char c , create b char A , char B , char C , create c char 1 , char 2 , char 3 ,
- main
3 0 do cr a i cells + @ emit b i cells + @ emit c i cells + @ emit loop cr a b c 3 0 do cr 3 0 do rot dup @ emit cell+ loop loop drop drop drop
- </lang>
Fortran
<lang fortran>program main
implicit none
integer,parameter :: n_vals = 3 character(len=*),dimension(n_vals),parameter :: ls = ['a','b','c'] character(len=*),dimension(n_vals),parameter :: us = ['A','B','C'] integer,dimension(n_vals),parameter :: ns = [1,2,3]
integer :: i !counter
do i=1,n_vals write(*,'(A1,A1,I1)') ls(i),us(i),ns(i) end do
end program main </lang>
If the arrays are of different length (say, array ns has no third element), then when its turn comes the next unit of storage along from the second element will be accessed, its content interpreted as an integer, and its decimal value printed... If however, array bound checking is activated (or there is a memory access protection scheme that would detect this), a feature unavailable via many compilers and not the default on the rest, then an error will be detected and the run will be terminated, possibly with a somewhat helpful message.
If instead of reading the action had been to store a value into the array, then in the absence of bound checking, arbitrary damage will be done (to code or data) that will possibly result in something going wrong. And if you're lucky, it will happen swiftly.
FreeBASIC
<lang freebasic>' FB 1.05.0 Win64
Function min(x As Integer, y As Integer) As Integer
Return IIf(x < y, x, y)
End Function
Dim arr1(1 To 3) As String = {"a", "b", "c"} Dim arr2(1 To 3) As String = {"A", "B", "C"} Dim arr3(1 To 3) As Integer = {1, 2, 3}
For i As Integer = 1 To 3
Print arr1(i) & arr2(i) & arr3(i)
Next
' For arrays of different lengths we would need to iterate up to the mimimm length of all 3 in order ' to get a contribution from each one. For example:
Dim arr4(1 To 4) As String = {"A", "B", "C", "D"} Dim arr5(1 To 2) As Integer = {1, 2}
Dim ub As Integer = min(UBound(arr1), min(UBound(arr4), UBound(arr5))) For i As Integer = 1 To ub
Print arr1(i) & arr2(i) & arr3(i)
Next
Print Sleep</lang>
- Output:
aA1 bB2 cC3 aA1 bB2
Frink
<lang frink>a1 = ["a", "b", "c"] a2 = ["A", "B", "C"] a3 = ["1", "2", "3"] m = [a1, a2, a3] for row = m.transpose[]
println[join["",row]]</lang>
FunL
<lang funl>import lists.zip3
for x <- zip3( ['a', 'b', 'c'], ['A', 'B', 'C'], [1, 2, 3] )
println( x.mkString() )</lang>
- Output:
aA1 bB2 cC3
Gambas
Click this link to run this code <lang gambas>Public Sub Main() Dim a1 As String[] = ["a", "b", "c"] Dim a2 As String[] = ["A", "B", "C"] Dim a3 As String[] = ["1", "2", "3"] Dim siC As Short
For siC = 0 To a1.Max
Print a1[siC] & a2[siC] & a3[siC]
Next
End</lang>
Output:
aA1 bB2 cC3
GAP
<lang gap>
- The Loop function will apply some function to every tuple built by
taking
- the i-th element of each list. If one of them is exhausted before the
others,
- the loop continues at its begining. Only the longests lists will be
precessed only once. Loop := function(a, f)
local i, j, m, n, v; n := Length(a); v := List(a, Length); m := Maximum(v); for j in [1 .. m] do f(List([1 .. n], i -> a[i][1 + RemInt(j - 1, v[i])])); od;
end;
- Here we simply print each "row"
f := function(u)
Perform(u, Print); Print("\n");
end;
Loop([["a", "b", "c"], ["A", "B", "C"], [1, 2, 3]], f);
aA1 bB2 cC3
Loop([["a", "b"], ["A", "B", "C", "D", "E"], [1, 2, 3]], f);
aA1 bB2 aC3 bD1 aE2</lang>
Go
Go's "range clause" of a for statement only looks at a single iterable value (array, slice, etc). To access the three in parallel, they have to be explicitly indexed.
If a2
or a3
were
shorter, the program would panic with "runtime error: index out of
range".
If a2
or a3
were
longer, extra elements would be ignored.
Go's philosophy is that you should explicitly check for whatever
conditions are meaningful in your application and explicitly handle
whatever errors are plausible.
<lang go>package main
import "fmt"
var a1 = []string{"a", "b", "c"} var a2 = []byte{'A', 'B', 'C'} var a3 = []int{1, 2, 3}
func main() { for i := range a1 { fmt.Printf("%v%c%v\n", a1[i], a2[i], a3[i]) } }</lang>
Golfscript
<lang golfscript>["a" "b" "c"]:a; ["A" "B" "C"]:b; ["1" "2" "3"]:c; [a b c]zip{puts}/</lang>
If there are arrays of different size, the shorter are treated as "null-padded" array.
Groovy
Solution: <lang groovy>def synchedConcat = { a1, a2, a3 ->
assert a1 && a2 && a3 assert a1.size() == a2.size() assert a2.size() == a3.size() [a1, a2, a3].transpose().collect { "${it[0]}${it[1]}${it[2]}" }
}</lang>
Test: <lang groovy>def x = ['a', 'b', 'c'] def y = ['A', 'B', 'C'] def z = [1, 2, 3]
synchedConcat(x, y, z).each { println it }</lang>
- Output:
aA1 bB2 cC3
Harbour
Using FOR EACH ... NEXT statement <lang visualfoxpro> PROCEDURE Main() LOCAL a1 := { "a", "b", "c" }, ; a2 := { "A", "B", "C", "D" }, ; // the last element "D" of this array will be ignored a3 := { 1, 2, 3 } LOCAL e1, e2, e3
FOR EACH e1, e2, e3 IN a1, a2, a3 Qout( e1 + e2 + hb_ntos( e3 ) ) NEXT RETURN </lang> Output:
aA1 bB2 cC3
If the arrays are not of equal length, the iteration stops after the last item of the smaller array has been processed; any extra items of lengthier arrays are ignored (or in other words, the iteration counter never exceeds the length of the smaller array, thus preventing an 'out of subscript range' error).
Haskell
Using list comprehension
<lang haskell>{-# LANGUAGE ParallelListComp #-} main = sequence [ putStrLn [x, y, z] | x <- "abd" | y <- "ABC" | z <- "123"]</lang>
Using Transpose
In this special case of transposing strings.
<lang haskell>import Data.List main = mapM putStrLn $ transpose ["abd", "ABC", "123"]</lang>
Using ZipWith*
<lang haskell>import Data.List main = mapM putStrLn $ zipWith3 (\a b c -> [a,b,c]) "abc" "ABC" "123"</lang>
Using applicative ZipLists
ZipLists generalize zipWith to any number of parameters <lang haskell>import Control.Applicative main = sequence $ getZipList $ (\x y z -> putStrLn [x, y, z]) <$> ZipList "abd" <*> ZipList "ABC" <*> ZipList "123"</lang>
Haxe
<lang haxe>using Lambda; using Std;
class Main {
static function main() { var a = ['a', 'b', 'c']; var b = ['A', 'B', 'C']; var c = [1, 2, 3];
//Find smallest array var len = [a, b, c] .map(function(a) return a.length) .fold(Math.min, 0x0FFFFFFF) .int();
for (i in 0...len) Sys.println(a[i] + b[i] + c[i].string()); } }</lang>
HicEst
<lang HicEst>CHARACTER :: A = "abc" REAL :: C(3)
C = $ ! 1, 2, 3
DO i = 1, 3
WRITE() A(i), "ABC"(i), C(i)
ENDDO</lang>
Icon and Unicon
The first solution uses co-expressions to produce parallel evaluation. <lang Icon>procedure main() a := create !["a","b","c"] b := create !["A","B","C"] c := create !["1","2","3"] while write(@a,@b,@c) end</lang>
The second solution is more like other procedural languages and also handles unequal list lengths. <lang Icon>link numbers # for max
procedure main()
a := ["a","b","c"] b := ["A","B","C","D"] c := [1,2,3]
every i := 1 to max(*a,*b,*c) do
write(a[i]|"","\t",b[i]|"","\t",c[i]|"")
end</lang>
[http://www.cs.arizona.edu/icon/library/procs/numbers.htm Uses max from numbers]
J
For arrays of different types: <lang J> ,.&:(":"0@>)/ 'abc' ; 'ABC' ; 1 2 3 aA1 bB2 cC3</lang> This approach works by representing the digits as characters.
Where arrays are all the same type (all numeric or all string): <lang J> ,.&:>/ 'abc' ; 'ABC' ; '123' aA1 bB2 cC3</lang>
Both of these implementations reject arrays with conflicting lengths.
Other options include:
<lang J> |: 'abc', 'ABC' ,:;":&> 1 2 3 aA1 bB2 cC3</lang> <lang J> |: 'abc', 'ABC',: '123' aA1 bB2 cC3</lang> These implementations pad short arrays with spaces.
