Loop structures: Difference between revisions

From Rosetta Code
Content added Content deleted
m (Spelling/grammar/aesthetics)
Line 91: Line 91:
)
)


A non-obvious feature of the ALGOL 68 FOR loop is that the stripped
A non-obvious feature of the ALGOL 68 FOR loop is that the stripped DO ~ OD clause is valid. Moreover, the initialization and increment parts may be empty. That is, the following is an infinite loop:
DO ~ OD clause is valid. Moreover, the initialization and increment parts may be
empty. That is, the following is an infinite loop:


main: (
main: (
Line 101: Line 99:
)
)


A loop like this can still be left with <tt>GO TO</tt> (see below) or
A loop like this can still be left with <tt>GO TO</tt> (see below) or <tt>EXIT</tt> (which returns from the function, and therefore also ends the loop).
<tt>EXIT</tt> (which returns from the function, and therefore also ends
the loop).


This next loop has a loop variable "i" that starts at 1.
This next loop has a loop variable "i" that starts at 1.
Line 137: Line 133:


===WHILE with "break" equivalent===
===WHILE with "break" equivalent===
The break statement allows you to stop a loop. In ALGOL 68 uses GO TOs, hence
The break statement allows you to stop a loop. In ALGOL 68 uses GO TOs, hence can break from any level. You can also do this with do-WHILE and FOR.
can break from any level. You can also do this with do-WHILE and FOR.


main:(
main:(
Line 178: Line 173:
** only the ''<u>do</u> ~ <u>od</u>'' portion was compulsory, in which case the loop will iterate indefinitely.
** only the ''<u>do</u> ~ <u>od</u>'' portion was compulsory, in which case the loop will iterate indefinitely.
** thus the clause ''<u>to</u> 100 <u>do</u> ~ <u>od</u>'', will iterate only 100 times.
** thus the clause ''<u>to</u> 100 <u>do</u> ~ <u>od</u>'', will iterate only 100 times.
** the <u>while</u> "syntatic element" allowed a programmer to break from a <u>for</u> loop early. eg
** the <u>while</u> "syntactic element" allowed a programmer to break from a <u>for</u> loop early. eg
<u>int</u> sum sq:=0;
<u>int</u> sum sq:=0;
<u>for</u> i
<u>for</u> i
Line 188: Line 183:
<u>od</u>
<u>od</u>


Subsequent "extensions" to the standard Algol68 allowed the <u>to</u> syntatic element to be replaced with <u>upto</u> and <u>downto</u> to achieve a small optimisation. The same compilers also incorporated:
Subsequent "extensions" to the standard Algol68 allowed the <u>to</u> syntactic element to be replaced with <u>upto</u> and <u>downto</u> to achieve a small optimisation. The same compilers also incorporated:
* <u>until</u><sup>(C)</sup> - for late loop termination.
* <u>until</u><sup>(C)</sup> - for late loop termination.
* <u>foreach</u><sup>(S)</sup> - for working on arrays in [[Parallel computing|parallel]].
* <u>foreach</u><sup>(S)</sup> - for working on arrays in [[Parallel computing|parallel]].
Line 312: Line 307:
A loop like this can still be left with <tt>break</tt> (see below) or <tt>return</tt> (which returns from the function, and therefore also ends the loop).
A loop like this can still be left with <tt>break</tt> (see below) or <tt>return</tt> (which returns from the function, and therefore also ends the loop).


Another difference compared to <tt>while</tt> is the behavior of <tt>continue</tt>. In a for loop, continue will also execute the third clause (the increment part).
Another difference compared to <tt>while</tt> is the behavior of <tt>continue</tt>. In a for loop, continue will also execute the third clause (the increment part).


===while with continue===
===while with continue===
The continue statement allows you to continue execution
The continue statement allows you to continue execution at the beginning of the loop, skipping the rest of the loop.
In C you can only do this with the most inner loop. You can also do this with do-while and for.
at the beginning of the loop, skipping the rest of the loop.
In C you can only do this with the most inner loop.
You can also do this with do-while and for.


int main (int argc, char ** argv) {
int main (int argc, char ** argv) {
Line 335: Line 328:


===while with break===
===while with break===
The break statement allows you to stop a loop.
The break statement allows you to stop a loop. In C you can only break from most inner loop. You can also do this with do-while and for.
In C you can only break from most inner loop.
You can also do this with do-while and for.

int main (int argc, char ** argv) {
int main (int argc, char ** argv) {
int condition = 1;
int condition = 1;
Line 393: Line 383:
=={{header|Factor}}==
=={{header|Factor}}==
===Looping===
===Looping===
Most looping is done with recursion. Tail recursion is properly optimized.
Most looping is done with recursion. Tail recursion is properly optimized.
: forever ( quot -- ) dup slip forever ;
: forever ( quot -- ) dup slip forever ;
[ "A hungry raptor stalks you..." print flush 2000 random sleep ] forever
[ "A hungry raptor stalks you..." print flush 2000 random sleep ] forever
Line 446: Line 436:


===Mixed Structures===
===Mixed Structures===
Because Forth's compiler is laid bare to the programmer, it is quite easy to both define your own looping structures or combine existing structures in interesting ways. The rules for such combining are somewhat involved, though discussions can be found in the gforth user's manual, among other places. These more complex looping constructs can make up for Forth's lack of a "break" word, and can allow expressing complex loops without resorting to boolean variables. A practical example is also found in the [[Binary search]] task.
Because Forth's compiler is laid bare to the programmer, it is quite easy to both define your own looping structures or combine existing structures in interesting ways. The rules for such combining are somewhat involved, though discussions can be found in the gforth user's manual, among other places. These more complex looping constructs can make up for Forth's lack of a "break" word, and can allow expressing complex loops without resorting to boolean variables. A practical example is also found in the [[Binary search]] task.


