Loops/Wrong ranges

From Rosetta Code


Task
Loops/Wrong ranges
You are encouraged to solve this task according to the task description, using any language you may know.

Some languages have syntax or function(s) to generate a range of numeric values from a start value, a stop value, and an increment.

The purpose of this task is to select the range syntax/function that would generate at least two increasing numbers when given a stop value more than the start value and a positive increment of less than half the difference. You are than to use that same syntax/function but with different parameters; and show, here, what would happen.

Use these values if possible:

start stop increment Comment
-2 2 1 Normal
-2 2 0 Zero increment
-2 2 -1 Increments away from stop value
-2 2 10 First increment is beyond stop value
2 -2 1 Start more than stop: positive increment
2 2 1 Start equal stop: positive increment
2 2 -1 Start equal stop: negative increment
2 2 0 Start equal stop: zero increment
0 0 0 Start equal stop equal zero: zero increment
Related tasks



ALGOL W[edit]

Using the Algol W for loop and limiting the sequence to 10 values. The Algol W for loop considers the sign of the increment value when deciding whether to terminate when the loop counter exceeds the stop value (positive increment) or is smaller than the stop value (negative increment).

begin
 % sets the first n elements of s to the sequences of values specified by start, stop and increment  %
 % s( 0 ) is set to the number of elements of s that have been set, in case the sequence ends before n %
procedure sequence ( integer array s ( * )
 ; integer value n, start, stop, increment
) ;
begin
integer sPos;
for j := 0 until n do s( j ) := 0;
sPos  := 1;
for j := start step increment until stop do begin
if sPos > n then goto done;
s( sPos ) := j;
s( 0 ) := s( 0 ) + 1;
sPos  := sPos + 1;
end for_j ;
done:
end sequence ;
 % tests the sequence procedure %
procedure testSequence( integer value start, stop, increment
 ; string(48) value legend
) ;
begin
integer array s ( 0 :: 10 );
sequence( s, 10, start, stop, increment );
s_w := 0; % set output formating %
i_w := 4;
write( legend, ": " );
for i := 1 until s( 0 ) do writeon( s( i ) )
end testSequence ;
 % task trest cases %
testSequence( -2, 2, 1, "Normal" );
testSequence( -2, 2, 0, "Zero increment" );
testSequence( -2, 2, -1, "Increments away from stop value" );
testSequence( -2, 2, 10, "First increment is beyond stop value" );
testSequence( 2, -2, 1, "Start more than stop: positive increment" );
testSequence( 2, 2, 1, "Start equal stop: positive increment" );
testSequence( 2, 2, -1, "Start equal stop: negative increment" );
testSequence( 2, 2, 0, "Start equal stop: zero increment" );
testSequence( 0, 0, 0, "Start equal stop equal zero: zero increment" )
end.
Output:
Normal                                          :   -2  -1   0   1   2
Zero increment                                  :   -2  -2  -2  -2  -2  -2  -2  -2  -2  -2
Increments away from stop value                 :
First increment is beyond stop value            :   -2
Start more than stop: positive increment        :
Start equal stop: positive increment            :    2
Start equal stop: negative increment            :    2
Start equal stop: zero increment                :    2   2   2   2   2   2   2   2   2   2
Start equal stop equal zero: zero increment     :    0   0   0   0   0   0   0   0   0   0

C[edit]

C's 'for' statement appears to fit the bill here and so we use it directly to generate the required ranges of numbers though, as some of the ranges will be infinite, we limit the output to a maximum of 10 numbers.

#include <stdio.h>
 
#define TRUE 1
#define FALSE 0
 
typedef int bool;
 
typedef struct {
int start, stop, incr;
const char *comment;
} S;
 