Or:
<lang J> |:>]&.>L:_1 'abc';'ABC';<1 2 3 ┌─┬─┬─┐ │a│A│1│ ├─┼─┼─┤ │b│B│2│ ├─┼─┼─┤ │c│C│3│ └─┴─┴─┘</lang>
This implementation puts each item from each of the original lists into a box and forms an array of boxes. (A "box" is a immutable pointer to immutable data -- in other words value semantics instead of reference semantics -- and "putting an item into a box" is obtaining one of these pointers for that item.) This implementation extends any short array by providing empty boxes to represent the missing elements. (An "empty box" is what a programmer in another language might call "a pointer to a zero length array".)
Java
<lang java>String[][] list1 = {{"a","b","c"}, {"A", "B", "C"}, {"1", "2", "3"}};
for (int i = 0; i < list1.length; i++) { for (String[] lista : list1) { System.out.print(lista[i]); } System.out.println(); }</lang>
JavaScript
Imperative
This loops over the indices of the first array, and uses that to index into the others. <lang javascript>var a = ["a","b","c"],
b = ["A","B","C"], c = [1,2,3], output = "", i;
for (i = 0; i < a.length; i += 1) {
output += a[i] + b[i] + c[i] + "\n";
}</lang> If the b or c arrays are too "short", you will see the string "undefined" appear in the output.
Alternatively, we can nest a couple of calls to .forEach(): one for the array of three arrays, and one for each of the three index positions:
<lang JavaScript>var lstOut = [, , ];
[["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]].forEach(
function (a) { [0, 1, 2].forEach( function (i) { // side-effect on an array outside the function lstOut[i] += a[i]; } ); }
);
// lstOut --> ["aA1", "bB2", "cC3"]</lang>
Functional composition
ES5
Functional options include folding across an array of arrays with the built-in Array.reduce(), using a zipWith() function of suitable arity, or mapping over the output of a generic (any arity) zip() function.
(The generic zip function is the most tolerant – it simply ignores further elements in any arrays which are longer than the shortest array).
Reduce / fold:
<lang JavaScript>(function (lstArrays) {
return lstArrays.reduce( function (a, e) { return [ a[0] + e[0], a[1] + e[1], a[2] + e[2] ]; }, [, , ] // initial copy of the accumulator ).join('\n');
})([
["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]
]);</lang>
A fixed arity ZipWith:
<lang JavaScript>(function (x, y, z) {
// function of arity 3 mapped over nth items of each of 3 lists // (a -> b -> c -> d) -> [a] -> [b] -> [c] -> [d] function zipWith3(f, xs, ys, zs) { return zs.length ? [f(xs[0], ys[0], zs[0])].concat( zipWith3(f, xs.slice(1), ys.slice(1), zs.slice(1))) : []; }
function concat(x, y, z) { return .concat(x, y, z); }
return zipWith3(concat, x, y, z).join('\n')
})(["a", "b", "c"], ["A", "B", "C"], [1, 2, 3]);</lang>
Or we could write a generic zipListsWith applying some supplied function overs lists derived from the nth members of an arbitrary list of (equal-length) lists.
<lang JavaScript>(function () {
'use strict';
// zipListsWith :: ([a] -> b) -> a -> b function zipListsWith(f, xss) { return (xss.length ? xss[0] : []) .map(function (_, i) { return f(xss.map(function (xs) { return xs[i]; })); }); }
// concat :: [a] -> s function concat(lst) { return .concat.apply(, lst); }
// TEST return zipListsWith( concat, [["a", "b", "c"], ["A", "B", "C"], [1, 2, 3]] ) .join('\n');
})();</lang>
- Output:
<lang JavaScript>aA1 bB2 cC3</lang>
ES6
By transposition: <lang JavaScript>(() => {
'use strict';
// GENERIC FUNCTIONS -----------------------------------------------------
// concat :: a -> [a] const concat = xs => xs.length > 0 ? (() => { const unit = typeof xs[0] === 'string' ? : []; return unit.concat.apply(unit, xs); })() : [];
// map :: (a -> b) -> [a] -> [b] const map = (f, xs) => xs.map(f);
// transpose :: a -> a const transpose = xs => xs[0].map((_, col) => xs.map(row => row[col]));
// unlines :: [String] -> String const unlines = xs => xs.join('\n');
// TEST ------------------------------------------------------------------ const xs = [ ['a', 'b', 'c'], ['A', 'B', 'C'], [1, 2, 3] ];
return unlines( map(concat, transpose(xs)) );
})();</lang>
- Output:
aA1 bB2 cC3
jq
The following solution is based on the assumption that all the arrays can be presented as an array of arrays. This allows any number of arrays to be handled.
Specifically, zip/0 expects an array of 0 or more arrays as its input. The first array determines the number of items in the output; nulls are used for padding. <lang jq># zip/0 emits [] if input is [].
def zip:
. as $in | [range(0; $in[0]|length) as $i | $in | map( .[$i] ) ];</lang>
Example 1:
[["a","b","c"], ["A","B","C"], [1,2,3]] | zip
- Output:
[["a","A",1],["b","B",2],["c","C",3]]
To obtain the compact output used in the the task description, we can filter the results through a "pretty-print" function:
def pp: reduce .[] as $i (""; . + "\($i)");
Example 2:
[["a","b","c"], ["A","B","C"], [1,2,3]] | zip | map(pp)
- Output:
[ "aA1", "bB2", "cC3" ]
As already mentioned, the above definition of zip/0 privileges the first
array,
and if the subsequent arrays are of different lengths, null is used as a
filler.
Thus:
[["a","b","c"], ["A","B"], [1]] | zip
produces:
[["a","A",1],["b","B",null],["c",null,null]]
Handling jagged input An alternative approach would be use a variant of zip/0 that pads all arrays shorter than the longest with nulls. Here is such a variant: <lang jq>
- transpose a possibly jagged matrix
def transpose:
if . == [] then [] else (.[1:] | transpose) as $t | .[0] as $row | reduce range(0; [($t|length), (.[0]|length)] | max) as $i ([]; . + [ [ $row[$i] ] + $t[$i] ]) end;
</lang>
Jsish
<lang javascript>/* Loop over multiple arrays, in Jsish */ var a1 = ['a', 'b', 'c']; var a2 = ['A', 'B', 'C']; var a3 = [1, 2, 3];
puts('Equal sizes'); var arr = [a1, a2, a3]; var m = arr[0].length; for (var a of arr) if (a.length > m) m = a.length; for (var i = 0; i < m; i++) printf("%q%q%q\n", a1[i], a2[i], a3[i]);
puts('\nUnequal sizes'); var a4 = []; var a5 = [4,5,6,7];
arr = [a1, a2, a3, a4, a5]; m = arr[0].length; for (a of arr) if (a.length > m) m = a.length; for (i = 0; i < m; i++) printf("%q%q%q%q%q\n", a1[i], a2[i], a3[i], a4[i], a5[i]);
/*
!EXPECTSTART!
Equal sizes aA1 bB2 cC3
Unequal sizes aA1undefined4 bB2undefined5 cC3undefined6 /home/btiffin/forge/jsi/jsi-test/rosetta/loopOverMultipleArrays.jsi:19: warn: call with undefined var for argument arg 2 '...', in call to 'printf' <undefined>. (at or near "%q%q%q%q%q ")
undefinedundefinedundefinedundefined7
!EXPECTEND!
- /</lang>
- Output:
prompt$ jsish -u loopOverMultipleArrays.jsi [PASS] loopOverMultipleArrays.jsi
Julia
With a higher order function: <lang julia>foreach(println, ('a', 'b', 'c'), ('A', 'B', 'C'), (1, 2, 3))</lang>
With a loop: <lang julia>for (i, j, k) in zip(('a', 'b', 'c'), ('A', 'B', 'C'), (1, 2, 3))
println(i, j, k)
end</lang>
- Output:
aA1 bB2 cC3
K
<lang K>{,/$x}'+("abc";"ABC";1 2 3)</lang>
- Output:
("aA1" "bB2" "cC3")
If the length of the arrays are different, then K croaks with "length error".
The following is a more general approach where <lang K>
&/#:'x
</lang>
calculates the minimum length of the arrays and is used to index the first elements in each array.
<lang K>
{+x[;!(&/#:'x)]}("abc";"ABC";"1234")
</lang>
- Output:
("aA1" "bB2" "cC3")
If the arrays are of different type, then the arrays must be converted to strings.