A good example of a useful combination is this complex loop:
A good example of a useful combination is this complex loop:
Line 549: Line 539:
=={{header|Haskell}}==
=={{header|Haskell}}==


Most of the usual applications for loops are realized in Haskell by operations on (lazy) lists, like '''map''', '''fold''' or '''filter'''. It's unusual to use loops in an imperative style.
Most of the usual applications for loops are realized in Haskell by operations on (lazy) lists, like '''map''', '''fold''' or '''filter'''. It's unusual to use loops in an imperative style. However, if one insists on it, it's easy to make your own implementation of any loop variant.
However, if one insists on it, it's easy to make your own implemenation of any loop variant.


Here are a few examples:
Here are a few examples:
Line 1,014: Line 1,003:
for x y in [a b c], [1 2 3] do [^x ^y] => endfor;
for x y in [a b c], [1 2 3] do [^x ^y] => endfor;


in first iteration sets x to "a" and y to 1, in the second x is "b"
in first iteration sets x to "a" and y to 1, in the second x is "b" and y is 2, in the third (last) iteration x is "c" and y is 3. The iteration ends when the shortest list is exhausted.
and y is 2, in the third (last) iteration x is "c" and y is 3.
The iteration ends when the shortest list is exhausted.


Somtimes one wants to process tails of the list, to do this use
Sometimes one wants to process tails of the list, to do this use on keyword instead of in keyword:
on keyword instead of in keyword:


for x on [a b c] do x => endfor;
for x on [a b c] do x => endfor;
Line 1,046: Line 1,032:


=== quitloop quitif quitunless ===
=== quitloop quitif quitunless ===
Inside loops one can use control transfers to prematuraly exit the
Inside loops one can use control transfers to prematurely exit the loop or end current iteration and start the next one:
loop or end current iteration and start the next one:


while true do n - 1 -> n; quitif(n=0); endwhile;
while true do n - 1 -> n; quitif(n=0); endwhile;
Line 1,055: Line 1,040:


=== nextloop nextif nextunless ===
=== nextloop nextif nextunless ===
Similarely to quitloop nextloop unconditionally ends current iteration
Similarely to quitloop nextloop unconditionally ends current iteration and starts the new one, nextif(x) ends current iteration when x is true, nextunless(x) is equivalent to nextif(not(x)). The loop control transfers can be also used inside for (and until) loops.
and starts the new one, nextif(x) ends current iteration when x is true,
nextunless(x) is equivalent to nextif(not(x)). The loop control transfers
can be also used inside for (and until) loops.


Finally, it is frequently possible to avoid explicit iteration
Finally, it is frequently possible to avoid explicit iteration using higher order map functions (like appdata and mapdata).
using higher order map functions (like appdata and mapdata).


=={{header|PostScript}}==
=={{header|PostScript}}==
Line 1,218: Line 1,199:


=={{header|Seed7}}==
=={{header|Seed7}}==
In Seed7 new statements can be declared easily.
In Seed7 new statements can be declared easily. This feature is not explained here. Here are examples of the predefined loops:
This feature is not explained here.
Here are examples of the predefined loops:


===while===
===while===

Revision as of 16:25, 18 January 2008

Task
Loop structures
You are encouraged to solve this task according to the task description, using any language you may know.
Control Structures

These are examples of control structures. You may also be interested in:


In this task, we document loop structures offered by different languages.

Ada

Simple Loop

The simple loop in Ada produces an infinite loop.

loop
   -- do something
end loop;

Any loop can be terminated by the exit reserved word.

loop
   -- do something
   if some-condition then
      exit;
   end if;
   -- do somthing else
end loop;

The short-hand notation for expressing the exit condition is:

loop
   -- do something
   exit when some-conditon;
   -- do something else
end loop;

While loop

while some-condition loop
   -- do something
end loop;

For loop

The for loop in Ada iterates over a specified range of discrete values. Ranges are an important concept in Ada. Ranges are always expressed in the form of lowest..highest. The loop control variable always takes on the type of the specified range. The loop control variable is read-only within the loop, and has a scope only within the loop.