S examples[9] = {
{-2, 2, 1, "Normal"},
{-2, 2, 0, "Zero increment"},
{-2, 2, -1, "Increments away from stop value"},
{-2, 2, 10, "First increment is beyond stop value"},
{2, -2, 1, "Start more than stop: positive increment"},
{2, 2, 1, "Start equal stop: positive increment"},
{2, 2, -1, "Start equal stop: negative increment"},
{2, 2, 0, "Start equal stop: zero increment"},
{0, 0, 0, "Start equal stop equal zero: zero increment"}
};
 
int main() {
int i, j, c;
bool empty;
S s;
const int limit = 10;
for (i = 0; i < 9; ++i) {
s = examples[i];
printf("%s\n", s.comment);
printf("Range(%d, %d, %d) -> [", s.start, s.stop, s.incr);
empty = TRUE;
for (j = s.start, c = 0; j <= s.stop && c < limit; j += s.incr, ++c) {
printf("%d ", j);
empty = FALSE;
}
if (!empty) printf("\b");
printf("]\n\n");
}
return 0;
}
Output:
Normal
Range(-2, 2, 1) -> [-2 -1 0 1 2]

Zero increment
Range(-2, 2, 0) -> [-2 -2 -2 -2 -2 -2 -2 -2 -2 -2]

Increments away from stop value
Range(-2, 2, -1) -> [-2 -3 -4 -5 -6 -7 -8 -9 -10 -11]

First increment is beyond stop value
Range(-2, 2, 10) -> [-2]

Start more than stop: positive increment
Range(2, -2, 1) -> []

Start equal stop: positive increment
Range(2, 2, 1) -> [2]

Start equal stop: negative increment
Range(2, 2, -1) -> [2 1 0 -1 -2 -3 -4 -5 -6 -7]

Start equal stop: zero increment
Range(2, 2, 0) -> [2 2 2 2 2 2 2 2 2 2]

Start equal stop equal zero: zero increment
Range(0, 0, 0) -> [0 0 0 0 0 0 0 0 0 0]

Factor[edit]

<range> divides by the step value, so a step of 0 causes a divide by zero exception. For the purpose of getting through all the examples, the exceptions are dropped and execution continues, which in general should be avoided.

USING: continuations formatting io kernel math.ranges
prettyprint sequences ;
 
: try-range ( from length step -- )
[ <range> { } like . ]
[ 4drop "Exception: divide by zero." print ] recover ;
 
{
{ -2 2 1 } { 2 2 0 } { -2 2 -1 } { -2 2 10 } { 2 -2 1 }
{ 2 2 1 } { 2 2 -1 } { 2 2 0 } { 0 0 0 }
}
[
first3
[ "%2d %2d %2d <range> => " printf ]
[ try-range ] 3bi
] each
Output:
-2  2  1 <range>  =>  { -2 -1 0 1 2 }
 2  2  0 <range>  =>  Exception: divide by zero.
-2  2 -1 <range>  =>  { }
-2  2 10 <range>  =>  { -2 }
 2 -2  1 <range>  =>  { }
 2  2  1 <range>  =>  { 2 }
 2  2 -1 <range>  =>  { 2 }
 2  2  0 <range>  =>  Exception: divide by zero.
 0  0  0 <range>  =>  Exception: divide by zero.

Go[edit]

Go has only one loop, a 'for' statement, which supports four different syntactical forms commonly found in other C-family languages:

1. A C-like 'for' loop with initialization, condition and increment sections.

2. The 'while' loop functionality (condition only)

3. Infinite loop, equivalent to for(;;) (all sections omitted)

4. Looping over a range of values, similar to foreach etc. (using 'range' keyword).

It appears that either #1 or #4 fits the requirements of this task so I've written a function which generates the appropriate sequence using #1 (limited to a maximum of 10 elements as some sequences will be infinite). I've then applied #4 to the resulting sequence. All sequences include the stop value if it's actually reached.

package main
 
import "fmt"
 
type S struct {
start, stop, incr int
comment string
}
 
var examples = []S{
{-2, 2, 1, "Normal"},
{-2, 2, 0, "Zero increment"},
{-2, 2, -1, "Increments away from stop value"},
{-2, 2, 10, "First increment is beyond stop value"},
{2, -2, 1, "Start more than stop: positive increment"},
{2, 2, 1, "Start equal stop: positive increment"},
{2, 2, -1, "Start equal stop: negative increment"},
{2, 2, 0, "Start equal stop: zero increment"},
{0, 0, 0, "Start equal stop equal zero: zero increment"},
}
 
func sequence(s S, limit int) []int {
var seq []int
for i, c := s.start, 0; i <= s.stop && c < limit; i, c = i+s.incr, c+1 {
seq = append(seq, i)
}
return seq
}
 
func main() {
const limit = 10
for _, ex := range examples {
fmt.Println(ex.comment)
fmt.Printf("Range(%d, %d, %d) -> ", ex.start, ex.stop, ex.incr)
fmt.Println(sequence(ex, limit))
fmt.Println()
}
}
Output:
Normal
Range(-2, 2, 1) -> [-2 -1 0 1 2]