<lang K>
{a:,/'($:'x);+a[;!(&/#:'a)]}("abc";"ABC";1 2 3 4)
</lang>
Kotlin
<lang scala>// version 1.0.6
fun main(args: Array<String>) {
val a1 = charArrayOf('a', 'b', 'c') val a2 = charArrayOf('A', 'B', 'C') val a3 = intArrayOf(1, 2, 3) for(i in 0 .. 2) println("${a1[i]}${a2[i]}${a3[i]}") println() // For arrays of different sizes we would need to iterate up to the mimimm size of all 3 in order // to get a contribution from each one. val a4 = intArrayOf(4, 5, 6, 7) val a5 = charArrayOf('d', 'e') val minSize = Math.min(a2.size, Math.min(a4.size, a5.size)) // minimum size of a2, a4 and a5 for(i in 0 until minSize) println("${a2[i]}${a4[i]}${a5[i]}")
}</lang>
- Output:
aA1 bB2 cC3 A4d B5e
Lambdatalk
<lang scheme> 1) loop over 3 sentences of equal length and returning 3 sentences
{def A a b c} -> A {def B A B C} -> B {def C 1 2 3} -> C
{S.map {lambda {:i} {br}{S.get :i {A}}
{S.get :i {B}} {S.get :i {C}} } {S.serie 0 {- {S.length {A}} 1}}}
-> a A 1 b B 2 c C 3
2) loop over 3 arrays of equal length and returning 3 arrays
{def maps
{lambda {:a :b :c} {S.map {{lambda {:a :b :c :i} {br}{A.new {A.get :i :a} {A.get :i :b} {A.get :i :c}} } :a :b :c} {S.serie 0 {- {A.length :a} 1}}}}}
-> maps
{def P {A.new a b c}} -> P {def Q {A.new A B C}} -> Q {def R {A.new 1 2 3}} -> R
{maps {P} {Q} {R}} -> [a,A,1] [b,B,2] [c,C,3]
3) loop over 3 words of inegal length and returning words
{def X James} -> X {def Y Bond} -> Y {def Z 007} -> Z
{S.map {lambda {:i} {br}{W.get :i {X}}
{W.get :i {Y}} {W.get :i {Z}} } {S.serie 0 {- {W.length {X}} 1}}}
-> J B 0 a o 0 m n 7 e d s </lang>
LFE
<lang lisp> (lists:zipwith3
(lambda (i j k) (io:format "~s~s~p~n" `(,i ,j ,k))) '(a b c) '(A B C) '(1 2 3))
</lang>
If any of the data lists differ in size from the other,
the results will print out up to the shortest data list,
and then raise a function_clause
error.
Erlang, and thus LFE, have zipwith
and
zipwith3
for working with 2 and 3 simultaneous
sets of data respectively.
If you need more than that, you'll need to create your own "zip"
function with something like (: lists map
...)
.
Liberty BASIC
<lang lb>a$(1)="a" : a$(2)="b" : a$(3)="c" b$(1)="A" : b$(2)="B" : b$(3)="C" c(1)=1 : c(2)=2 : c(3)=3
for i = 1 to 3
print a$(i);b$(i);c(i)
next</lang>
Lisaac
<lang Lisaac>Section Header
+ name := ARRAY_LOOP_TEST;
Section Public
- main <- (
+ a1, a2 : ARRAY[CHARACTER]; + a3 : ARRAY[INTEGER];
a1 := ARRAY[CHARACTER].create 1 to 3; a2 := ARRAY[CHARACTER].create 1 to 3; a3 := ARRAY[INTEGER].create 1 to 3;
1.to 3 do { i : INTEGER; a1.put ((i - 1 + 'a'.code).to_character) to i; a2.put ((i - 1 + 'A'.code).to_character) to i; a3.put i to i; };
1.to 3 do { i : INTEGER; a1.item(i).print; a2.item(i).print; a3.item(i).print; '\n'.print; };
);</lang>
LiveCode
Arrays <lang LiveCode>command loopArrays
local lowA, uppA, nums, z put "a,b,c" into lowA put "A,B,C" into uppA put "1,2,3" into nums split lowA by comma split uppA by comma split nums by comma
repeat with n = 1 to the number of elements of lowA put lowA[n] & uppA[n] & nums[n] & return after z end repeat put z
end loopArrays</lang> "list" processing <lang LiveCode>command loopDelimitedList
local lowA, uppA, nums, z put "a,b,c" into lowA put "A,B,C" into uppA put "1,2,3" into nums
repeat with n = 1 to the number of items of lowA put item n of lowA & item n of uppA & item n of nums
& return after z
end repeat put z
end loopDelimitedList</lang> Output - both behave similarly for this exercise.
aA1 bB2 cC3
When there are fewer elements than the first (or whatever the loop is based on), livecode will add an "empty" value. If we add a "d" to lowA and a 4 to nums we get the following:
aA1 bB2 cC3 d4
Logo
<lang logo>show (map [(word ?1 ?2 ?3)] [a b c] [A B C] [1 2 3])
; [aA1 bB2 cC3]
(foreach [a b c] [A B C] [1 2 3] [print (word ?1 ?2 ?3)]) ; as above, one per line</lang>
Lua
This can be done with a simple for loop: <lang lua> a1, a2, a3 = {'a' , 'b' , 'c' } , { 'A' , 'B' , 'C' } , { 1 , 2 , 3 } for i = 1, 3 do print(a1[i]..a2[i]..a3[i]) end </lang> but it may be more enlightening (and in line with the spirit of the challenge) to use the generic for: <lang lua> function iter(a, b, c)
local i = 0 return function() i = i + 1 return a[i], b[i], c[i] end
end
for u, v, w in iter(a1, a2, a3) do print(u..v..w) end </lang>
Maple
<lang Maple># Set up L := [["a", "b", "c"],["A", "B", "C"], ["1", "2", "3"]]; M := Array(1..3, 1..3, L);
multi_loop := proc(M) local i, j; for i from 1 to upperbound(M, 1) do for j from 1 to upperbound(M, 2) do printf("%s", M[j, i]); end do; printf("\n"); end do; end proc:
multi_loop(M);</lang>
- Output:
aA1 bB2 cC3
Mathematica /Wolfram Language
This can be done with a built-in function: <lang mathematica>MapThread[Print, {{"a", "b", "c"}, {"A", "B", "C"}, {1, 2, 3}}];</lang> All arguments must be lists of the same length.
Mercury
<lang>
- - module multi_array_loop.
- - interface.
- - import_module io.
- - pred main(io::di, io::uo) is det.
- - implementation.
- - import_module char, list, string.
main(!IO) :-
A = ['a', 'b', 'c'], B = ['A', 'B', 'C'], C = [1, 2, 3], list.foldl_corresponding3(print_elems, A, B, C, !IO).
- - pred print_elems(char::in, char::in, int::in, io::di, io::uo) is det.
print_elems(A, B, C, !IO) :-
io.format("%c%c%i\n", [c(A), c(B), i(C)], !IO).
</lang> The foldl_corresponding family of procedures all throw a software_error/1 exception if the lengths of the lists are not the same.
Modula-3
<lang modula3>MODULE MultiArray EXPORTS Main;
IMPORT IO, Fmt;
TYPE ArrIdx = [1..3];
VAR
arr1 := ARRAY ArrIdx OF CHAR {'a', 'b', 'c'}; arr2 := ARRAY ArrIdx OF CHAR {'A', 'B', 'C'}; arr3 := ARRAY ArrIdx OF INTEGER {1, 2, 3};
BEGIN
FOR i := FIRST(ArrIdx) TO LAST(ArrIdx) DO IO.Put(Fmt.Char(arr1[i]) & Fmt.Char(arr2[i]) &
Fmt.Int(arr3[i]) & "\n");
END;
END MultiArray.</lang>
MUMPS
Pieces of String version <lang MUMPS> LOOPMULT
N A,B,C,D,% S A="a,b,c,d" S B="A,B,C,D" S C="1,2,3" S D="," F %=1:1:$L(A,",") W !,$P(A,D,%),$P(B,D,%),$P(C,D,%) K A,B,C,D,% Q
</lang> When there aren't enough elements, a null string will be returned from the $Piece function.
- Output:
USER>d LOOPMULT^ROSETTA aA1 bB2 cC3 dD
Local arrays version <lang MUMPS> LOOPMULU
N A,B,C,D,% S A(1)="a",A(2)="b",A(3)="c",A(4)="d" S B(1)="A",B(2)="B",B(3)="C",B(4)="D" S C(1)="1",C(2)="2",C(3)="3" ; will error S %=$O(A("")) F Q:%="" W !,A(%),B(%),C(%) S
%=$O(A(%))
S %=$O(A("")) F Q:%="" W !,$G(A(%)),$G(B(%)),$G(C(%)) S %=$O(A(%)) K A,B,C,D,%
</lang>
The commented out line will throw an <UNDEFINED> error when trying
to look up D(4). Using the $Get function as a wrapper means that if the subscript for the array doesn't exist, a null string will be returned.
This same syntax is used for globals (permanent variables, that have a caret "^" as the first character).
