Integer sequence: Difference between revisions

 

=={{header|0815}}==

 

=={{header|11l}}==

 

=={{header|360 Assembly}}==

For maximum compatibility, this program uses only the basic instruction set (S/360).

INTSEQED CSECT

USING INTSEQED,12

DW DS 0D,PL8 pack dec 15num

EM12 DC X'402020202020202020202120' mask CL12 11num

{{out}}

<pre>

...

</pre>

 

=={{header|8080 Assembly}}==

Actually printing the numbers out would depend on the hardware and operating system.

MVI A, 0 ; move immediate

LOOP: INR A ; increment

JMP LOOP ; jump unconditionally

 

 

A more complex, arbitrary precision version that can count as high as you have free bytes of memory to use. (This does assemble with CP/M's MAC assembler, but since it doesn't implement PRBUFR, it's only useful for exposition purposes, or for loading into DDT.)

 

ORG 0100H

BITS EQU 128 ; 128 bits of precision

BUFR: ; This space will hold our number

; We zero this memory before the loop

 

=={{header|Action!}}==

CARD i

 

UNTIL i=0

OD

{{out}}

[https://gitlab.com/amarok8bit/action-rosetta-code/-/raw/master/images/Integer_sequence.png Screenshot from Atari 8-bit computer]

 

=={{header|Ada}}==

procedure Integers is

Value : Integer := 1;

Value := Value + 1; -- raises exception Constraint_Error on overflow

end loop;

Alternative (iterating through all values of Positive (positive part of Integer) without endless loop):

procedure Positives is

begin

Ada.Text_IO.Put_Line (Positive'Image (Value));

end loop;

 

=={{header|ALGOL 68}}==

{{wont work with|ELLA ALGOL 68|Any (with appropriate job cards) - tested with release [http://sourceforge.net/projects/algol68/files/algol68toc/algol68toc-1.8.8d/algol68toc-1.8-8d.fc9.i386.rpm/download 1.8-8d] - due to extensive use of '''format'''[ted] ''transput''.}}

The upper limit of the loop variable ''i'' is ''max int'' currently ''+2147483647'' for [[ALGOL 68G]].

(

FOR i DO

printf(($g(0)","$,i))

OD

Partial output:

<pre>

 

=={{header|ALGOL W}}==

% print the integers from 1 onwards %

% Algol W only has 32-bit integers. When i reaches 2^32, %

i := i + 1

end loop_forever ;

 

=={{header|Applesoft BASIC}}==

Integer variables can be within the range of -32767 to 32767.

20 PRINT I%;

30 I% = I% + 1

40 PRINT ", ";

Last screen of scrolled output:

51, 32652, 32653, 32654, 32655, 32656, 3

2657, 32658, 32659, 32660, 32661, 32662,

, 32766, 32767

?ILLEGAL QUANTITY ERROR IN 30

 

=={{header|ARM Assembly}}==

.global main

 

@ call to 'PRINT' routine

add r0, r0, #1 @ increment R0

 

=={{header|ArnoldC}}==

HEY CHRISTMAS TREE n

YOU SET US UP @NO PROBLEMO

ENOUGH TALK

CHILL

 

=={{header|Arturo}}==

while ø [

print i

inc 'i

 

=={{header|AutoHotkey}}==

This uses traytip to show the results. A msgbox, tooltip, or fileappend could also be used.

Loop

 

=={{header|AWK}}==

for( i=0; i != i + 1; i++ )

print( i )

 

Awk uses floating-point numbers. This loop terminates when <code>i</code> becomes too large for integer precision. With IEEE doubles, this loop terminates when <code>i</code> reaches <code>2 ^ 53</code>.

Integers in Axe are limited to 16 bits, or a maximum of 65535. This script will run infinitely until either the variable overflows or a key is pressed.

 

End

0→I

Disp I▶Dec,i

I++

 

=={{header|BASIC}}==

{{works with|ZX Spectrum Basic}}

20 LET A = A + 1

30 PRINT A

{{works with|QBasic}}

 

=={{header|BASIC256}}==

 

do

print i

i += 1

 

 

=={{header|Batch File}}==

Variables are limited to 32bit integer, capable of a maximum value of <code>2,147,483,647</code>

@echo off

set number=0

echo %number%

goto loop

{{out}}

<pre>

{{works with|BBC BASIC for Windows}}

Native version, limited to 53-bit integers (maximum output 9007199254740992):

REPEAT

i += 1

PRINT TAB(0,0) i;

Version using Huge Integer Math and Encryption library (up to 2^31 bits, but this program limited to 65535 decimal digits because of maximum string length):

PROC_himeinit("")

reg% = 1

SYS `hi_Incr`, ^reg%, ^reg%

PRINT TAB(0,0) FN_higetdec(reg%);

 

=={{header|bc}}==

 

=={{header|beeswax}}==

Using an ordinary loop structure:

 

Using a jump instruction:

 

Numbers in beeswax are unsigned 64-bit integers, so after reaching 2^64-1 the counter wraps around to 0.