Iteration over a literal range:

for I in 1..10 loop
   Put(I);
end loop;

You can also employ a range defined earlier in your code.

subtype Month_Number is Integer range 1..12;

for I in Month_Number loop
   Put(I);
end loop;

Ranges are also defined for the indices for arrays. Ada array indices can begin at any value, not just 0 or 1.

type Balanced_Index is range -10..10;
type My_Array_Type is array(Balanced_Index) of Integer;
My_Array : My_Array_Type;

for I in My_Array'Range loop
   Put(My_Array(I));
end loop;

ALGOL 68

WHILE

main: (
  BOOL condition := TRUE;

  WHILE condition DO
    # Do something #
    # Don't forget to change the value of condition. #
    # If it remains nonzero, we'll have an infinite loop. #
  OD
)

do-WHILE

main: (
  BOOL condition := ~;
 
  WHILE
    # Do something #
    # The difference with the first loop is that the #
    # code in the loop will be executed at least once, #
    # even IF the condition is 0 at the beginning, #
    # because it is only checked at the end. #
    # Don't forget to change the value of condition. #
    # If it remains nonzero, we'll have an infinite loop. #
  condition DO SKIP OD
)

FOR

main: (
 
  FOR i FROM 1 TO 10 BY DO
    # The code here will be performed 10 times. #
    # The first part in the FOR-statement i FROM 1 is the initialization, #
    # and is executed once before the loop begins. #
    # The second part is the end condition TO 10, which is checked #
    # every time the loop is started, also the first time; #
    # the loop ends IF it is false. #
    # The third part BY 1 is performed every time the code in the loop #
    # is at the end, just before the end condition is checked. #
  OD
)

A non-obvious feature of the ALGOL 68 FOR loop is that the stripped DO ~ OD clause is valid. Moreover, the initialization and increment parts may be empty. That is, the following is an infinite loop:

main: (
  DO # forever #
    SKIP
  OD
)

A loop like this can still be left with GO TO (see below) or EXIT (which returns from the function, and therefore also ends the loop).

This next loop has a loop variable "i" that starts at 1.

main: (
  FOR I DO # forever #
    SKIP
  OD
)

The next loop iterates only a maximum of 100 times:

main: (
  TO 100 DO # 100 times only #
    SKIP
  OD
)

WHILE with "continue" equivalent

main:(
  BOOL condition := TRUE;

  WHILE condition DO
    # Do something #

    IF other condition THEN
      SKIP # do nothing, hence start another iteration #
    ELSE
    # Do something else #
    # This part is not executed IF other condition was true #
    FI 
  OD
)

WHILE with "break" equivalent

The break statement allows you to stop a loop. In ALGOL 68 uses GO TOs, hence can break from any level. You can also do this with do-WHILE and FOR.

main:(
  BOOL condition := TRUE;

  WHILE condition DO
    # Do something #

    IF other condition THEN
      GO TO break x # Continue after loop #
    FI 

    # Do something else #
    # This part is not executed IF other condition was true #
  OD;
  break x: SKIP
)

A more domestic version of a break allows for only one condition to exit a loop:

main:(
  BOOL break condition := FALSE;
  WHILE 
    # Do something #

    NOT break condition DO

    # Do something else #
    # This part is not executed IF other condition was true #

  OD
)

The ALGOL 68 "universal" for/whileloop

 [ for index ] [ from first ] [ by increment ] [ to last ] [ while condition ] do statements od
 The minimum form of a "loop clause" is thus: do statements od

This was considered the "universal" loop, the full syntax is:

for i from 1 by 1 to 3 while i≠4 do ~ od

There are several unusual aspects of the construct:

    • only the do ~ od portion was compulsory, in which case the loop will iterate indefinitely.
    • thus the clause to 100 do ~ od, will iterate only 100 times.
    • the while "syntactic element" allowed a programmer to break from a for loop early. eg
int sum sq:=0;
for i
while 
  print(("So far:",i,newline));
  sum sq≤1000 
do 
  sum sq+:=i↑2
od

Subsequent "extensions" to the standard Algol68 allowed the to syntactic element to be replaced with upto and downto to achieve a small optimisation. The same compilers also incorporated:

  • until(C) - for late loop termination.
  • foreach(S) - for working on arrays in parallel.

AppleScript

repeat-until

set i to 5
repeat until i is less than 0
       set i to i - 1
end repeat
repeat
       --endless loop
end repeat

repeat-with

       repeat with i from 1 to 20
               --do something
       end repeat
       set array to {1,2,3,4,5}
       repeat with i in array
               display dialog i
       end repeat

BASIC

Compiler: QuickBasic 4.5

for

By 1's:

FOR I = 1 TO 10
'...
NEXT I

By X's (any number you want):

FOR I = 1 TO 10 STEP X
'...
NEXT I

while

DO WHILE(SomeNumber = SomeOtherNumber)
'...
LOOP

do-while

DO
'...
LOOP WHILE(SomeNumber = SomeOtherNumber)

Note: The sign for "not equal to" in BASIC is not !=, but rather <> .

Befunge

Befunge does not have specific loop statements, but any type of loop construct can be created using the branch (_|) skip (#) and direction (<>v^) commands. For example, the following is one possible structure for a while loop:

enter > test v
      ^ body _ exit

This program is like a for-loop in other languages, printing the digits from nine down to zero.

9>:.:#v_@
 ^ -1 <

Brainf***

BF's only control flow construct of any kind is a loop. Two of the eight commands define the start and end of a conditional loop.