Zero increment
Range(-2, 2, 0) -> [-2 -2 -2 -2 -2 -2 -2 -2 -2 -2]

Increments away from stop value
Range(-2, 2, -1) -> [-2 -3 -4 -5 -6 -7 -8 -9 -10 -11]

First increment is beyond stop value
Range(-2, 2, 10) -> [-2]

Start more than stop: positive increment
Range(2, -2, 1) -> []

Start equal stop: positive increment
Range(2, 2, 1) -> [2]

Start equal stop: negative increment
Range(2, 2, -1) -> [2 1 0 -1 -2 -3 -4 -5 -6 -7]

Start equal stop: zero increment
Range(2, 2, 0) -> [2 2 2 2 2 2 2 2 2 2]

Start equal stop equal zero: zero increment
Range(0, 0, 0) -> [0 0 0 0 0 0 0 0 0 0]

Huginn[edit]

Huginn has the Range generator in Algorithms package. Instantiation of an a priori invalid range is a fatal error.

import Algorithms as algo;
 
class Example {
_start = none;
_stop = none;
_step = none;
_comment = none;
}
 
main() {
examples = [
Example( -2, 2, 1, "Normal" ),
Example( 2, 2, 0, "Start equal stop: zero increment" ),
Example( 0, 0, 0, "Start equal stop equal zero: zero increment" ),
Example( 2, 2, 1, "Start equal stop: positive increment" ),
Example( 2, 2, -1, "Start equal stop: negative increment" ),
Example( -2, 2, 10, "First increment is beyond stop value" ),
Example( -2, 2, 0, "Zero increment, stop greater than start" ),
Example( -2, 2, -1, "Increments away from stop value" ),
Example( 2, -2, 1, "Start more than stop: positive increment" )
];
for ( ex : examples ) {
print(
"{}\nRange( {}, {}, {} ) -> ".format(
ex._comment, ex._start, ex._stop, ex._step
)
);
r = algo.range( ex._start, ex._stop, ex._step );
print(
"{}\n\n".format(
algo.materialize( algo.slice( r, 22 ), list )
)
);
}
}
Output:
Normal
Range( -2, 2, 1 ) -> [-2, -1, 0, 1]

Start equal stop: zero increment
Range( 2, 2, 0 ) -> []

Start equal stop equal zero: zero increment
Range( 0, 0, 0 ) -> []

Start equal stop: positive increment
Range( 2, 2, 1 ) -> []

Start equal stop: negative increment
Range( 2, 2, -1 ) -> []

First increment is beyond stop value
Range( -2, 2, 10 ) -> [-2]

Zero increment, stop greater than start
Range( -2, 2, 0 ) -> ./range.hgn:32:17: Invalid range.
Exit 3

Kotlin[edit]

Although Kotlin's 'for' statement can deal with a range of integers, the increment must be positive and so it cannot be used for this task. We therefore use instead a 'while' statement to generate the same sequence as a C language 'for' statement would (limited to a maximum of 10 elements as some sequences will be infinite) and wrap it in a function.

// Version 1.2.70
 
class Example(val start: Int, val stop: Int, val incr: Int, val comment: String)
 
var examples = listOf(
Example(-2, 2, 1, "Normal"),
Example(-2, 2, 0, "Zero increment"),
Example(-2, 2, -1, "Increments away from stop value"),
Example(-2, 2, 10, "First increment is beyond stop value"),
Example(2, -2, 1, "Start more than stop: positive increment"),
Example(2, 2, 1, "Start equal stop: positive increment"),
Example(2, 2, -1, "Start equal stop: negative increment"),
Example(2, 2, 0, "Start equal stop: zero increment"),
Example(0, 0, 0, "Start equal stop equal zero: zero increment")
)
 