Also note that the range of values written to the code page or 'playfield' is often much smaller - frequently only supporting 8 bits, sometimes signed, sometimes unsigned.

 

 

=={{header|BQN}}==

 

While the input is lesser than or equal to infinity, print, then increment.

 

=={{header|Bracmat}}==

{{trans|Ruby}}

Bracmat uses big numbers. Numbers are stored with a radix 10, each decimal digit occupying one byte. When multiplying or dividing, numbers are temporarily converted to radix 10000 (32-bit systems: 1 digit occupies two bytes) or radix 100000000 (64-bit systems: 1 digit occupies four bytes) to speed up the computation.

 

=={{header|Brainf***}}==

This program assumes that decrementing past zero wraps around, but it doesn't rely on cell size, other than that a cell can hold at least six bits. It also assumes the ASCII character set. This is an arbitrarily large number implementation.

-------------------[>>-<++++++++++<[+>-<]]>[-<+>]<++++++++++++++++++

++++++++++++++++++++++++++++++>]<[<]>>[-<+++++++++++++++++++++++++++

 

This modification of the previous program will print out 1 to the maximum cell value, still assuming wrapping. On many implementations, this will print out 1-255.

--------------------[>>-<++++++++++<[+>-<]]>[-<+>]<+++++++++++++++++

+++++++++++++++++++++++++++++++>]<[<]>>[-<++++++++++++++++++++++++++

 

This program can count in any base counting system under 256. '''Note:''' Change the characters in quotes equal to the base counting system you want to use.

 

=={{header|Brat}}==

 

loop {

p i

i = i + 1

 

=={{header|Burlesque}}==

 

1R@

 

=={{header|C}}==

Prints from 1 to max unsigned integer (usually 2**32 -1), then stops.

 

int main()

 

return 0;

 

==={{libheader|GMP}}===

 

int main()

 

return 0;

 

==={{libheader|OpenSSL}}===

OpenSSL provides arbitrarily large integers.

 

#include <openssl/err.h> /* ERR_*() */

#include <stdio.h> /* fprintf(), puts() */

}

/* NOTREACHED */

 

=={{header|C sharp|C#}}==

using System.Numerics;

 

}

}

 

=={{header|C++}}==

#include <iostream>

#include <limits>

while (i < std::numeric_limits<decltype(i)>::max())

std::cout << ++i << '\n';

<!--

 

#include <iostream>

// Do nothing

}

-->

 

=={{header|ChucK}}==

Math.INT_MAX is a constant value that represents the greater integer, 32 bit , 64 bit systems.

for(1 => int i; i < Math.INT_MAX; i ++)

{

<<< i >>>;

}

 

=={{header|Clean}}==

In Clean this example has a limit of basically 2147483648.

 

import StdEnv

 

 

Output:

 

=={{header|Clojure}}==

 

=={{header|CLU}}==

% overflows. It is a signed machine-sized integer; so 64 bits on

% a modern machine. After that it will raise an exception.

stream$putl(po, int$unparse(i))

end

 

=={{header|COBOL}}==

PROGRAM-ID. Int-Sequence.

 

 

GOBACK

 

=={{header|CoffeeScript}}==

Like with most languages, counting is straightforward in CoffeeScript, so the program below tries to handle very large numbers. See the comments for starting the sequence from 1.

 

# This very limited BCD-based collection of functions

# makes it easy to count very large numbers. All arrays

console.log BcdInteger.render big_int

big_int = BcdInteger.succ big_int

 

output

> coffee foo.coffee | head -5

199999999999999999999999999999999999999999999999999999

200000000000000000000000000000000000000000000000000002

200000000000000000000000000000000000000000000000000003

 

=={{header|Common Lisp}}==

 

 

If your compiler does tail call elimination (note: this has absolutely no advantage over normal loops):

 

=={{header|Component Pascal}}==

BlackBox Component Builder

MODULE IntegerSequence;

IMPORT StdLog;

 

END IntegerSequence.

Execute: ^Q IntegerSequence.Do<br/>

Output:

=={{header|Computer/zero Assembly}}==

This program counts up to 255 in the accumulator, after which it starts again from zero.

JMP start

 

=={{header|Cowgol}}==

This program will count up to 2^32-1, and then stop.

 

 

var n: uint32 := 1;

print_nl();

n := n + 1;

 

The following program will keep going until it runs out of memory, using one byte per digit.