  • [ - branch forward past matching ']' if the current cell is zero
  • ] - branch back to the matching '[' if the current cell is non-zero

In practice this is equivalent to a "while not zero" loop in other languages.

[-]  set current cell to zero
[->+>+<<]>>[-<<+>>] copy cell 0 to cell 1, using cell 2 as temporary storage

C

while

Compiler: gcc 4.1.2

int main (int argc, char ** argv) {
  int condition = 1;

  while ( condition ) {
    // Do something
    // Don't forget to change the value of condition.
    // If it remains nonzero, we'll have an infinite loop.
  }
}

do-while

int main (int argc, char ** argv) {
  int condition = ...;
 
  do {
    // Do something
    // The difference with the first loop is that the
    // code in the loop will be executed at least once,
    // even if the condition is 0 at the beginning,
    // because it is only checked at the end.
    // Don't forget to change the value of condition.
    // If it remains nonzero, we'll have an infinite loop.
  } while ( condition );
}

for

int main (int argc, char ** argv) {
  int i;
 
  for (i=0; i<10; ++i) {
    // The code here will be performed 10 times.
    // The first part in the for-statement (i=0) is the initialization,
    // and is executed once before the loop begins.
    // The second part is the end condition (i<10), which is checked
    // every time the loop is started, also the first time;
    // the loop ends if it is false.
    // The third part (++i) is performed every time the code in the loop
    // is at the end, just before the end condition is checked.
  }
}

A non-obvious feature of the C for loop is that an empty condition is equivalent to a true condition. Moreover, the initialization and increment parts may be empty. That is, the following is an infinite loop:

int main()
{
  for (;;) // forever
  {
  }
}

A loop like this can still be left with break (see below) or return (which returns from the function, and therefore also ends the loop).

Another difference compared to while is the behavior of continue. In a for loop, continue will also execute the third clause (the increment part).

while with continue

The continue statement allows you to continue execution at the beginning of the loop, skipping the rest of the loop. In C you can only do this with the most inner loop. You can also do this with do-while and for.

int main (int argc, char ** argv) {
  int condition = 1;

  while ( condition ) {
    // Do something

    if (other condition)
      continue; // Continue at the beginning of the loop

    // Do something else
    // This part is not executed if other condition was true
  }
}

while with break

The break statement allows you to stop a loop. In C you can only break from most inner loop. You can also do this with do-while and for.

int main (int argc, char ** argv) {
  int condition = 1;

  while ( condition ) {
    // Do something

    if (other condition)
      break; // Continue after the the loop

    // Do something else
    // This part is not executed if other condition was true
  }
}

C++

Run-Time Control Structures

See C

Compile-Time Control Structures

A compile-time for loop can be generated with template metaprogramming. Example:

// the loop
template<int start, int finish, template<int n, typename T> class X> struct loop
{
  typedef typename X<start, typename loop<start+1, finish, X>::type>::type type;
};

// the termination of the loop
template<int finish, template<int n, typename T> class X> struct loop<finish, finish, X>
{
  typedef typename X<finish, void>::type type;
};
// example usage: This implements just a very complicated way of building a multi-dimensional array

// the loop body
template<int n, typename T> struct build_array
{
  typedef T type[n];
};

template<int n> struct build_array<n, void>
{
  typedef double type;
};

// the loop execution: This is equivalent to
// typedef double array_type[2][3][4][5];
typedef loop<2,6,build_array>::type array_type;

Delphi

See Pascal

Factor

Looping

Most looping is done with recursion. Tail recursion is properly optimized.

   : forever ( quot -- ) dup slip forever ;
   [ "A hungry raptor stalks you..." print flush 2000 random sleep ] forever

Iteration

Most indices are implicit or not present at all.

   3 [ "pint" drink ] times
   { "high" "level" "language" } [ print ] each
       high
       level
       language
   10 [ sq ] map
       { 0 1 4 9 16 25 36 49 64 81 }
   { 1 2 3 } { 4 5 6 } [ * ] 2map .
       { 4 10 18 }
   10 [ even? ] subset .
       V{ 0 2 4 6 8 }
   0 10 3 <range> >array .
       { 0 3 6 9 }
   10 1 -2 <range> >array .
       { 10 8 6 4 2 }
   2222 [ dup 0 > ] [ 2/ dup ] [ ] unfold nip .
       { 1111 555 277 138 69 34 17 8 4 2 1 0 }

Iterating with an index:

   : indexed-alphabet. ( -- )
       "abcdefghijklmnopqrstuvwxyz" dup length
       [ unparse write ": " write 1string print ] 2each ;

Forth

DO-LOOP

( limit start ) DO ( iterated statements ) LOOP
( limit start ) DO ( iterated statements ) ( increment ) +LOOP
LEAVE \ exits a DO loop
UNLOOP EXIT \ cleans up loop counters from return stack before returning from the current word

example: Two standard iterations

10 0 DO I . LOOP      \ Prints the numbers from 0 to 9
10 0 DO I . 2 +LOOP   \ Prints the even numbers from 0 to 8