fun sequence(ex: Example, limit: Int) =
if (ex.incr == 0) {
List(limit) { ex.start }
}
else {
val res = mutableListOf<Int>()
var c = 0
var i = ex.start
while (i <= ex.stop && c < limit) {
res.add(i)
i += ex.incr
c++
}
res
}
 
fun main(args: Array<String>) {
for (ex in examples) {
println(ex.comment)
System.out.printf("Range(%d, %d, %d) -> ", ex.start, ex.stop, ex.incr)
println(sequence(ex, 10))
println()
}
}
Output:
Normal
Range(-2, 2, 1) -> [-2, -1, 0, 1, 2]

Zero increment
Range(-2, 2, 0) -> [-2, -2, -2, -2, -2, -2, -2, -2, -2, -2]

Increments away from stop value
Range(-2, 2, -1) -> [-2, -3, -4, -5, -6, -7, -8, -9, -10, -11]

First increment is beyond stop value
Range(-2, 2, 10) -> [-2]

Start more than stop: positive increment
Range(2, -2, 1) -> []

Start equal stop: positive increment
Range(2, 2, 1) -> [2]

Start equal stop: negative increment
Range(2, 2, -1) -> [2, 1, 0, -1, -2, -3, -4, -5, -6, -7]

Start equal stop: zero increment
Range(2, 2, 0) -> [2, 2, 2, 2, 2, 2, 2, 2, 2, 2]

Start equal stop equal zero: zero increment
Range(0, 0, 0) -> [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

Perl[edit]

None of these sequences are 'errors', though some are of infinite length, and #5 has a length of zero.

for $i (
[ -2, 2, 1], #1 Normal
[ -2, 2, 0], #2 Zero increment
[ -2, 2, -1], #3 Increments away from stop value
[ -2, 2, 10], #4 First increment is beyond stop value
[ 2, -2, 1], #5 Start more than stop: positive increment
[ 2, 2, 1], #6 Start equal stop: positive increment
[ 2, 2, -1], #7 Start equal stop: negative increment
[ 2, 2, 0], #8 Start equal stop: zero increment
[ 0, 0, 0], #9 Start equal stop equal zero: zero increment
) {
$iter = gen_seq(@$i);
printf "start: %3d stop: %3d incr: %3d | ", @$i;
printf "%4s", &$iter for 1..10;
print "\n";
}
 
sub gen_seq {
my($start,$stop,$increment) = @_;
$n = 0;
return sub {
$term = $start + $n++ * $increment;
return $term > $stop ? '' : $term;
}
}
Output:
start:  -2  stop:   2  incr:   1 |   -2  -1   0   1   2
start:  -2  stop:   2  incr:   0 |   -2  -2  -2  -2  -2  -2  -2  -2  -2  -2
start:  -2  stop:   2  incr:  -1 |   -2  -3  -4  -5  -6  -7  -8  -9 -10 -11
start:  -2  stop:   2  incr:  10 |   -2
start:   2  stop:  -2  incr:   1 |
start:   2  stop:   2  incr:   1 |    2
start:   2  stop:   2  incr:  -1 |    2   1   0  -1  -2  -3  -4  -5  -6  -7
start:   2  stop:   2  incr:   0 |    2   2   2   2   2   2   2   2   2   2
start:   0  stop:   0  incr:   0 |    0   0   0   0   0   0   0   0   0   0

Perl 6[edit]

Works with: Rakudo version 2018.08

It would be odd to call ANY of these sequences "wrong" in Perl 6. Perl 6 specifically has built in capability of working with infinite sequences. Just because a sequence is infinite, doesn't mean you can't define it, work with it or use values from it. Sure, if you try to reify the whole thing you may be waiting a while, but there is nothing preventing you from using a portion of it.

Perl 6 sequence definitions specifically allow "ending points" that may never occur in the sequence. Since that is the case, you don't even really need to specify a stop value. You can just say stop at "whatever". Whatever is spelled "*" in Perl 6.

There is additional syntax you can add to stop at the nearest value, last value previous or first value successor to the "stop value" (Note I didn't say less than or greater than the stop value since the sequence can be ascending, descending or non-monotonic).