 

 

sub print_back(s: [uint8]) is

infnum := incr(infnum);

print_back(infnum);

 

=={{header|Crystal}}==

Will run as long as enough memory to represent numbers.

 

 

=={{header|D}}==

 

void main() {

while (true)

writeln(++i);

Alternative:

 

void integerSequence(T)() if (isIntegral!T || is(T == BigInt)) {

default: writeln("\nBye bye!"); break;

}

 

=={{header|Dc}}==

 

=={{header|DCL}}==

$ loop:

$ write sys$output i

$ i = i + 1

{{out}}

<pre>1

 

=={{header|Delphi}}==

 

{$APPTYPE CONSOLE}

for i := 1 to High(i) do

WriteLn(i);

 

=={{header|DWScript}}==

High(i) returns the maximum supported value, typically, it is the highest signed 64 bit integer.

var i: Integer;

 

for i:=1 to High(i) do

PrintLn(i);

 

=={{header|Dyalect}}==

 

while true {

n += 1

print(n)

 

=={{header|Déjà Vu}}==

 

while /= -- dup dup:

++

 

 

This continues to print numbers until double precision IEEE 754 cannot represent adjacent integers any more (9007199254740992, to be exact).

=={{header|E}}==

 

 

=={{header|EchoLisp}}==

(lib 'bigint) ;; arbitrary length integers

(for ((n (in-naturals))) (writeln n))

 

=={{header|EDSAC order code}}==

================

P0D [ constant: 1 ]

 

=={{header|Eiffel}}==

class

APPLICATION

number:INTEGER_64

end

 

=={{header|Elena}}==

ELENA 4.x :

public program()

i += 1u

}

 

=={{header|Elixir}}==

 

=={{header|Emacs Lisp}}==

Displays in the message area interactively, or to standard output under <code>-batch</code>.

 

 

=={{header|Erlang}}==

 

 

=={{header|ERRE}}==

.............

A%=0

END LOOP

.............

% is integer-type specificator. Integer type works on 16-bit signed numbers (reserved constant MAXINT is 32767). Beyond this limit execution will give Runtime error #6 (overflow).

 

=={{header|Euphoria}}==

i = 0

while 1 do

? i

i += 1

 

=={{header|F_Sharp|F#}}==

 

let rec integers i =

seq { yield i

yield! integers (i+1) }

 

 

lazy sequence of int32 starting from 0

 

lazy sequence of int32 starting from n

 

lazy sequence (not necessarily of int32) starting from n (using unfold anamorphism)

<div>

> numbers 0 |> Seq.take 10;;

 

=={{header|Factor}}==

 

=={{header|Fantom}}==

 

class Main

{

}

}

 

Fantom's integers are 64-bit signed, and so the numbers will return to 0 and continue again, if you wait long enough!

 

=={{header|Fermat}}==

 

=={{header|Fish}}==

Since there aren't really libraries in Fish and I wouldn't know how to program arbitarily large integers, so here's an example that just goes on until the interpreter's number limit:

 

=={{header|Forth}}==

 

=={{header|Fortran}}==

{{works with|Fortran|90 and later}}

implicit none

n = n + 1

end do

 

=={{header|FreeBASIC}}==

 

' FB does not natively support arbitrarily large integers though support can be added

Loop Until i = 0 ' will wrap back to 0 when it reaches 4,294,967,296

 

 

=={{header|Frink}}==

All of Frink's numbers can be arbitrarily-sized:

i=0

while true

i = i + 1

}

 

=={{header|FunL}}==

The following has no limit since FunL has arbitrary size integers.

 

 

=={{header|Futhark}}==

accepts an input indicating how many integers to generate. It encodes the size of the returned array in its type.

 

fun main(n: int): [n]int = iota n

 

=={{header|GAP}}==

local n;

n := 1;

 

# Prepare some coffee

 

=={{header|Go}}==

Size of <tt>int</tt> type is implementation dependent. After the maximum positive value, it rolls over to maximum negative, without error. Type <tt>uint</tt> will roll over to zero.

 

import "fmt"

fmt.Println(i)

}

The <tt>big.Int</tt> type does not roll over and is limited only by available memory, or practically, by whatever external factor halts CPU execution: human operator, lightning storm, CPU fan failure, heat death of universe, etc.

 

import (

fmt.Println(i)

}

 

=={{header|Gridscript}}==

#INTEGER SEQUENCE.

 

(9,1):INCREMENT

(11,1):GOTO 0

 

=={{header|Groovy}}==

for (def i = 1; i > 0; i++) { println i }

 

 

// Arbitrarily-long binary signed integer (BigInteger)

 

=={{header|GUISS}}==

Graphical User Interface Support Script makes use of installed programs. There are no variables, no loop structures and no jumps within the language so iteration is achieved by repetative instructions. In this example, we will just use the desktop calculator and keep adding one to get a counter. We stop after counting to ten in this example.