- Output:
USER>D LOOPMULU^ROSETTA aA1 bB2 cC3 dD USER>D LOOPMULV^ROSETTA aA1 bB2 cC3 dD S %=$O(A("")) F Q:%="" W !,A(%),B(%),C(%) S %=$O(A(%)) ^ <UNDEFINED>LOOPMULV+5^ROSETTA *C(4)
Nanoquery
<lang Nanoquery>list1 = {{"a","b","c"}, {"A","B","C"}, {"1","2","3"}} for i in range(0, len(list1) - 1)
for lista in list1 print lista[i] end for println
end for</lang>
Nemerle
It "feels" better to use zip() for this, unfortunately the built in zip() only takes two lists. <lang Nemerle>using System; using System.Console;
module LoopMultiple {
Zip3[T1, T2, T3] (x : list[T1], y : list[T2], z : list[T3]) :
list[T1 * T2 * T3]
{ |(x::xs, y::ys, z::zs) => (x, y, z)::Zip3(xs, ys, zs) |([], [], []) => [] |(_, _, []) => throw ArgumentNullException() |(_, [], _) => throw ArgumentNullException() |([], _, _) => throw ArgumentNullException() } Main() : void { def first = ['a', 'b', 'c']; def second = ["A", "B", "C"]; def third = [1, 2, 3]; foreach ((x, y, z) in Zip3(first, second, third)) WriteLine($"$x$y$z"); }
}</lang>
Alternately:
<lang Nemerle>using System.Console;
module LoopMult {
Main() : void { def first = array['a', 'b', 'c']; def second = array['A', 'B', 'C']; def third = array[1, 2, 3]; when (first.Length == second.Length && second.Length ==
third.Length)
foreach (i in [0 .. (first.Length - 1)]) WriteLine("{0}{1}{2}", first[i], second[i], third[i]); }
}</lang>
NetRexx
<lang NetRexx>/* NetRexx */ options replace format comments java crossref savelog symbols nobinary
say 'Using arrays' aa = ['a', 'b', 'c', 'd'] bb = ['A', 'B', 'C'] cc = [1, 2, 3, 4]
loop x_ = 0 for aa.length
do ax = aa[x_] catch ArrayIndexOutOfBoundsException ax = ' ' end do bx = bb[x_] catch ArrayIndexOutOfBoundsException bx = ' ' end do cx = cc[x_] catch ArrayIndexOutOfBoundsException cx = ' ' end
say ax || bx || cx end x_
say 'Using indexed strings (associative arrays)' ai = sampleData('a b c d') bi = sampleData('A B C') ci = sampleData('1 2 3 4')
loop x_ = 1 to ai[0]
say ai[x_] || bi[x_] || ci[x_] end x_
method sampleData(arg) public static returns Rexx
smp = ' ' smp[0] = arg.words loop i_ = 1 to smp[0] smp[i_] = arg.word(i_) end i_ return smp
</lang>
- Output:
Using arrays aA1 bB2 cC3 d 4 Using indexed strings (associative arrays) aA1 bB2 cC3 d 4
NewLISP
<lang NewLISP>(map println '(a b c) '(A B C) '(1 2 3))</lang>
Nim
<lang nim>let
a = @['a','b','c'] b = @["A","B","C"] c = @[1,2,3]
for i in 0..2:
echo a[i], b[i], c[i]</lang>
NS-HUBASIC
<lang NS-HUBASIC>10 DIM A$(3) 20 DIM B$(3) 30 DIM C$(3) 40 A$(1)="THIS" 50 A$(2)=" LOOPS" 60 A$(3)=" ARRAYS" 70 B$(1)=" NS-HUBASIC" 80 B$(2)=" OVER" 90 B$(3)=" AT" 100 C$(1)=" PROGRAM" 110 C$(2)=" MULTIPLE" 120 C$(3)=" ONCE." 130 FOR I=1 TO 3 140 PRINT A$(I)B$(I)C$(I) 150 NEXT</lang>
Oberon-2
Works with oo2c version 2 <lang oberon2> MODULE LoopMArrays; IMPORT Out; VAR x,y: ARRAY 3 OF CHAR; z: ARRAY 3 OF INTEGER;
PROCEDURE DoLoop; VAR i: INTEGER; BEGIN i := 0; WHILE i < LEN(x) DO Out.Char(x[i]);Out.Char(y[i]);Out.LongInt(z[i],0);Out.Ln; INC(i) END END DoLoop;
BEGIN
x[0] := 'a';y[0] := 'A';z[0] := 1;
x[1] := 'b';y[1] := 'B';z[1] := 2;
x[2] := 'c';y[2] := 'C';z[2] := 3;
DoLoop
END LoopMArrays.
</lang>
Output:
aA1 bB2 cC3
Objeck
<lang objeck> class MultipleArrayAccess {
function : Main(args : String[]) ~ Nil { a := ["a", "b", "c"]; b := ["A", "B", "C"]; c := [1, 2, 3]; each(i : a) { a[i]->Append(b[i]); a[i]->Append(c[i]); a[i]->PrintLine(); }; }
} </lang>
If the arrays are different lengths, then an out-of-bounds error will be raised.
OCaml
an immediate solution: <lang ocaml>let a1 = [| 'a'; 'b'; 'c' |] and a2 = [| 'A'; 'B'; 'C' |] and a3 = [| '1'; '2'; '3' |] ;;
Array.iteri (fun i c1 ->
print_char c1; print_char a2.(i); print_char a3.(i); print_newline()
) a1 ;;</lang>
a more generic solution could be to use a function which iterates over a list of arrays:
<lang ocaml>let n_arrays_iter ~f = function
| [] -> () | x::xs as al -> let len = Array.length x in let b = List.for_all (fun a -> Array.length a = len) xs in if not b then invalid_arg "n_arrays_iter: arrays of different
length";
for i = 0 to pred len do let ai = List.map (fun a -> a.(i)) al in f ai done</lang>
this function raises Invalid_argument exception if arrays have different
length,
and has this signature:
<lang ocaml>val n_arrays_iter : f:('a list -> unit) -> 'a array list -> unit</lang>
how to use it with arrays a1, a2 and a3 defined before:
<lang ocaml>let () =
n_arrays_iter [a1; a2; a3] ~f:(fun l -> List.iter print_char l; print_newline());
- </lang>
Oforth
If arrays don't have the same size, zipAll reduces to the minimum size
<lang Oforth>[ "a", "b", "c" ] [ "A", "B", "C" ] [ 1, 2, 3 ] zipAll(3) apply(#[ apply(#print) printcr ])</lang>
- Output:
aA1 bB2 cC3
ooRexx
<lang ooRexx> x = .array~of("a", "b", "c") y = .array~of("A", "B", "C") z = .array~of(1, 2, 3)
loop i = 1 to x~size
say x[i]y[i]z[i]
end </lang>
Oz
<lang oz>for
I in [a b c] J in ['A' 'B' 'C'] K in [1 2 3]
do
{System.showInfo I#J#K}
end</lang>
The loop will stop when the shortest list is exhausted.
PARI/GP
This version stops when the shortest vector is exhausted. <lang parigp>loopMultiple(V)={
my(l=#V[1]); for(i=2,#V,l=min(l,#V[i])); for(i=1,#V[1], for(j=1,#V, print1(V[j][i]) ); print() )
};</lang>
This version prints blanks when a vector is exhausted. <lang parigp>loopMultiple(V)={
my(l=0); for(i=1,#V,l=max(l,#V[i])); for(i=1,#V[1], for(j=1,#V, if(#V[j]<i, print1(" ") , print1(V[j][i]) ) ); print() )
};</lang>
Pascal
See Delphi
Perl
<lang perl>sub zip (&@) {
my $code = shift; my $min; $min = $min && $#$_ > $min ? $min : $#$_ for @_;
for my $i(0..$min){ $code->(map $_->[$i] ,@_) }
} my @a1 = qw( a b c ); my @a2 = qw( A B C ); my @a3 = qw( 1 2 3 );
zip { print @_,"\n" }\(@a1, @a2, @a3);</lang>
This implementation will stop producing items when the shortest array
ends.
Phix
Assumes a and b are strings and c is a sequence of integers.
If the arguments were not all the same length, attempts to retrieve non-existent elements could trigger a fatal run-time error, were it not for the min(). In print3, fairly obviously, we only extract up to the shortest length. The builtin columnize() routine can perform a similar task: I have provided a space defval and replaced the 3rd array with a string to ensure we get strings back, and extended it to show how columnize uses that default value for missing entries off the end of the first two arrays.
procedure print3(sequence a, b, c) for i=1 to min({length(a),length(b),length(c)}) do printf(1, "%s%s%g\n", {a[i], b[i], c[i]}) end for end procedure print3("abc","ABC",{1, 2, 3}) ?columnize({"abc","ABC","1234"},{},' ')
- Output:
aA1 bB2 cC3 {"aA1","bB2","cC3"," 4"}
Phixmonti
<lang Phixmonti>include ..\Utilitys.pmt
( "abc" "ABC" "123" )
len var dim1 1 get len var dim2 drop
dim1 for
var i dim2 for var j ( j i ) sget tochar print endfor nl
endfor </lang>
PHP
<lang PHP>$a = array('a', 'b', 'c'); $b = array('A', 'B', 'C'); $c = array('1', '2', '3'); //These don't *have* to be strings, but it saves PHP from casting them later
if ((sizeOf($a) !== sizeOf($b)) || (sizeOf($b) !== sizeOf($c))){
throw new Exception('All three arrays must be the same length');
} foreach ($a as $key => $value){
echo "{$a[$key]}{$b[$key]}{$c[$key]}\n";
}</lang>
This implementation throws an exception if the arrays are not all the same length.
PicoLisp
<lang PicoLisp>(mapc prinl
'(a b c) '(A B C) (1 2 3) )</lang>
The length of the first argument list controls the operation. If subsequent lists are longer, their remaining values are ignored. If they are shorter, NIL is passed to the function.
Pike
Could be done with for, but foreachs included index counter avoids the usual off-by-one errors
<lang Pike>
array a1 = ({ "a", "b", "c" }); array a2 = ({ "A", "B", "C" }); array a3 = ({ "1", "2", "3" });
foreach(a1; int index; string char_dummy) write("%s%s%s\n", a1[index], a2[index], a3[index]);
</lang>
- Output:
aA1 bB2 cC3
PL/I
<lang pli> declare P(3) character (1) initial ('a', 'b', 'c'),
Q(3) character (1) initial ('A', 'B', 'C'), R(3) fixed decimal (1) initial (1, 2, 3);
do i = lbound(P,1) to hbound(P,1);
put skip edit (P(i), Q(i), R(i)) (2 A, F(1));
end; </lang>
PostScript
<lang postscript> % transpose is defined in initlib like this. /transpose {
[ exch { { {empty? exch pop} map all?} {pop exit} ift [ exch {} {uncons {exch cons} dip exch} fold counttomark 1 roll] uncons } loop ] {reverse} map
}.