Also note: The iterator function for the sequence is literally a function. It is any expression that produces a value. These sequences all use simple arithmatic increments but that is not a limitation of the sequence operator.

# Given sequence definitions
# start stop inc. Comment
for -2, 2, 1, # Normal
-2, 2, 0, # Zero increment
-2, 2, -1, # Increments away from stop value
-2, 2, 10, # First increment is beyond stop value
2, -2, 1, # Start more than stop: positive increment
2, 2, 1, # Start equal stop: positive increment
2, 2, -1, # Start equal stop: negative increment
2, 2, 0, # Start equal stop: zero increment
0, 0, 0, # Start equal stop equal zero: zero increment
 
# Additional "problematic" sequences
1, Inf, 3, # Endpoint literally at infinity
0, π, τ/8, # Floating point numbers
1.4, *, -7.1 # Whatever
 
-> $start, $stop, $inc {
my $seq = flat ($start, *+$inc$stop);
printf "Start: %3s, Stop: %3s, Increment: %3s | ", $start, $stop.Str, $inc;
# only show up to the first 15 elements of possibly infinite sequences
put $seq[^15].grep: +*.defined
}
 
# For that matter the start and end values don't need to be numeric either. Both
# or either can be a function, list, or other object. Really anything that a
# "successor" function can be defined for and produces a value.
say "\nDemonstration of some other specialized sequence operator functionality:";
# Start with a list, iterate by multiplying the previous 3 terms together
# and end with a term defined by a function.
put 1, -.5, 2.sqrt, * * * * **.abs < 1e-2;
 
# Start with an array, iterate by rotating, end when 0 is in the last place.
say [0,1,2,3,4,5], *.rotate(-1)!*.tail;
 
# Iterate strings backwards.
put 'xp''xf';
Output:
Start:  -2, Stop:   2, Increment:   1 | -2 -1 0 1 2
Start:  -2, Stop:   2, Increment:   0 | -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2 -2
Start:  -2, Stop:   2, Increment:  -1 | -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 -16
Start:  -2, Stop:   2, Increment:  10 | -2 8 18 28 38 48 58 68 78 88 98 108 118 128 138
Start:   2, Stop:  -2, Increment:   1 | 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Start:   2, Stop:   2, Increment:   1 | 2
Start:   2, Stop:   2, Increment:  -1 | 2
Start:   2, Stop:   2, Increment:   0 | 2
Start:   0, Stop:   0, Increment:   0 | 0
Start:   1, Stop: Inf, Increment:   3 | 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43
Start:   0, Stop: 3.141592653589793, Increment: 0.7853981633974483 | 0 0.7853981633974483 1.5707963267948966 2.356194490192345 3.141592653589793
Start: 1.4, Stop:   *, Increment: -7.1 | 1.4 -5.7 -12.8 -19.9 -27 -34.1 -41.2 -48.3 -55.4 -62.5 -69.6 -76.7 -83.8 -90.9 -98

Demonstration of some other specialized sequence operator functionality:
1 -0.5 1.4142135623730951 -0.7071067811865476 0.5000000000000001 -0.5000000000000002 0.176776695296637 -0.04419417382415928 0.0039062500000000095
([0 1 2 3 4 5] [5 0 1 2 3 4] [4 5 0 1 2 3] [3 4 5 0 1 2] [2 3 4 5 0 1] [1 2 3 4 5 0])
xp xo xn xm xl xk xj xi xh xg xf

Phix[edit]

Phix for loops do not allow a zero step (neither are any floating point values permitted).
The following shows the behaviour of both for and while loops, and the latter has a couple of additional commented out termination checks that might be appropriate in some cases.