 

Button:[plus],Button:1,Button:[equals],Button:[plus],Button:1,Button:[equals],

Button:[plus],Button:1,Button:[equals],Button:[plus],Button:1,Button:[equals],

Button:[plus],Button:1,Button:[equals],Button:[plus],Button:1,Button:[equals],

Button:[plus],Button:1,Button:[equals],Button:[plus],Button:1,Button:[equals],

 

=={{header|GW-BASIC}}==

20 PRINT A#

30 A#=A#+1

 

=={{header|Haskell}}==

 

Or less imperatively:

 

 

=={{header|HolyC}}==

Prints from 1 to max unsigned 64 bit integer (2**64 -1), then stops.

while (++i) Print("%d\n", i);

 

=={{header|Icon}} and {{header|Unicon}}==

Icon and Unicon support large integers by default. The built-in generator seq(i,j) yields the infinite sequence i, i+j, i+2*j, etc. Converting the results to strings for display will likely eat your lunch before the sequence will take its toll.

 

every write(seq(1)) # the most concise way

 

=={{header|IS-BASIC}}==

110 PRINT I;

 

INF = 9.999999999E62

The following will count indefinitely but once the 32-bit (or 64-bit depending on J engine version) limit is reached, the results will be reported as floating point values (which would immediately halt on 64 bit J and halt with the 53 bit precision limit is exceeded on 32 bit J). Since that could take many, many centuries, even on a 32 bit machine, more likely problems include the user dying of old age and failing to pay the electric bill resulting in the machine being powered off.

 

 

The above works with both fixed sized integers and floating point numbers (fixed sized integers are automatically promoted to floating point, if they overflow), but also works with extended precision integers (which will not overflow, unless they get so large that they cannot be represented in memory, but that should exceed lifetime of the universe, let alone lifetime of the computer).

 

This adds support for extended precision (in that it converts non-extended precision arguments to extended precision arguments) and will display integers to ∞ (or at least until the machine is turned off or interrupted or crashes).

 

=={{header|Java}}==

Long limit:

for(long i = 1; ;i++) System.out.println(i);

}

"Forever":

 

for(BigInteger i = BigInteger.ONE; ;i = i.add(BigInteger.ONE)) System.out.println(i);

}

 

=={{header|JavaScript}}==

This code is accurate up to 2^53 where it will be stuck an 2^53 because a IEEE 64-bit double can not represent 2^53 + 1.

 

while (true)

This example uses a BigInt[https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/BigInt] literal to support arbitrary large integers.

 

while (true)

 

=={{header|Joy}}==

 

 

=={{header|jq}}==

 

 

 

Integers can of course also be represented by strings of decimal digits, and if this representation is satisfactory, a stream of consecutive integers thus represented can be generated using the same technique as is employed on the

 

=={{header|Julia}}==

while true

println(i += 1)

The built-in <code>BigInt</code> type is an arbitrary precision integer (based on the GMP library), so the value of <code>i</code> is limited only by available memory. To use (much faster) hardware fixed-width integer types, use e.g. <code>zero(Int32)</code> or <code>zero(Int64)</code>. (Initializing <code>i = 0</code> will use fixed-width integers that are the same size as the hardware address width, e.g. 64-bit on a 64-bit machine.)

 

=={{header|K}}==

 

Using a <code>while</code> loop:

 

 

=={{header|Kotlin}}==

 

// version 1.0.5-2

n += BigInteger.ONE

}

 

=={{header|Lambdatalk}}==

The long_add primitive allow counting beyond the javascript numbers limits, depending on the system memory.

{def infinite_set

{lambda {:i}

{infinite_set 0}

-> 0 1 2 3 ... forever

 

=={{header|Lang5}}==

 

=={{header|Lasso}}==

while(#number > 0) => {^

#number++

' '

//#number > 100 ? #number = -2 // uncomment this row if you want to halt the run after proving concept

This will run until you exhaust the system resources it's run under.

 

=={{header|Liberty BASIC}}==

Liberty BASIC handles extremely large integers. The following code was halted by user at 10,000,000 in test run.

i=i+1

locate 1,1

scan

wend

 

=={{header|Limbo}}==

The int (32 bits) and big (64 bits) types are both signed, so they wrap around. This version uses the infinite precision integer library:

 

 

include "sys.m"; sys: Sys;

}

}

 

=={{header|Lingo}}==

repeat while i>0

put i

i = i+1

 

Lingo uses signed 32 bit integers, so max. supported integer value is 2147483647:

 

Beyond this limit values behave like negative numbers:

-- -2147483648

put the maxInteger+2

Up to the (quite high) number where floats (double-precission) start rounding, floats can be used to exceed the integer limit:

 

put float(the maxInteger)+1

-- 2

-- 3

 

=={{header|LLVM}}==

{{trans|C}}

; LLVM does not provide a way to print values, so the alternative would be

; to just load the string into memory, and that would be boring.