BEGIN-UNTIL

BEGIN ( iterated statements ) ( conditional ) UNTIL

example: Counts down from a given number to zero

: COUNTDOWN ( n -- )  BEGIN  DUP CR .  1- DUP 0< UNTIL  DROP ;

BEGIN-AGAIN

BEGIN ( iterated statements ) AGAIN

example: echo user's input

: FOREVER ( -- )   BEGIN  KEY EMIT  AGAIN ;

BEGIN-WHILE-REPEAT

BEGIN  ( conditional ) WHILE  ( iterated statements )  REPEAT

example: counts down from a given number to one

: COUNTDOWN ( n -- )  BEGIN  DUP WHILE  CR DUP . 1-  REPEAT  DROP ;

Additional WHILE clauses may be added to a loop, but each extra WHILE requires a matching THEN after the REPEAT.

Mixed Structures

Because Forth's compiler is laid bare to the programmer, it is quite easy to both define your own looping structures or combine existing structures in interesting ways. The rules for such combining are somewhat involved, though discussions can be found in the gforth user's manual, among other places. These more complex looping constructs can make up for Forth's lack of a "break" word, and can allow expressing complex loops without resorting to boolean variables. A practical example is also found in the Binary search task.

A good example of a useful combination is this complex loop:

BEGIN
  ( condition 1 )
WHILE
  ( condition 2 )
UNTIL
  ( condition 2 succeeded )
ELSE
  ( condition 1 failed )
THEN

An example of using this idiom in practice might be this pseudo-Forth

BEGIN
  read-next-record
WHILE
  found-record
UNTIL
  process-record
ELSE
  error" Ran out of records looking for the right one!"
THEN

Groovy

While Loops

while (true) {
  println 'groovy'
}

For Loops


 // iterate over a range
 x = 0
 for (i in 1..3) { x += i }
 assert x == 6

 // iterate over a list
 x = 0
 for (i in [1, 2, 3]) { x += i }
 assert x == 6

 // iterate over an array
 x = 0
 for (i in (1..3).toArray()) { x += i }
 assert x == 6

 // iterate over a map's key/value pairs
 x = 0
 for (i in map) { x += i.value }
 assert x = 6

 // iterate over a map's values
 x = 0
 for (i in map.values()) { x += i }
 assert x == 6

 // iterate over the characters in a string
 text = 'abc'
 list = []
 for (c in text) { list.add(c) }
 assert list == ['a', 'b', 'c']

Each

 def stringList = [ "java", "perl", "python", "ruby" ];

 def stringMap = [ "Su" : "Sunday", "Mo" : "Monday", "Tu" : "Tuesday",
                  "We" : "Wednesday", "Th" : "Thursday", "Fr" : "Friday",
                  "Sa" : "Saturday" ];

 stringList.each() { print " ${it}" }; println "";
 // java perl python ruby

 stringMap.each() { key, value -> println "${key} == ${value}" };
 // Su == Sunday
 // We == Wednesday
 // Mo == Monday
 // Sa == Saturday
 // Th == Thursday
 // Tu == Tuesday
 // Fr == Friday  

 stringList.eachWithIndex() { obj, i -> println " ${i}: ${obj}" };
 // 0: java
 // 1: perl
 // 2: python
 // 3: ruby

 stringMap.eachWithIndex() { obj, i -> println " ${i}: ${obj}" };
 // 0: Su=Sunday
 // 1: We=Wednesday
 // 2: Mo=Monday
 // 3: Sa=Saturday
 // 4: Th=Thursday
 // 5: Tu=Tuesday
 // 6: Fr=Friday

Haskell

Most of the usual applications for loops are realized in Haskell by operations on (lazy) lists, like map, fold or filter. It's unusual to use loops in an imperative style. However, if one insists on it, it's easy to make your own implementation of any loop variant.

Here are a few examples:

Pre-checked loop (while)

whileM :: Monad m => m Bool -> m a -> m ()
whileM cond body = 
  cond >>= \b -> if b then body >> untilM cond body else return ()

Post-checked loop (repeat-until)

untilM :: Monad m => m Bool -> m a -> m ()
untilM cond body = 
  body >> cond >>= \b -> if b then return () else body >> untilM cond body

For-style loop

Simplest done by iterating over a list:

forM :: Monad m => [a] -> (a -> m ()) -> m ()
forM []     f = return ()
forM (x:xs) f = f x >> forM xs f

IDL

It should be noted that IDL programmers tend to avoid loops -- most of the time loops are used to access the elements of arrays or vectors, and since IDL is an array language the same purpose can almost always be served in a faster, more elegant and more readable way though any of the array operations.

for

 for i=0,50,2 do print,i

prints out every second number starting at 0 stopping at 50. Wherever a single command can go in IDL, there can always go a begin...end pair with arbitrary amount of code in between. Thus the above can also read

 for variable = start, stop [,increment] do begin
   ;some code here
   ;some more code
 endfor

It is allowed but not required to use the appropriate "type of end" - i.e. it would be allowed to just say "end" instead of "endfor". However "endfor" (and "endwhile", "endif" etc) will throw an error if the wrong one is encountered at compile time and thus it is recommended to always use the more descriptive form since it makes debugging a lot easier.

while

 while condition do command

Same extensions as above:

 while running do begin
   ; various snippets of code that might change the variable 
   ; "running" from something true to something false
 end[while]

repeat

 repeat command until condition

etc

goto

'Goto' exists and can in principle be forced to make a loop:

 label: 
   ;some code
 [if condition then $]
 goto label

break

The break statement will immediately terminate the current innermost for, while, repeat, if, case or switch without having to resort to a goto.