% using it. [[/a /b /c] [/A /B /C] [1 2 3]] transpose </lang>
PowerBASIC
<lang powerbasic>FUNCTION PBMAIN () AS LONG
DIM x(2), y(2) AS STRING * 1 DIM z(2) AS LONG
'data ARRAY ASSIGN x() = ("a", "b", "c") ARRAY ASSIGN y() = ("A", "B", "C") ARRAY ASSIGN z() = (1, 2, 3)
'set upper bound C& = UBOUND(x) IF UBOUND(y) > C& THEN C& = UBOUND(y) IF UBOUND(z) > C& THEN C& = UBOUND(z)
OPEN "output.txt" FOR OUTPUT AS 1 FOR L& = 0 TO C& IF L& <= UBOUND(x) THEN PRINT #1, x(L&); IF L& <= UBOUND(y) THEN PRINT #1, y(L&); IF L& <= UBOUND(z) THEN PRINT #1, TRIM$(STR$(z(L&))); PRINT #1, NEXT CLOSE
END FUNCTION</lang>
PowerShell
A cheap and chEasy 'zip' function: <lang PowerShell> function zip3 ($a1, $a2, $a3) {
while ($a1) { $x, $a1 = $a1 $y, $a2 = $a2 $z, $a3 = $a3 [Tuple]::Create($x, $y, $z) }
} </lang> <lang PowerShell> zip3 @('a','b','c') @('A','B','C') @(1,2,3) </lang>
- Output:
Item1 Item2 Item3 ----- ----- ----- a A 1 b B 2 c C 3
<lang PowerShell> zip3 @('a','b','c') @('A','B','C') @(1,2,3) | ForEach-Object {$_.Item1 + $_.Item2 + $_.Item3} </lang>
- Output:
aA1 bB2 cC3
Prolog
Works with SWI-Prolog <lang Prolog>multiple_arrays(L1, L2, L3) :- maplist(display, L1, L2, L3).
display(A,B,C) :- writef('%s%s%s\n', [[A],[B],[C]]). </lang>
- Output:
?- multiple_arrays("abc", "ABC", "123"). aA1 bB2 cC3 true. ?- multiple_arrays("abc", "AB", "123"). aA1 bB2 false.
PureBasic
<lang PureBasic>OpenConsole()
- Fill arrays
Dim a.s(2) Dim b.s(2) Dim c(2) For Arrayposition = 0 To ArraySize(a())
a(Arrayposition) = Chr(Asc("a") + Arrayposition) b(Arrayposition) = Chr(Asc("A") + Arrayposition) c(Arrayposition) = Arrayposition + 1
Next
- loop over them
For Arrayposition = 0 To ArraySize(a())
PrintN(a(Arrayposition) + b(Arrayposition) + Str(c(Arrayposition)))
Next Input() ;wait for Enter before ending</lang>
If they have different lengths there are two cases:
a() is the shortest one: Only elements up to maximum index of a() are
printed
a() is bigger than another one: if exceeding index to much, program
crashes,
else it may work because there is some "free space" after end of
assigned array memory.
For example if a has size 4, line dD4 will also be printed. size 20
leads to an crash
This is because ReDim becomes slow if everytime there is a change to
array size new memory has to be allocated.
Python
Using zip(): <lang python>>>> print ( '\n'.join(.join(x) for x in zip('abc', 'ABC', '123')) ) aA1 bB2 cC3 >>></lang> If lists are different lengths, zip() stops after the shortest one.
Using map(): <lang python>>>> print(*map(.join, zip('abc', 'ABC', '123')), sep='\n') aA1 bB2 cC3 >>></lang> If lists are different lengths, map() in Python 2.x pretends that the shorter lists were extended with None items; map() in Python 3.x stops after the shortest one.
Using itertools.imap() (Python 2.x): <lang python>from itertools import imap
def join3(a,b,c):
print a+b+c
imap(join3,'abc','ABC','123')</lang> If lists are differnt lengths, imap() stops after the shortest is exhausted.
Python 3.X has zip_longest which fills shorter iterables with its fillvalue argument which defaults to None (similar to the behavior of map() in Python 2.x): <lang python>>>> from itertools import zip_longest >>> print ( '\n'.join(.join(x) for x in zip_longest('abc', 'ABCD', '12345', fillvalue='#')) ) aA1 bB2 cC3
- D4
- 5
>>></lang> (The Python 2.X equivalent is itertools.izip_longest)
Quackery
The code presented here will loop as many times as the number of characters in the first nest (i.e. "abc" in the example). If either of the other two nests are shorter than the first then the program will report a problem.
<lang Quackery> [ rot witheach
[ emit over i^ peek emit dup i^ peek emit cr ] 2drop ] is task ( $ $ $ --> )
$ "abc" $ "ABC" $ "123" task</lang>
- Output:
aA1 bB2 cC3
R
<lang R>multiloop <- function(...) {
# Retrieve inputs and convert to a list of character strings arguments <- lapply(list(...), as.character) # Get length of each input lengths <- sapply(arguments, length)
# Loop over elements for(i in seq_len(max(lengths))) { # Loop over inputs for(j in seq_len(nargs())) { # print a value or a space (if that input has finished) cat(ifelse(i <= lengths[j], argumentsj[i], " ")) } cat("\n") }
} multiloop(letters[1:3], LETTERS[1:3], 1:3)</lang>
Same thing as a single function call. But throws error if the arrays differ in length.
<lang R> apply(data.frame(letters[1:3], LETTERS[1:3], 1:3), 1,
function(row) { cat(row, "\n", sep=) })
</lang>
Racket
Racket for loops can loop over an arbitrary number of sequences of any kind at once:
<lang racket>
- lang racket
(for ([x '(a b c)] ; list
[y #(A B C)] ; vector [z "123"] [i (in-naturals 1)]) ; 1, 2, ... infinitely (printf "~s: ~s ~s ~s\n" i x y z))
</lang>
The loop stops as soon as the first sequence terminates -- in the above case i can iterate forever but looping stops when we reach the end of the list/vector/string. (The same holds for multiple containers of different sizes.)
Raku
(formerly Perl 6)
Note that all of the following work with any iterable object, (array, list, range, sequence; anything that does the Iterable role), not just arrays.
Basic functionality
<lang perl6>for <a b c> Z <A B C> Z 1, 2, 3 -> ($x, $y, $z) {
say $x, $y, $z;
}</lang>
The Z
operator stops emitting items as soon as the shortest input list is exhausted. However, short lists are easily extended by replicating all or part of the list, or by appending any kind of lazy list generator to supply default values as necessary.
Since Z
will return a list of lists (in this example, the first list is ('a', 'A', 1)
, parentheses are used around in the lambda signature ($x, $y, $z)
to unpack the list for each iteration.
Factoring out concatenation
Note that we can also factor out the concatenation by making the Z metaoperator apply the ~ concatenation operator across each triple:
<lang perl6>.say for <a b c> Z~ <A B C> Z~ 1, 2, 3;</lang>
We could also use the zip-to-string with the reduction metaoperator:
<lang perl6>.say for [Z~] <a b c>, <A B C>, (1,2,3);</lang>
We could also write that out "long-hand":
<lang perl6>.say for zip :with(&infix:<~>), <a b c>, <A B C>, (1,2,3);</lang>
returns the exact same result so if you aren't comfortable with the concise operators, you have a choice.
A list and its indices
The common case of iterating over a list and a list of its indices can be done using the same method:
<lang perl6>for ^Inf Z <a b c d> -> ($i, $letter) { ... }</lang>
or by using the .kv
(key and value) method on the list (and dropping the parentheses because the list returned by .kv
is a flattened list):
<lang perl6>for <a b c d>.kv -> $i, $letter { ... }</lang>
Iterate until all exhausted
If you have different sized lists that you want to pull a value from each per iteration, but want to continue until all of the lists are exhausted, we have roundrobin
.
<lang perl6>.put for roundrobin <a b c>, 'A'..'G', ^5;</lang>
- yields:
a A 0 b B 1 c C 2 D 3 E 4 F G
Red
The word repeat evaluates a given block! a specified number of times and exposes the count value to the block! being executed. When a variable is used in a path notation, we put a colon in front of it. :counter
<lang Red>>>blk: [["a" "b" "c"] ["A" "B" "C"] [1 2 3]] == [["a" "b" "c"] ["A" "B" "C"] [1 2 3]]
>> repeat counter 3 [print [blk/1/:counter blk/2/:counter blk/3/:counter]] a A 1 b B 2 c C 3</lang>
REXX
same size arrays
If any of the array's elements are missing or it is a short list,
a blank is substituted to retain visual fidelity in the output.
When all elements are blank, then it signifies the end of the arrays.
<lang rexx>/*REXX program shows how to simultaneously loop over multiple arrays.*/
x. = ' '; x.1 = "a"; x.2 = 'b'; x.3 = "c"
y. = ' '; y.1 = "A"; y.2 = 'B'; y.3 = "C"
z. = ' '; z.1 = "1"; z.2 = '2'; z.3 = "3"
do j=1 until output= output = x.j || y.j || z.j say output end /*j*/ /*stick a fork in it, we're done.*/</lang>
output
aA1 bB2 cC3
dissimilar sized arrays
In this example, two of the arrays are extended (past the 1st example).
Also note that REXX doesn't require quotes around non-negative numbers (they're optional).