}

 

 

=={{header|Lua}}==

i = 1

 

i = i + 1

end

 

=={{header|M2000 Interpreter}}==

\\ easy way

a=1@

\\ this flag reset to false before restart.

{loop : Print a : a++}

 

=={{header|Maple}}==

Maple has arbitrary-precision integers so there are no built-in limits on the size of the integers represented.

 

print(n)

 

=={{header|Mathematica}} / {{header|Wolfram Language}}==

Built in arbitrary precision support means the following will not overflow.

x = 1;

Monitor[While[True, x++], x]

 

=={{header|MATLAB}} / {{header|Octave}}==

 

 

Typically, numbers are stored as double precision floating point numbers, giving accurate integer values up to about 2^53=bitmax('double')=9.0072e+15. Above this limit, round off errors occur. This limitation can be overcome by defining the numeric value as type int64 or uint64

 

 

This will run up to 2^64 and then stop increasing, there will be no overflow.

Matlab and Octave recommend vectorizing the code, one might pre-allocate the sequence up to a specific N.

 

 

The main limitation is the available memory on your machine. The standard version of Octave has a limit that a single data structure can hold at most 2^31 elements. In order to overcome this limit, Octave must be compiled with "./configure --enable-64", but this is currently not well supported.

 

=={{header|Maxima}}==

 

=={{header|min}}==

{{works with|min|0.19.3}}

min's integers are 64-bit signed. This will eventually overflow.

 

=={{header|МК-61/52}}==

 

=={{header|ML/I}}==

"" Integer sequence

"" Will overflow when it reaches implementation-defined signed integer limit

MCGO L1

>

 

=={{header|Modula-2}}==

FROM FormatString IMPORT FormatString;

FROM Terminal IMPORT WriteString,ReadChar;

END;

ReadChar

 

=={{header|Nanoquery}}==

All native integers in Nanoquery can become arbitrarily large by default, so this program would run until it ran out of memory.

while true

println i

i += 1

 

=={{header|Necromantus}}==

In Necromantus integer size is limited by the java's int.

let i = 0;

while true

i = i + 1;

}

 

=={{header|NetRexx}}==

===Rexx Built In===

NetRexx provides built-in support for very large precision arithmetic via the <tt>Rexx</tt> class.

options replace format comments java crossref symbols binary

 

say k_

end k_

 

===Using BigInteger===

Java's <tt>BigInteger</tt> class is also available for very large precision arithmetic.

options replace format comments java crossref symbols binary

 

say k_.toString(int radix)

end

 

=={{header|NewLISP}}==

 

=={{header|Nim}}==

while true:

inc i

 

Using BigInts:

 

var i = 0.initBigInt

while true:

i += 1

 

=={{header|Oberon-2}}==

Works with oo2c Version 2

MODULE IntegerSeq;

IMPORT

 

END IntegerSeq.

 

=={{header|Objeck}}==

bundle Default {

class Count {

}

}

 

=={{header|OCaml}}==

with an imperative style:

let i = ref 0 in

while true do

print_newline ();

incr i;

 

with a functional style:

let rec aux i =

print_int i;

aux (succ i)

in

 

=={{header|Oforth}}==

The loop will stop when out of memory

 

 

=={{header|Ol}}==

Ol does not limit the size of numbers. So maximal number depends only on available system memory.

(let loop ((n 1))

(print n)

(loop (+ 1 n)))

 

Sample sequence with break for large numbers:

(let loop ((n 2))

(print n)

(unless (> n 100000000000000000000000000000000)

(loop (* n n))))

Output:

<pre>

OpenEdge has three data types that can be used for this task:

<ol><li>INTEGER (32-bit signed integer)

 

DO WHILE TRUE:

ii = ii + 1.

DISPLAY ii.

When an integer rolls over its maximum of 2147483647 error 15747 is raised (Value # too large to fit in INTEGER.).

</li>

<li>INT64 (64-bit signed integer)

 

DO WHILE TRUE:

ii = ii + 1.

DISPLAY ii.

When a 64-bit integer overflows no error is raised and the signed integer becomes negative.

</li>

<li>DECIMAL (50 digits)

 

DO WHILE TRUE:

de = de + 1.

DISPLAY de.

When a decimal requires mores than 50 digits error 536 is raised (Decimal number is too large.).