Java

while

   while(booleanConditionOrVariable)
   {
       foo();
   }

do-while

   do
   {
       foo();
   }
   while (booleanConditionOrVariable)

for

   for(int i = 0; i < 5; i++)
   {
       foo();
   }

foreach

Platform: J2SE 1.5.0+

 Object[] objects;
 // ...
 for (Object current : objects[]) {
   // ...
 }
 int[] numbers;
 // ...
 for (int i : numbers) {
   // ...
 }
 Collection<Type> collect;
 // ...
 for(Type i : collect){
   // ...
 }

while with continue

while(booleanConditionOrVariable){
   //[code]
   if(anotherBooleanConditionOrVariable){
      continue;
      //"continue" skips this code and goes through the loop again
   }
   //"continue" skips this code and goes through the loop again
}

This also works in a for loop, but the conditional and iteration code (second and third arguments) will be run before the loop begins again.

while with break

while(booleanConditionOrVariable){
   //[code]
   if(anotherBooleanConditionOrVariable){
      break;
      //"break" skips this code and leaves the loop
   }
   //"break" skips this code and leaves the loop
}

This also works in a for loop, and no code from the arguments of the loop will be run again.

JavaScript

while

   while(true) {
       foo();
   }

do while

   do {
       foo();
   } while(test);

for

   for(var i = 0; i < 5; i++) {
       foo();
   }

for in

//iterate through property names of an object

var obj = {prop1:"a", prop2:"b", prop3:"c"};
for (var key in obj)
  alert( key + ' is set to ' + obj[key] );

for each in

//iterate through property values of an object

var obj = {prop1:"a", prop2:"b", prop3:"c"};
for each(var element in obj)
  alert( element );

LSE64

LSE's loop words all work via tail recursion, like Scheme, by re-executing the current word. If used interactively, "repeat" works upon the current line. Exiting words, like "&&", "||", and "exit" can be used to exit an otherwise infinite loop (see other tasks).

infinite : " again and " ,t repeat
limited : continue? &repeat

Counted loops execute a specified word n times. Within that word, "count" accesses a loop counter, counted down to zero.

body : count , sp
10 body iterate    # 10 9 8 7 6 5 4 3 2 1 

body? : count , sp  count 5 >
10 body? &iterate   # 10 9 8 7 6 5

newLISP

dotimes

Interpreter: newLISP v.9.0

(dotimes (x 10) (println (+ x 1)))

do-until

Interpreter: newLISP v.9.0

(set 'x 1)
(do-until (= x 10) (inc 'x) (println x))

do-while

Interpreter: newLISP v.9.0

(set 'x 1)
(do-while (< x 10) (inc 'x) (println x))

for

Interpreter: newLISP v.9.0

(for (x 1 10) (println x))

Object Pascal

See Pascal

Objective-C

See C

Pascal

while

Compiler: Turbo Pascal 7.0

 WHILE condition1 DO
   BEGIN
     procedure1;
     procedure2;
   END;

repeat-until

Compiler: Turbo Pascal 7.0

 REPEAT
   procedure1;
   procedure2;
 UNTIL condition1;

for

Compiler: Turbo Pascal 7.0

 FOR counter := 1 TO 10 DO
   BEGIN
     procedure1;
     procedure2;
   END;

Compiler: Delphi 1.x and above

 Var
  Counter : char ;
 FOR counter := 'A' TO 'Z' DO
   BEGIN
     procedure1;
     procedure2;
   END;

Perl

while

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my $condition1 = 0;

while ( $condition1 ) {
 # Do something.
 # Remember to change the value of condition1 at some point.
}

do-while

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my $condition1 = 0;

do {
 # Do something.
 # Remember to change the value of condition1 at some point.
} while ( $condition1 );

until

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my $condition1 = 1;

until ( $condition1 ) {
 # Do something.
 # Remember to change the value of condition1 at some point.
}


do-until

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my $condition1 = 1;

do {
 # Do something.
 # Remember to change the value of condition1 at some point.
} until ( $condition1 );

for

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my $limit = 5;

for ( my $iterator = 0; $iterator < $limit; $iterator++ ) {
  # Do something
}

# for-variant, implicit iteration
for (0..$limit) {
  # Do something
}

do_something() for 0..$limit;

foreach

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my @numbers = (1, 2, 3);
my %names = (first => "George", last => "Jetson");

foreach my $number (@numbers) {
  # Do something with $number
}

foreach my $key (keys %names) {
  # Do something with $key (values are accessible as %names{$key} )
}

map

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my @numbers = (1, 2, 3);
my @target;

@target = map {
  # Do something with $_
} @numbers;