<lang rexx>/*REXX program shows how to simultaneously loop over multiple arrays.*/
x.=' '; x.1="a"; x.2='b'; x.3="c"; x.4='d'
y.=' '; y.1="A"; y.2='B'; y.3="C";
z.=' '; z.1= 1 ; z.2= 2 ; z.3= 3 ; z.4= 4; z.5= 5
do j=1 until output= output=x.j || y.j || z.j say output end /*j*/ /*stick a fork in it, we're done.*/</lang>
output
aA1 bB2 cC3 d 4 5
dissimilar sized lists
<lang rexx>/*REXX program shows how to simultaneously loop over multiple lists.*/ x = 'a b c d' y = 'A B C' z = 1 2 3 4
do j=1 until output= output = word(x,j) || word(y,j) || word(z,j) say output end /*j*/ /*stick a fork in it, we're done.*/</lang>
output
aA1 bB2 cC3 d4
idiomatic method for lists
<lang rexx>/*REXX program shows how to simultaneously loop over multiple lists.*/ x = 'a b c d' y = 'A B C' z = 1 2 3 4 ..LAST
do j=1 for max(words(x), words(y), words(z)) say word(x,j) || word(y,j) || word(z,j) end /*j*/ /*stick a fork in it, we're done.*/</lang>
output
aA1 bB2 cC3 d4 ..LAST
Ring
<lang ring> array1 = ["a", "b", "c"] array2 = ["A", "B", "C"] array3 = [1, 2, 3]
for n = 1 to 3
see array1[n] + array2[n] + array3[n] + nl
next </lang>
Ruby
<lang ruby>['a','b','c'].zip(['A','B','C'], [1,2,3]) {|i,j,k| puts "#{i}#{j}#{k}"}</lang> or <lang ruby>['a','b','c'].zip(['A','B','C'], [1,2,3]) {|a| puts a.join}</lang>
Both of these loops print aA1
, bB2
, cC3
.
Array#zip
iterates once for each element of the receiver.
If an argument array is longer, the excess elements are ignored.
If an argument array is shorter, the value nil
is supplied.
<lang ruby>irb(main):001:0> ['a','b','c'].zip(['A','B'], [1,2,3,4]) {|a| puts a.join}
aA1
bB2
c3
=> nil
irb(main):002:0> ['a','b','c'].zip(['A','B'], [1,2,3,4])
=> [["a", "A", 1], ["b", "B", 2], ["c", nil, 3]]</lang>
Run BASIC
<lang runbasic>for i = 1 to 3
a$(i) = chr$(i+96) b$(i) = chr$(i+64) c(i) = i
next i
for i = 1 to 3
print a$(i);b$(i);c(i)
next</lang>
Rust
<lang rust>fn main() {
let a1 = ["a", "b", "c"]; let a2 = ["A", "B", "C"]; let a3 = [1, 2, 3];
for ((&x, &y), &z) in a1.iter().zip(a2.iter()).zip(a3.iter()) { println!("{}{}{}", x, y, z); }
}</lang>
- Output:
aA1 bB2 cC3
Salmon
<lang Salmon>// First, we'll define a general-purpose zip() to zip
any
// number of lists together. function zip(...)
{ variable result; variable list_num := 0; iterate(arg; arguments) { variable elem_num := 0; iterate (x; arg) { result[elem_num][list_num] := x; ++elem_num; }; ++list_num; }; return result; };
immutable a := ["a", "b", "c"],
b := ["A", "B", "C"], c := [1, 2, 3];
iterate (x; zip(a, b, c))
print(x[0], x[1], x[2], "\n");;</lang>
The preceding code will throw an exception if the lists aren't the same length. Here's an example that will print a number of lines equal to the length of the longest list and print nothing for elements that are missing if some lists are shorter than the longest:
<lang Salmon>// First, we'll define a general-purpose zip() to zip
any
// number of lists together. function zip(...)
{ variable result := []; variable list_num := 0; iterate(arg; arguments) { variable elem_num := 0; iterate (x; arg) { if (elem_num >= length(result)) result[elem_num] := <<(* --> "")>>;; result[elem_num][list_num] := x; ++elem_num; }; ++list_num; }; return result; };
immutable a := ["a", "b", "c"],
b := ["A", "B", "C"], c := [1, 2, 3];
iterate (x; zip(a, b, c))
print(x[0], x[1], x[2], "\n");;</lang>
Sather
<lang sather>class MAIN is
main is a :ARRAY{STR} := |"a", "b", "c"|; b :ARRAY{STR} := |"A", "B", "C"|; c :ARRAY{STR} := |"1", "2", "3"|; loop i ::= 0.upto!(2); #OUT + a[i] + b[i] + c[i] + "\n"; end; end;
end;</lang>
If the index i is out of bounds, a runtime error is raised.
Scala
<lang scala> ("abc", "ABC", "123").zipped foreach { (x, y, z) =>
println(x.toString + y + z)
} </lang>
Scheme
Scheme provides for-each
and
map
to iterate a function over one or more
lists.
The map
form is used to collect the results
into a new list.
<lang scheme> (let ((a '("a" "b" "c"))
(b '("A" "B" "C")) (c '(1 2 3))) (for-each (lambda (i1 i2 i3) (display i1) (display i2) (display i3) (newline)) a b c))
</lang>
Scheme has a vector
datatype with constant-time
retrieval of items held in an ordered sequence. Use srfi-43 to get
similar iterators for vectors, vector-for-each
and vector-map
:
<lang scheme> (let ((a (vector "a" "b" "c"))
(b (vector "A" "B" "C")) (c (vector 1 2 3))) (vector-for-each (lambda (current-index i1 i2 i3) (display i1) (display i2) (display i3) (newline)) a b c))
</lang>
Note, the lists or vectors must all be of the same length.
Sidef
The simplest way is by using the Array.zip{} method: <lang ruby>[%w(a b c),%w(A B C),%w(1 2 3)].zip { |i,j,k|
say (i, j, k)
}</lang>
- Output:
aA1 bB2 cC3
Smalltalk
<lang smalltalk>|a b c| a := OrderedCollection new addAll: #('a' 'b' 'c'). b := OrderedCollection new addAll: #('A' 'B' 'C'). c := OrderedCollection new addAll: #(1 2 3).
1 to: (a size) do: [ :i |
(a at: i) display. (b at: i) display. (c at: i) displayNl.
].</lang>
If index i is out of bound, a runtime error is raised.
Actually, there is no need for the extra OrderedCollections as in the above example.
Also, most Smalltalks (all?) can concatenate non-string args¹.
At least in ST/X, the following works ¹:
<lang smalltalk>|a b c|
a := #('a' 'b' 'c'). b := #('A' 'B' 'C'). c := #(1 2 3). 1 to: (a size) do: [ :i |
((a at: i),(b at: i),(c at: i)) displayNl.
].</lang>
Another alternative is to use a multi-collection enumerator, which hides the element access (transparent to how elements are stored inside the collection): <lang smalltalk>|a b c|
a := #('a' 'b' 'c'). b := #('A' 'B' 'C'). c := #(1 2 3). a with:b with:c do:[:ai :bi :ci |
(ai,bi,ci) displayNl.
].</lang>
1) concatenation of integer objects as shown above may require a change in the , (comma) implementation, to send "asString" to the argument.
Standard ML
The below code will combine arbitrarily many lists of strings into a single list with length equal to that of the shortest list. <lang Standard ML> (*
* val combine_lists : string list list -> string list *)
fun combine_lists nil = nil | combine_lists (l1::ls) = List.foldl (ListPair.map (fn (x,y) => y ^
x)) l1 ls;
(* ["a1Ax","b2By","c3Cz"] *) combine_lists[["a","b","c"],["1","2","3"],["A","B","C"],["x","y","z"]]; </lang>
Stata
Use an index variable.
<lang stata>local u a b c local v A B C matrix w=1,2,3 forv i=1/3 { di "`: word `i' of `u`: word `i' of `v`=el("w",1,`i')'" }</lang>
Mata
<lang stata>mata u="a","b","c" v="A","B","C" w=1,2,3
for (i=1; i<=3; i++) {
printf("%s%s%f\n",u[i],v[i],w[i])
} end</lang>
SuperCollider
Using three variables and indexing (SuperCollider posts the last statement in the REPL) <lang SuperCollider>
- x, y, z = [["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]];
3.collect { |i| x[i] ++ y[i] ++ z[i] } </lang>
A more idiomatic way of writing it, independent of the number of dimensions: <lang SuperCollider> [["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]].flop.collect { |x| x.join } </lang>
Or simpler: <lang SuperCollider> [["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]].flop.collect(_.join) </lang>
Same with lamination (a concept from APL/J):
<lang SuperCollider>
["a", "b", "c"] +++ ["A", "B", "C"] +++ ["1", "2", "3"]
</lang>
Independent of dimensions: <lang SuperCollider> [["a", "b", "c"], ["A", "B", "C"], ["1", "2", "3"]].reduce('+++') </lang>
Swift
<lang Swift>let a1 = ["a", "b", "c"] let a2 = ["A", "B", "C"] let a3 = [1, 2, 3]
for i in 0 ..< a1.count {
println("\(a1[i])\(a2[i])\(a3[i])")
}</lang>
- Output:
aA1 bB2 cC3
Tailspin
Simplest iteration with an ordinary "loop" that will error on uneven sizes <lang tailspin> def x: ['a', 'b', 'c']; def y: ['A', 'B', 'C']; def z: [1, 2, 3];
1..$x::length -> '$x($);$y($);$z($); ' -> !OUT::write </lang>
- Output:
aA1 bB2 cC3
A simple transpose method that gives the same output and also errors on uneven sizes <lang tailspin> templates transpose
def a: $; def n: $(1)::length; [ 1..$n -> $a(1..last; $) ] !