</li>

=={{header|Order}}==

Order supports arbitrarily-large positive integers natively. However, the simple version:

 

#define ORDER_PP_DEF_8printloop ORDER_PP_FN( \

8printloop(8inc(8N)))) )

 

... while technically fulfilling the task, will probably never display anything, as most C Preprocessor implementations won't print their output until the file is done processing. Since the C Preprocessor is not technically Turing-complete, the Order interpreter has a maximum number of steps it can execute - but this number is very, very large (from the documentation: "the Order interpreter could easily be extended with a couple of hundred macros to prolong the wait well beyond the estimated lifetime of the sun"), so the compiler is rather more likely to simply run out of memory.

 

To actually see anything with GCC, add a maximum limit so that the task can complete:

 

#define ORDER_PP_DEF_8printloop ORDER_PP_FN( \

8when(8less(8N, 99), 8printloop(8inc(8N))))) )

 

 

=={{header|PARI/GP}}==

 

=={{header|Pascal}}==

{{works with|Free_Pascal}}

Quad word has the largest positive range of all ordinal types

var

Number: QWord = 0; // 8 bytes, unsigned: 0 .. 18446744073709551615

inc(Number);

until false;

{{libheader|GMP}}

With the gmp library your patience is probably the limit :-)

 

uses

mpz_add_ui(Number, Number, 1); //* increase Number *//

until false;

 

=={{header|Perl}}==

 

On 64-bit Perls this will get to <tt>2^64-1</tt> then print <tt>1.84467440737096e+19</tt> forever. On 32-bit Perls using standard doubles this will get to <tt>999999999999999</tt> then start incrementing and printing floats until they lose precision. This behavior can be changed by adding something like:

which makes almost all integers large (ranges are excluded). Faster alternatives exist with non-core modules, e.g.

* <tt>use bigint lib=>"GMP";</tt>

=={{header|Phix}}==

This will crash at 1,073,741,824 on 32 bit, or 4,611,686,018,427,387,904 on 64-bit, and as indicated best not to try this or any below under pwa/p2js:

<span style="color: #008080;">without</span> <span style="color: #008080;">javascript_semantics</span>

<span style="color: #004080;">integer</span> <span style="color: #000000;">i</span> <span style="color: #0000FF;">=</span> <span style="color: #000000;">0</span>

<span style="color: #000000;">i</span> <span style="color: #0000FF;">+=</span> <span style="color: #000000;">1</span>

<span style="color: #008080;">end</span> <span style="color: #008080;">while</span>

This will stall at 9,007,199,254,740,992 on 32-bit, and about twice the above on 64-bit.

(after ~15 or 19 digits of precision, adding 1 will simply cease to have any effect)

<span style="color: #008080;">without</span> <span style="color: #008080;">javascript_semantics</span>

<span style="color: #004080;">atom</span> <span style="color: #000000;">a</span> <span style="color: #0000FF;">=</span> <span style="color: #000000;">0</span>

<span style="color: #000000;">a</span> <span style="color: #0000FF;">+=</span> <span style="color: #000000;">1</span>

<span style="color: #008080;">end</span> <span style="color: #008080;">while</span>

{{libheader|Phix/mpfr}}

This will probably carry on until the number has over 300 million digits (32-bit, you can

square that on 64-bit) which would probably take zillions of times longer than the

universe has already existed, if your hardware/OS/power grid kept going that long.

<span style="color: #008080;">without</span> <span style="color: #008080;">javascript_semantics</span>

<span style="color: #008080;">include</span> <span style="color: #004080;">mpfr</span><span style="color: #0000FF;">.</span><span style="color: #000000;">e</span>

<span style="color: #7060A8;">mpfr_printf</span><span style="color: #0000FF;">(</span><span style="color: #000000;">1</span><span style="color: #0000FF;">,</span><span style="color: #008000;">"%Zd\n"</span><span style="color: #0000FF;">,</span><span style="color: #000000;">b</span><span style="color: #0000FF;">)</span>

<span style="color: #008080;">end</span> <span style="color: #008080;">while</span>

Lastly, a gui version you can run online [http://phix.x10.mx/p2js/Integers.htm here].

{{libheader|Phix/pGUI}}

{{libheader|Phix/online}}

<span style="color: #008080;">with</span> <span style="color: #008080;">javascript_semantics</span>

<span style="color: #008080;">include</span> <span style="color: #000000;">pGUI</span><span style="color: #0000FF;">.</span><span style="color: #000000;">e</span>

<span style="color: #7060A8;">IupClose</span><span style="color: #0000FF;">()</span>

<span style="color: #008080;">end</span> <span style="color: #008080;">if</span>

 

=={{header|PicoLisp}}==

 

=={{header|Piet}}==

 

=={{header|Pike}}==

while(true)

 

=={{header|PILOT}}==

*InfiniteLoop

T :#n

C :n = n + 1

 

=={{header|PL/I}}==

infinity: procedure options (main);

declare k fixed decimal (30);

((k do k = 1 to 999999999999999999999999999998))(f(31));

end infinity;

 

=={{header|PL/M}}==

will print numbers until the <code>ADDRESS</code> variable overflows.