@target = map($_ + 1, @numbers);

sub a_sub {
  # Do something with $_
}

@target = map a_sub @numbers;

grep

Interpreter: Perl 5.8.8

#!/usr/bin/perl -w

use strict;

my @people = qw/Bobbie Charlie Susan/;
my @target;

@target = grep {
  # Discriminate based on $_
} @people;

# Feed grep into map, this picks out elements 1, 3, 5, etc.
@target = map($people[$_], grep($_ & 1, 0..$#people));

# Pick out the diminutive names
@target = grep(/ie$/, @people);

sub a_sub {
  # Do something with $_, and return a true or false value
}

@target = grep a_sub @people;

PHP

while

while(ok()) {
    foo();
    bar();
    baz();
}

do-while

$i = 0;
do {
    echo $i;
} while ($i > 0);

for

for($i = 0; $i < 10; ++$i) {
    echo $i;
}

foreach

foreach(range(0, 9) as $i) {
    echo $i;
}

foreach is only used for arrays, which is not obvious from the above example

foreach($array as $key => $value) {
    echo $key.' is '.$value;
}

Pop11

while

Pop11 offers multiple looping constructs. Basic one is while loop:

 5 -> n;
 while n > 0 do 
     printf(n, '%p\n');
     n - 1 -> n;
 endwhile;

until

Variant of while is until loop:

until condition do /* Action */ enduntil;

is equivalent to:

while mot(condition) do /* Action */ endwhile;

for

One can process all elements of a list:

 for x in [a b c] do x => endfor;

It is possible to simultaneously process multiple lists:

for x y in [a b c], [1 2 3] do [^x ^y] => endfor;

in first iteration sets x to "a" and y to 1, in the second x is "b" and y is 2, in the third (last) iteration x is "c" and y is 3. The iteration ends when the shortest list is exhausted.

Sometimes one wants to process tails of the list, to do this use on keyword instead of in keyword:

for x on [a b c] do x => endfor;

in first iteration sets x to [a b c], in the second to [b c], etc...

There is also "counting" version of for loop:

for x from 2 by 2 to 7 do x => endfor;

goes trough 2, 4 and 6. Ommited by frase means by 1.

There is alse a C-like for loop:

for action1 step  action2 till condition do /* Actions */ endfor;

is equivalent to

action1
while not(condition) do
   /* Actions */
   action2
endwhile;

There are more specialized kinds of loops, but we skip them here.

quitloop quitif quitunless

Inside loops one can use control transfers to prematurely exit the loop or end current iteration and start the next one:

while true do n - 1 -> n; quitif(n=0); endwhile;

quits loop when n=0. quitloop unconditionally quits loop, quitunless(x) is equivalent to quitif(not(x)).

nextloop nextif nextunless

Similarely to quitloop nextloop unconditionally ends current iteration and starts the new one, nextif(x) ends current iteration when x is true, nextunless(x) is equivalent to nextif(not(x)). The loop control transfers can be also used inside for (and until) loops.

Finally, it is frequently possible to avoid explicit iteration using higher order map functions (like appdata and mapdata).

PostScript

The "for" operator expects three numbers and a procedure on the stack. It will consume all four arguments then it will push the first number on the stack, execute the procedure, increase the first number by the second number, push it on the stack, execute the procedure etc until the third number is reached. For example

10 12 200 
  {dup moveto 100 0 lineto} 
  for 
stroke

will add lines to the currentpath that start at coordinates {10,10}; {22,22}; {34,34} ... and all end at {100,0}. The "stroke" operator then renders these lines on the current output device (usually a screen or a piece of paper).

Python

for

Typically for is used to iterate over items in an iteratable. A lot of built-in objects support iteration: list, tuple, str, unicode, dict, set, xrange and file. Any object can support iteration by adding special methods.

for x in ["foo", "bar", "baz"]:
    print x

The built-in range() and/or xrange() functions are used to create indexes. range returns a list while xrange returns an iterator without building the list in memory:

for i in xrange(10):
    print numbers[i]

Optional else suite is executed unless break was called during iteration:

for i in numbers:
    if item < 0:
        break
else:
    print 'all numbers are positive'

Because of the pervasive support for iteration and generators in Python classes and libraries the for loop structure is far more commonly used than while.

list comprehension

Typically used when you want to create a list and there is little logic involved. Faster then for loop:

positives = [n for n in numbers if n > 0]

while

Typical use:

while 1:
    # Do stuff...
    if condition:
        break

You can add optional else, which is executed only if the expression tested was false. Typically not used.

while running:
    # Do stuff...
    if condition:
        running = 0
else:
   print 'stopped'

Since Python has no "bottom-tested" loop construct (such as "do ... until") ... it's common Python practice to either rethink the design in terms of iteration or to use an while 1 (infinite loop) and break out of it as appropriate.