procedure test(integer start, stop, step, string legend, bool bFor)
sequence res = {}
if bFor then
try
for i=start to stop by step do
res &= i
if length(res)>9 then exit end if
end for
res = sprint(res)
catch e
res = e[E_USER]
end try
else
integer i = start
while (step>=0 and i<=stop)
or (step<=0 and i>=stop) do
res &= i
if length(res)>9 then exit end if
-- if i=stop then exit end if
-- if step=0 then exit end if
i += step
end while
res = sprint(res)
end if
printf(1,"%-43s: %s\n",{legend,res})
end procedure
 
for i=1 to 2 do
 ?iff(i=1?"for":"while")
test(-2, 2, 1, "Normal" ,i=1)
test(-2, 2, 0, "Zero increment" ,i=1)
test(-2, 2,-1, "Increments away from stop value" ,i=1)
test(-2, 2,10, "First increment is beyond stop value" ,i=1)
test( 2,-2, 1, "Start more than stop: positive increment" ,i=1)
test( 2, 2, 1, "Start equal stop: positive increment" ,i=1)
test( 2, 2,-1, "Start equal stop: negative increment" ,i=1)
test( 2, 2, 0, "Start equal stop: zero increment" ,i=1)
test( 0, 0, 0, "Start equal stop equal zero: zero increment",i=1)
puts(1,"\n")
end for
Output:
"for"
Normal                                     : {-2,-1,0,1,2}
Zero increment                             : for loop error, step is 0
Increments away from stop value            : {}
First increment is beyond stop value       : {-2}
Start more than stop: positive increment   : {}
Start equal stop: positive increment       : {2}
Start equal stop: negative increment       : {2}
Start equal stop: zero increment           : for loop error, step is 0
Start equal stop equal zero: zero increment: for loop error, step is 0

"while"
Normal                                     : {-2,-1,0,1,2}
Zero increment                             : {-2,-2,-2,-2,-2,-2,-2,-2,-2,-2}
Increments away from stop value            : {}
First increment is beyond stop value       : {-2}
Start more than stop: positive increment   : {}
Start equal stop: positive increment       : {2}
Start equal stop: negative increment       : {2}
Start equal stop: zero increment           : {2,2,2,2,2,2,2,2,2,2}
Start equal stop equal zero: zero increment: {0,0,0,0,0,0,0,0,0,0}

Python[edit]

Python has the range function.

import re
from itertools import islice # To limit execution if it would generate huge values
# list(islice('ABCDEFG', 2)) --> ['A', 'B']
# list(islice('ABCDEFG', 4)) --> ['A', 'B', 'C', 'D']
 
 
data = '''
start stop increment Comment
-2 2 1 Normal
-2 2 0 Zero increment
-2 2 -1 Increments away from stop value
-2 2 10 First increment is beyond stop value
2 -2 1 Start more than stop: positive increment
2 2 1 Start equal stop: positive increment
2 2 -1 Start equal stop: negative increment
2 2 0 Start equal stop: zero increment
0 0 0 Start equal stop equal zero: zero increment
'''

 
table = [re.split(r'\s\s+', line.strip()) for line in data.strip().split('\n')]
#%%
for _start, _stop, _increment, comment in table[1:]:
start, stop, increment = [int(x) for x in (_start, _stop, _increment)]
print(f'{comment.upper()}:\n range({start}, {stop}, {increment})')
values = None
try:
values = list(islice(range(start, stop, increment), 999))
except ValueError as e:
print('  !!ERROR!!', e)
if values is not None:
if len(values) < 22:
print(' =', values)
else:
print(' =', str(values[:22])[:-1], '...')
 
Output:
NORMAL:
  range(-2, 2, 1)
    = [-2, -1, 0, 1]
ZERO INCREMENT:
  range(-2, 2, 0)
  !!ERROR!! range() arg 3 must not be zero
INCREMENTS AWAY FROM STOP VALUE:
  range(-2, 2, -1)
    = []
FIRST INCREMENT IS BEYOND STOP VALUE:
  range(-2, 2, 10)
    = [-2]
START MORE THAN STOP: POSITIVE INCREMENT:
  range(2, -2, 1)
    = []
START EQUAL STOP: POSITIVE INCREMENT:
  range(2, 2, 1)
    = []
START EQUAL STOP: NEGATIVE INCREMENT:
  range(2, 2, -1)
    = []
START EQUAL STOP: ZERO INCREMENT:
  range(2, 2, 0)
  !!ERROR!! range() arg 3 must not be zero
START EQUAL STOP EQUAL ZERO: ZERO INCREMENT:
  range(0, 0, 0)
  !!ERROR!! range() arg 3 must not be zero

REXX[edit]

Note that a do loop with zero by value, or a do loop that goes in the "wrong" direction is not considered an error in REXX as there are other methods of limiting the range (or stopping condition) within the loop body.   A special check was made in this REXX version to check for a runaway (race) condition.