 

 

/* CP/M CALL AND NUMBER OUTPUT ROUTINE */

 

CALL BDOS(0,0);

 

To get around this limitation, the following program stores the number as an array of digits.

On a 64K CP/M system it will keep going until it has over 50.000 digits.

 

 

/* CP/M CALL */

END;

 

 

 

=={{header|Plain English}}==

Numbers are signed 32-bit values, so this will overflow somewhere in the neighborhood of 2.1 billion.

Start up.

Put 1 into a number.

Bump the number.

Repeat.

 

=={{header|PostScript}}==

{{libheader|initlib}}

1 {succ dup =} loop

 

=={{header|PowerShell}}==

{

for ([int]$i = 0;;$i++)

}

}

 

=={{header|Prolog}}==

writeln(I),

I1 is I+1,

loop(I1).

 

===Constraint Handling Rules===

Works with SWI-Prolog and library '''CHR''' written by '''Tom Schrijvers''' and '''Jan Wielemaker'''

 

:- chr_constraint loop/1.

 

loop(N) <=> writeln(N), N1 is N+1, loop(N1).

 

=={{header|PureBasic}}==

Repeat

a.q+1

PrintN(Str(a))

 

=={{header|Python}}==

while i:

print(i)

 

Or, alternatively:

 

for i in count():

 

Pythons integers are of arbitrary large precision and so programs would probably keep going until OS or hardware system failure.

 

=={{header|Q}}==

 

Using converge (the <tt>\</tt> adverb):

 

Using <tt>while</tt>:

 

=={{header|Quackery}}==

Quackery uses bignums.

 

 

=={{header|R}}==

repeat {

print(z)

z <- z + 1

 

=={{header|Racket}}==

Racket uses bignums, so counting should continue up to very large numbers. Naturally, printing these numbers will consume quite a bit of power.

 

(for ([i (in-naturals)]) (displayln i))

 

=={{header|Raku}}==

(formerly Perl 6)

 

=={{header|Raven}}==

Raven uses signed 32 bit integer values.

repeat TRUE while

 

=={{header|Red}}==

 

i: 1

print i

i: i + 1

 

=={{header|Retro}}==

Retro uses signed integer values.

 

 

=={{header|REXX}}==

/*keep counting. According to some pundits in-the-know, one version of */

/*the big-bang theory is that the universe will collapse back to where */

 

/*It only took Deep Thought 7.5 million years to come up with the */

 

=={{header|Ring}}==

size = 10

 

n = n + 1

end

 

=={{header|Ruby}}==

 

The step method of Numeric takes two optional arguments. The limit defaults to infinity, the step size to 1.

Ruby 2.6 introduced open-ended ranges:

 

 

=={{header|Run BASIC}}==

i = i + 1

print i

Eventually as it gets larger it becomes a floating point.

 

=={{header|Rust}}==

{{works with|Rust 1.2}}

for i in 0.. {

println!("{}", i);

}

 

 

Looping endlessly:

 

use num::bigint::BigUint;

i = i + BigUint::one();

}

 

=={{header|Salmon}}==

Salmon has built-in unlimited-precision integer arithmetic, so these examples will all continue printing decimal values indefinitely, limited only by the amount of memory available (it requires O(log(n)) bits to store an integer n, so if your computer has 1 GB of memory, it will count to a number with on the order of <math>2^{80}</math> digits).

 

 

or

 

 

or

 

while (true)

{

i!;

++i;

 

=={{header|Scala}}==

 

=={{header|Scheme}}==

 

(let loop ((i 1))

(display i) (newline)

(loop (+ 1 i)))

 

Scheme does not limit the size of numbers.

 

=={{header|Seed7}}==

Limit 2147483647:

 

const proc: main is func

writeln(number);

until number = 2147483647;

"Forever":

include "bigint.s7i";

 

incr(number);

until FALSE;

 

=={{header|Sidef}}==

No limit:

 

=={{header|Smalltalk}}==

[

Stdout print:i; cr.

i := i + 1

will run forever.

 

=={{header|SSEM}}==

Since we have no Add instruction, we subtract -1 on each iteration instead of adding 1. The same -1 also serves as a jump target, taking advantage of a quirk of the SSEM architecture (the Current Instruction counter is incremented after the instruction has been executed, not before—so <tt>GOTO address</tt> has to be coded as <tt>GOTO address - 1</tt>).

01000000000000000000000000000000 1. 2 to CI goto -1 + 1

 

=={{header|Standard ML}}==

IntInf.int (arbitrary precision).