end transpose
[$x, $y, $z] -> transpose... -> '$...; ' -> !OUT::write </lang> A more complex transpose that uses "foreach" more in line with the task proposal and handles uneven arrays <lang tailspin> def u: ['a', 'b']; def v: ['A', 'B', 'C']; def w: [1];
templates transpose2
@: []; $... -> \[i]( when <?($i <..$@transpose2::length>)> do ..|@transpose2($i): $; otherwise ..|@transpose2: [$];\) -> !VOID $@ !
end transpose2
[$x, $y, $z] -> transpose2... -> '$...; ' -> !OUT::write
' ' -> !OUT::write
[$u,$v,$w] -> transpose2... -> '$...; ' -> !OUT::write </lang>
- Output:
aA1 bB2 cC3 aA1 bB C
Tcl
<lang tcl>set list1 {a b c} set list2 {A B C} set list3 {1 2 3} foreach i $list1 j $list2 k $list3 {
puts "$i$j$k"
}</lang> If lists are different lengths, the manual [1] says: "The total number of loop iterations is large enough to use up all the values from all the value lists. If a value list does not contain enough elements for each of its loop variables in each iteration, empty values are used for the missing elements."
TorqueScript
<lang Torquescript> $var[0] = "a b c" $var[1] = "A B C"; $var[2] = "1 2 3";
for(%i=0;%i<3;%i++) echo(getWord($var[0],%i) @ getWord($var[1],%i) @ getWord($var[2],%i)); </lang>
TUSCRIPT
<lang tuscript> $$ MODE TUSCRIPT arr1="a'b'c" arr2="a'b'C" arr3="1'2'3" LOOP a=arr1,b=arr2,c=arr3 PRINT a,b,c ENDLOOP </lang>
- Output:
aa1 bb2 cC3
TXR
Pattern language
<lang bash>$ txr -c '@(bind a ("a" "b" "c")) @(bind b ("A" "B" "C")) @(bind c ("1" "2" "3")) @(output) @ (repeat) @a@b@c @ (end) @(end)' aA1 bB2 cC3</lang>
TXR Lisp, using mapcar
Here we actually loop over four things: three strings and an infinite list of newlines. The output is built up as one string object that is finally printed in one go.
<lang bash>$ txr -e '(pprint (mappend (op list) "abc" "ABC" "123" (repeat "\n")))' aA1 bB2 cC3</lang>
TXR Lisp, using each
<lang bash>$ txr -e '(each ((x "abc") (y "ABC") (z "123")) (put-line `@x@y@z`))' aA1 bB2 cC3</lang>
Translation of Scheme
<lang txrlisp>;; Scheme's vector-for-each: a one-liner in TXR
- that happily works over strings and lists.
- We don't need "srfi-43".
(defun vector-for-each (fun . vecs)
[apply mapcar fun (range) vecs])
(defun display (obj : (stream *stdout*))
(pprint obj stream))
(defun newline (: (stream *stdout*))
(display #\newline stream))
(let ((a (vec "a" "b" "c"))
(b (vec "A" "B" "C")) (c (vec 1 2 3))) (vector-for-each (lambda (current-index i1 i2 i3) (display i1) (display i2) (display i3) (newline)) a b c))</lang>
Translation of Logo
<lang txrlisp>(macro-time
(defun question-var-to-meta-num (var) ^(sys:var ,(int-str (cdr (symbol-name var))))))
(defmacro map (square-fun . square-args)
(tree-bind [(fun . args)] square-fun ^[apply mapcar (op ,fun ,*[mapcar question-var-to-meta-num args]) (macrolet ([(. args) ^(quote ,args)]) (list ,*square-args))]))
(defun word (. items)
[apply format nil "~a~a~a" items])
(defun show (x) (pprinl x))
(show (map [(word ?1 ?2 ?3)] [a b c] [A B C] [1 2 3]))</lang>
- Output:
(aA1 bB2 cC3)
UNIX Shell
With the Bourne shell, its for
loop (from
Loops/Foreach#UNIX Shell) can iterate only one list.
We use an index i
to access the other lists:
set -- $list
loads the positional parameters,
and shift $i
moves our element to
$1
.
<lang bash>a=a:b:c b=A:B:C c=1:2:3
oldifs=$IFS IFS=: i=0 for wa in $a; do set -- $b; shift $i; wb=$1 set -- $c; shift $i; wc=$1
printf '%s%s%s\n' $wa $wb $wc
i=`expr $i + 1` done IFS=$oldifs</lang>
- Output:
aA1 bB2 cC3
When the lists have different lengths, this code uses the length of list
a
. Longer lists ignore their extra elements,
and shorter lists give extra empty strings.
Inspired by the previous example, below is the way to
loop over two arrays simultaneously using set --
$ARGS
.
It is less general than the previous example
but it is shorter and works just fine.
<lang bash>A='a1 a2 a3' B='b1 b2 b3'
set -- $B for a in $A do
printf "$a $1\n" shift
done</lang>
- Output:
a1 b1 a2 b2 a3 b3
Some shells have real arrays, so the iteration is much more simple and easy.
<lang bash>a=(a b c) b=(A B C) c=(1 2 3) for ((i = 0; i < ${#a[@]}; i++)); do
echo "${a[$i]}${b[$i]}${c[$i]}"
done</lang>
- Output:
aA1 bB2 cC3
<lang bash>set -A a a b c set -A b A B C set -A c 1 2 3 ((i = 0)) while ((i < ${#a[@]})); do
echo "${a[$i]}${b[$i]}${c[$i]}" ((i++))
done</lang>
<lang bash>a=(a b c) b=(A B C) c=(1 2 3) for ((i = 1; i <= $#a; i++)); do
echo "$a[$i]$b[$i]$c[$i]"
done</lang>
C Shell
Uses the length of array a. Longer arrays ignore their extra elements, but shorter arrays force the shell to exit with an error like b: Subscript out of range.
<lang csh>set a=(a b c) set b=(A B C) set c=(1 2 3) @ i = 1 while ( $i <= $#a ) echo "$a[$i]$b[$i]$c[$i]" @ i += 1 end</lang>
Ursa
Looping over multiple arrays in an interactive session:
<lang ursa>> decl string<> a b c
> append (split "abc" "") a
> append (split "ABC" "") b
> append (split "123" "") c
> for (decl int i) (< i (size a)) (inc i)
.. out a b c endl console
..end
aA1
bB2
cC3
> _</lang>
If either of the arrays are smaller than (size a), then an indexerror is thrown. This could be caught with a try...catch
block.