Raven

each

10 each as i                                     # counted loop
    "index is %(i)d\n" print

'hello world' each as c                          # string characters
    "%(c)c" print

[ 'a' 'b' 'c' ] each as item                     # list items
    "%(item)s" print

{ 'a' 1 'b' 2 } each pair as k, v                # hash key/val pairs
    "%(k)s => %(v)d\n" print

'SELECT * FROM table' mysql query each as row    # iterable resource 
    row print

repeat while/until

repeat <some_condition> while
    <some_process>

repeat <come_condition> until
    <some_process>

Ruby

while

while true do
  foo
end

for

for i in 0..4 do
  foo
end

Optionally:

0.upto(4) do #startNumber.upto(endNumber)
  foo
end

or:

5.times do
  foo
end

Step size other than 1:

0.upto(10,2) do #startNumber.upto(endNumber, stepSize)
  foo
end

each

['foo', 'bar', 'baz'].each do |x|
  puts x
end

collect

array = ['foo', 'bar', 'baz'].collect do |x|
  foo x
end

map

array = ['foo', 'bar', 'baz'].map {|x| foo x }

inject

string = ['foo', 'bar', 'baz'].inject("") do |s,x|
  s << x
end
sum = ['foo', 'bar', 'baz'].inject(0) do |s,x|
  s + x.size
end
product = ['foo', 'bar', 'baz'].inject(1) do |p,x|
  p * x.size
end
boolean = ['foo', 'bar', 'baz'].inject(true) do |b,x|
  b &&= x != 'bar'
end

Seed7

In Seed7 new statements can be declared easily. This feature is not explained here. Here are examples of the predefined loops:

while

while TRUE do
  foo;
end while;

repeat

repeat
  foo;
until FALSE;

for

for i range 0 to 4 do
  foo;
end for;
for i range 4 downto 0 do
  foo;
end for;
for stri range []("foo", "bar", "baz") do
  writeln(stri);
end for;

Smalltalk

timesRepeat

 10 timesRepeat: [ expression ].

iterating over a collection

 ( collection ) do: [ expression ].
 ( collection ) collect: [:element | element sendMessageToElement].
 #(1 2 3 4) inject: 5 into: [:sum :number | sum + number]

whileTrue/whileFalse

 x := 0.
 [ x < 100 ]
      whileTrue: [ x := x + 10.].
 [ x = 0 ]
      whileFalse: [ x := x - 20.].

simple looping

 1 to: 5 do: [:number | Transcript show: 'Here is a number: ' number printString ; cr]
 (-50 to: 50 by: 4) do: [:number | Transcript show: number ; cr]

Tcl

foreach

foreach i {foo bar baz} {
    puts "$i"
}

Note that foreach also accepts multiple variables:

foreach {x y} {1 2 3 4} {
    puts "$x,$y"
}

And also multiple lists:

foreach i {1 2 3} j {a b c} {
    puts "$i,$j"
}

Or any combination of variables/list:

foreach i {1 2 3} {x y} {a b c d e f} {
    puts "$i,$x,$y"
}

for

 for {set i 0} {$i < 10} {incr i} {
     puts $i
 }

It bears noting that the three parts of the for loop do not have to consist of "initialize variable", "test value of variable" and "increment variable". This is a common way to think of it as it resembles the "for" loop in other languages, but many other things make sense. For example this for-loop will read a file line-by-line:

 set line ""
 for { set io [open test.txt r] } { ![eof $io] } { gets $io line } {
   if { $line != "" } { ...do something here... }
 }

(This is a somewhat awkward example; just to show what is possible)

while

 set i 0
 while {$i < 10} {
     puts [incr i]
 }

Toka

countedLoop

Counts up or down until the boundaries are met.

10 0 [ i . ] countedLoop
0 10 [ i . ] countedLoop

whileTrue

Repeatedly executes a quote until the quote returns true.

100 [ dup . 1 - dup 0 <> ] whileTrue

whileFalse

Repeatedly executes a quote until the quote returns true.

0 [ dup . 1 + dup 101 = ] whileFalse

UNIX Shell

for

Interpreter:' Bourne Again SHell

A simple loop that counts from 1 to 10.

#!/bin/bash
for ((x = 1; $x <= 10; x = $x + 1 ))
do
 echo $x
done

for...in

Interpreter: Bourne Again SHell

A loop that iterates through all of the (whitespace-delimited) members of ARRAY.

#!/bin/bash
ARRAY="VALUE1 VALUE2 VALUE3 VALUE4 VALUE5"

for ELEMENT in $ARRAY
do
 echo $ELEMENT # Print $ELEMENT
done

while

Interpreter: bourne SHell

#!/bin/sh
x=1
while [ $x -lt 101 ]
do
  echo $x
  x=`expr "$x" + 1`
done


Interpreter: Bourne Again SHell

 #!/bin/bash
 while true
 do
   ...
 done

XSLT

The <xsl:for-each> element allows explicit iteration over a collection of nodes:

<xsl:for-each select="node set expression">
  <!-- executed for each node in the node set. Each node becomes the context for this block. -->
</xsl:for-each>

The XPath node set expression for "select" must return a set of nodes:

PRE                       <!-- all <PRE> blocks in an HTML page -->
group[ignore != 'true']   <!-- all "group" nodes in the current context without an ignore attribute set to 'true' -->

Much other iteration in XSLT is implicit. For example, you can specify a "match" attribute to apply a template to all nodes of a particular type.