 

fun printInts(n) =

(

in

printInts(1)

 

{{out}}

=={{header|SuperCollider}}==

The SuperCollider language has a 32-bit signed int, and a 64 bit signed float. Instead of locking the interpreter with an infinite loop, we post the values over time.

i = Routine { inf.do { |i| i.yield } }; // return all integers, represented by a 64 bit signed float.

fork { inf.do { i.next.postln; 0.01.wait } }; // this prints them incrementally

 

A shorter form of the first line above, using list comprehensions:

i = {:i, i<-(0..) };

 

=={{header|Swift}}==

while true {

println(i++)

 

=={{header|Symsyn}}==

| The following code will run forever

| Symsyn uses a 64 bit signed integer

+ x

go lp

 

=={{header|Tcl}}==

 

=={{header|Tiny BASIC}}==

REM will overflow after 32767

LET N = 0

LET N = N + 1

GOTO 10

 

 

=={{header|True BASIC}}==

 

DO

LOOP

 

 

 

=={{header|TUSCRIPT}}==

LOOP n=0,999999999

n=n+1

 

=={{header|UNIX Shell}}==

 

num=0

while true; do

echo $num

num=`expr $num + 1`

 

=={{header|Ursa}}==

# integer sequence

#

out i endl console

inc i

 

=={{header|Vala}}==

uint i = 0;

while (++i < uint.MAX)

stdout.printf("%u\n", i);

 

 

=={{header|Verilog}}==

integer i;

 

$finish ;

end

 

=={{header|Visual Basic .NET}}==

Note that attempting to store any value larger than the maximum value of any given type (say 2 147 483 648 for an Integer) will result in an OverflowException being thrown (<i>"Arithmetic operation resulted in an overflow."</i>)

 

Console.WriteLine(i)

 

===Arbitrarily large numbers===

One could use the '''System.Numerics''' library as the C# example did, or one can do the following.<br/>A list of Long Integers is maintained as the incremented number. As the incremented value approaches the maximum allowed (''base'') in the first element of ''ar'', a new item is inserted at the beginning of the list to extend the incremented number. The process has the limitation of when the ''ar'' array is enlarged to the point where the program exhausts the available memory, it ought to indicate failure and terminate. It is my understanding that a '''List''' count is backed by an '''Integer.MaxValue''' limitation and there may also be a 2 GB per object limitation involved. Since writing to the Console is such a slow process, I lack the patience to wait for the program (as written) to fail. If the program is tweaked to fail early, the practical limit seems to be a number 2,415,919,086 digits in length.

 

Module Module1

TimeStamp("keypress")

End Sub

{{out}}

<pre>1

 

=={{header|WDTE}}==

 

s.new 0 (+ 1)

-> s.map (io.writeln io.stdout)

-> s.drain

 

WDTE's number type is, at the time of writing, backed by Go's <code>float64</code> type, so all of the same limitations that apply there apply here. Also, this should '''not''' be run in the WDTE playground, as it will run with no output until the browser crashes or is killed.

 

Also, the ''System.print'' method in the standard library will only display a maximum of 14 digits before switching to scientific notation. To get around this one can use instead the ''Fmt.print'' method of the ''Wren-fmt'' module which displays integers 'normally' up to the maximum and also caters for BigInts as well.

import "./big" for BigInt

 

Fmt.print("$i", bi)

bi = bi + 1

 

=={{header|XLISP}}==

(print x)

(integer-sequence-from (+ x 1)) )

 

 

=={{header|XPL0}}==

code CrLf=9, IntOut=11;

int N;

N:= N+1;

until N<0;

 

 

=={{header|Yabasic}}==

 

repeat

i = i + 1

until i = 0

 

 

===16-Bit===

The Amstrad CPC's screen isn't big enough to show it all at once, but here you go. This prints numbers out (in hexadecimal) from <tt>0x0001</tt> to <tt>0xFFFF</tt>.

PrintChar equ &BB5A

ld hl,1 ;START AT ONE

adc a,&40

jp PrintChar

 

===Arbitrarily Large Integers===

This version displays an ever-increasing 64-bit unsigned integer. Unlike the previous version, this one continues forever and underflows to 0 after it reaches <tt>0xFFFFFFFFFFFFFFFF</tt>. This logic can be extended to integers of up to 128 bytes in size (since <code>IX+#</code> uses a signed offset, you'd need some way to alter the pointer to memory if you wanted even larger numbers than that, it's possible but a bit cumbersome. Not that this method wasn't cumbersome to begin with.)

PrintChar equ &BB5A

ld ix,NumberRam

 

NumberRam: ;a 64-bit value, stored little-endian

 

=={{header|zkl}}==

m:=(1).MAX; [1..m].pump(Console.println) // (1).MAX is 9223372036854775807