Ursala
Compute the transpose of the list formed of the three lists. If they're of unequal lengths, an exception occurs. <lang Ursala>#show+
main = ~&K7 <'abc','ABC','123'></lang>
- Output:
aA1 bB2 cC3
Vala
<lang vala>const char a1[] = {'a','b','c'}; const char a2[] = {'A','B','C'}; const int a3[] = {1, 2, 3};
void main() {
for (int i = 0; i < 3; i++) stdout.printf("%c%c%i\n", a1[i], a2[i], a3[i]);
}</lang>
VBA
<lang vb>' Loop over multiple arrays simultaneously - VBA - 08/02/2021
Sub Main() a = Array("a","b","c") b = Array("A","B","C") c = Array(1,2,3) For i = LBound(a) To UBound(a) buf = buf & vbCrLf & a(i) & b(i) & c(i) Next Debug.Print Mid(buf,3) End Sub </lang>
- Output:
aA1 bB2 cC3
VBScript
<lang vb>' Loop over multiple arrays simultaneously - VBScript - 08/02/2021
a = Array("a","b","c") b = Array("A","B","C") c = Array(1,2,3) For i = LBound(a) To UBound(a)
buf = buf & vbCrLf & a(i) & b(i) & c(i)
Next WScript.Echo Mid(buf,3) </lang>
- Output:
aA1 bB2 cC3
Visual Basic .NET
Two implementations: one determines the shortest of the arrays and uses a simple For loop with element accesses to each array separately; one uses Enumerable.Zip (which can only zip two sequences at once) twice to create 3-tuples. Enumerable.Zip stops when either source runs out of elements, so the behavior of the two implementations is identical for arrays of different lengths. <lang vbnet> Module Program
Sub Main() Dim a As Char() = {"a"c, "b"c, "c"c} Dim b As Char() = {"A"c, "B"c, "C"c} Dim c As Integer() = {1, 2, 3}
Dim minLength = {a.Length, b.Length, c.Length}.Min() For i = 0 To minLength - 1 Console.WriteLine(a(i) & b(i) & c(i)) Next
Console.WriteLine()
For Each el As (a As Char, b As Char, c As Integer) In a.Zip(b, Function(l, r) (l, r)).Zip(c, Function(x, r) (x.l, x.r, r)) Console.WriteLine(el.a & el.b & el.c) Next End Sub
End Module</lang>
- Output:
aA1 bB2 cC3 aA1 bB2 cC3
Visual FoxPro
<lang vfp> LOCAL i As Integer, n As Integer, c As String LOCAL ARRAY a1[3], a2[3], a3[4], a[3]
- !* Populate the arrays and store the array lengths in a
a1[1] = "a" a1[2] = "b" a1[3] = "c" a[1] = ALEN(a1) a2[1] = "A" a2[2] = "B" a2[3] = "C" a[2] = ALEN(a2) a3[1] = "1" a3[2] = "2" a3[3] = "3" a3[4] = "4" a[3] = ALEN(a3)
- !* Find the maximum length of the arrays
- !* In this case, 4
n = MAX(a[1], a[2], a[3]) ? "Simple Loop" FOR i = 1 TO n
c = "" c = c + IIF(i <= a[1], a1[i], "#") c = c + IIF(i <= a[2], a2[i], "#") c = c + IIF(i <= a[3], a3[i], "#") ? c
ENDFOR
- !* Solution using a cursor
CREATE CURSOR tmp (c1 C(1), c2 C(1), c3 C(1), c4 C(3)) INSERT INTO tmp (c1, c2, c3) VALUES ("a", "A", "1") INSERT INTO tmp (c1, c2, c3) VALUES ("b", "B", "2") INSERT INTO tmp (c1, c2, c3) VALUES ("c", "C", "3") INSERT INTO tmp (c1, c2, c3) VALUES ("#", "#", "4") REPLACE c4 WITH c1 + c2 + c3 ALL ? "Solution using a cursor" LIST OFF FIELDS c4 </lang>
- Output:
Simple Loop aA1 bB2 cC3 ##4 Solution using a cursor aA1 bB2 cC3 ##4
Wart
<lang wart>each (x X n) (zip '(a b c) '(A B C) '(1 2 3))
prn x X n</lang>
Wren
The following script will work as expected provided the lengths of a1 and a2 are at least equal to the length of a3. Otherwise it will produce a 'Subscript out of bounds' error. <lang ecmascript>var a1 = ["a", "b", "c"] var a2 = ["A", "B", "C"] var a3 = [1, 2, 3] for (i in a3) System.print("%(a1[i-1])%(a2[i-1])%(i)")</lang>
- Output:
aA1 bB2 cC3
X86 Assembly
<lang asm> extern _printf
section .data
arr1 dd 3,"a","b","c" ;first dword saves the length arr2 dd 3,"A","B","C" arr3 dd 3,"1","2","3"
section .bss
arrLocation resd 4 tempOutput resd 2
section .text global _main _main:
mov [arrLocation], dword arr1 mov [arrLocation+4], dword arr2 mov [arrLocation+8], dword arr3 mov [arrLocation+12], dword 0 ;signales end mov [tempOutput+4], dword 0 ;dword 0 signales end of string mov ecx, 1 ;not 0 as in 0 the length is saved looping: mov ebx, 0 ;0 as arrLocation doesn't save length inloop: mov eax, [arrLocation+ebx*4] ;get ebxth arr address cmp eax, 0 ;if we don't get an address incresse index je incecx cmp ecx, [eax] ;when ecx is greater then the length of the current array we end jg end mov edx, [eax + ecx*4] ;get char at index ecx from arr mov [tempOutput], edx ;setup for _printf push ecx ;save ecx push tempOutput ;parameter for _printf call _printf add esp, 4 ;garbage collecting pop ecx ;restore ecx inc ebx ;get next arr jmp inloop incecx: mov [tempOutput], dword 0x0a ;after we print every element at the nth index we print a \n push ecx push tempOutput call _printf add esp, 4 pop ecx inc ecx ;increase index jmp looping end: xor eax, eax ret
</lang>
XBasic
<lang xbasic> PROGRAM "loopoverarrays"
DECLARE FUNCTION Entry()
FUNCTION Entry()
DIM arr1$[2], arr2$[2], arr3%[2] arr1$[0] = "a": arr1$[1] = "b": arr1$[2] = "c" arr2$[0] = "A": arr2$[1] = "B": arr2$[2] = "C" arr3%[0] = 1: arr3%[1] = 2: arr3%[2] = 3 FOR i% = 0 TO 2 PRINT arr1$[i%]; arr2$[i%]; FORMAT$("#", arr3%[i%]) NEXT i%
END FUNCTION END PROGRAM </lang>
XPL0
<lang XPL0>string 0; \use zero terminated strings include c:\cxpl\codes; \intrinsic 'code' declarations char A1, A2; int A3, I; [A1:= "abc";
A2:= "ABC"; A3:= [1,2,3]; for I:= 0 to 2 do [ChOut(0, A1(I)); ChOut(0, A2(I)); IntOut(0, A3(I)); CrLf(0); ];
]</lang>
- Output:
aA1 bB2 cC3
Z80 Assembly
<lang z80>org &1000
ld b,3
ld ix,array1
loop:
ld a,(ix)
call &bb5a ;prints character to screen
ld a,(ix+3)
call &bb5a
ld a,(ix+6)
call &bb5a
ld a,13
call &bb5a
ld a,10
call &bb5a
inc ix
djnz loop
ret array1:
db "abc"
array2:
db "ABC"
array3:
db "123"</lang>
- Output:
Ready call &1000 aA1 bB2 cC3 Ready
The program was only written to display the first 3 characters. If they were of different lengths, the wrong characters (or random bytes outside the program) would have been read, depending on which array "ran out" first. The three arrays defined with the db
directive are stored sequentially.
zkl
<lang zkl>foreach a,b,c in (["a".."c"].zip(T("A","B","C"),[1..])){ println(a,b,c) }</lang> or <lang zkl>Utils.zipWith(False,fcn{vm.arglist.concat().println()},
["a".."c"],T("A","B","C"),[1..])</lang>
- Output:
aA1 bB2 cC3
zip[With] stops at the end of the shortest sequence, which means it can work with infinite sequences
ZX Spectrum Basic
<lang zxbasic>10 LET sza = 3: REM size of a 20 LET szb = 3: REM size of b 30 LET szc = 3: REM size of c 40 DIM a$(sza): DIM b$(szb): DIM c$(szc) 50 LET max = sza: REM assume a is the biggest 60 IF szb > max THEN LET max = szb: REM now try b 70 IF szc > max THEN LET max = szc: REM or c 80 REM populate our arrays, and as a bonus we already have our demo loop 90 REM we might as well print as we populate showing the arrays in columns 100 FOR l = 1 TO max 110 IF l <= sza THEN READ a$(l): PRINT a$(l); 120 IF l <= szb THEN READ b$(l): PRINT b$(l); 130 IF l <= szc THEN READ c$(l): PRINT c$(l); 140 PRINT: REM newline 150 NEXT l 150 PRINT "The arrays are shown in columns." 160 PRINT "A$ runs down the left hand side," 170 PRINT "and C$ runs down the right." 180 STOP 200 DATA "a","b","c","A","B","C","1","2","3"</lang>
Simplification
<lang zxbasic>10 READ size: DIM a$(size): DIM b$(size): DIM c$(size) 20 FOR i=1 TO size 30 READ a$(i),b$(i),c$(i) 40 PRINT a$(i);b$(i);c$(i) 50 NEXT i 60 DATA 3,"a","A","1","b","B","2","c","C","3"</lang>
- Programming Tasks
- Iteration
- 11l
- 360 Assembly
- 8080 Assembly
- 8086 Assembly
- ACL2
- Action!
- Ada
- ALGOL 68
- ALGOL W
- APL
- AppleScript
- Arturo
- AutoHotkey
- AWK
- Axe
- Babel
- BaCon
- BASIC256
- BBC BASIC
- Beads
- Befunge
- C
- C sharp
- C++
- Chapel
- Clojure
- COBOL
- Common Lisp
- D
- Delphi
- DWScript
- E
- EchoLisp
- Efene
- Eiffel
- Ela
- Elena
- Elixir
- Erlang
- Euphoria
- F Sharp
- Factor
- Fantom
- Fermat
- Forth
- Fortran
- FreeBASIC
- Frink
- FunL
- Gambas
- GAP
- Go
- Golfscript
- Groovy
- Harbour
- Haskell
- Haxe
- HicEst
- Icon
- Unicon
- Icon Programming Library
- J
- Java
- JavaScript
- Jq
- Jsish
- Julia
- K
- Kotlin
- Lambdatalk
- LFE
- Liberty BASIC
- Lisaac
- LiveCode
- Logo
- Lua
- Maple
- Mathematica
- Wolfram Language
- Mercury
- Modula-3
- MUMPS
- Nanoquery
- Nemerle
- NetRexx
- NewLISP
- Nim
- NS-HUBASIC
- Oberon-2
- Objeck
- OCaml
- Oforth
- OoRexx
- Oz
- PARI/GP
- Pascal
- Perl
- Phix
- Phix/basics
- Phixmonti
- PHP
- PicoLisp
- Pike
- PL/I
- PostScript
- Initlib
- PowerBASIC
- PowerShell
- Prolog
- PureBasic
- Python
- Quackery
- R
- Racket
- Raku
- Red
- REXX
- Ring
- Ruby
- Run BASIC
- Rust
- Salmon
- Sather
- Scala
- Scheme
- Sidef
- Smalltalk
- Standard ML
- Stata
- SuperCollider
- Swift
- Tailspin
- Tcl
- TorqueScript
- TUSCRIPT
- TXR
- UNIX Shell
- C Shell
- Ursa
- Ursala
- Vala
- VBA
- VBScript
- Visual Basic .NET
- Visual FoxPro
- Wart
- Wren
- X86 Assembly
- XBasic
- XPL0
- Z80 Assembly
- Zkl
- ZX Spectrum Basic