The REXX language will cause the do loop index to be checked at the "head" of the do loop to see if the index falls within the specified iteration range   (if there is one).

/*REXX program demonstrates several versions of  DO  loops with  "unusual"  iterations. */
@.=; @.1= ' -2 2 1 ' /*"normal". */
@.2= ' -2 2 0 ' /*"normal", zero increment.*/
@.3= ' -2 2 -1 ' /*increases away from stop, neg increment.*/
@.4= ' -2 2 10 ' /*1st increment > stop, positive increment.*/
@.5= ' 2 -2 1 ' /*start > stop, positive increment.*/
@.6= ' 2 2 1 ' /*start equals stop, positive increment.*/
@.7= ' 2 2 -1 ' /*start equals stop, negative increment.*/
@.8= ' 2 2 0 ' /*start equals stop, zero increment.*/
@.9= ' 0 0 0 ' /*start equals stop, zero increment.*/
zLim= 10 /*a limit to check for runaway (race) loop.*/
/*a zero increment is not an error in REXX.*/
do k=1 while @.k\=='' /*perform a DO loop with several ranges. */
parse var @.k x y z . /*obtain the three values for a DO loop. */
say
say center('start of performing DO loop number ' k " with range: " x y z, 79, '═')
zz= 0
do j=x to y by z until zz>=zLim /* ◄─── perform the DO loop.*/
say ' j ───►' right(j, max(3, length(j) ) ) /*right justify J for alignment*/
if z==0 then zz= zz + 1 /*if zero inc, count happenings*/
end /*j*/
 
if zz>=zLim then say 'the DO loop for the ' k " entry was terminated (runaway)."
say center(' end of performing DO loop number ' k " with range: " x y z, 79, '─')
say
end /*k*/ /*stick a fork in it, we're all done. */
output:
══════════start of performing DO loop number  1  with range:  -2 2 1═══════════
   j ───►  -2
   j ───►  -1
   j ───►   0
   j ───►   1
   j ───►   2
────────── end  of performing DO loop number  1  with range:  -2 2 1───────────


══════════start of performing DO loop number  2  with range:  -2 2 0═══════════
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
   j ───►  -2
the DO loop for the  2  entry was terminated (runaway).
────────── end  of performing DO loop number  2  with range:  -2 2 0───────────


══════════start of performing DO loop number  3  with range:  -2 2 -1══════════
────────── end  of performing DO loop number  3  with range:  -2 2 -1──────────


══════════start of performing DO loop number  4  with range:  -2 2 10══════════
   j ───►  -2
────────── end  of performing DO loop number  4  with range:  -2 2 10──────────


══════════start of performing DO loop number  5  with range:  2 -2 1═══════════
────────── end  of performing DO loop number  5  with range:  2 -2 1───────────


═══════════start of performing DO loop number  6  with range:  2 2 1═══════════
   j ───►   2
─────────── end  of performing DO loop number  6  with range:  2 2 1───────────


══════════start of performing DO loop number  7  with range:  2 2 -1═══════════
   j ───►   2
────────── end  of performing DO loop number  7  with range:  2 2 -1───────────


═══════════start of performing DO loop number  8  with range:  2 2 0═══════════
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
   j ───►   2
the DO loop for the  8  entry was terminated (runaway).
─────────── end  of performing DO loop number  8  with range:  2 2 0───────────


═══════════start of performing DO loop number  9  with range:  0 0 0═══════════
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
   j ───►   0
the DO loop for the  9  entry was terminated (runaway).
─────────── end  of performing DO loop number  9  with range:  0 0 0───────────

zkl[edit]

// zero increment (ie infnite loop) throws an error
// if stop is "*", the loop is has no end (ie infinite)
// stop is included unless step steps skips it
// if start > stop is a dead loop
// ranges ([a..b,c]) are lazy lists
fcn looper([(start,stop,increment)]){
print(" %3s  %3s\t%2d --> ".fmt(start,stop,increment));
try{ foreach n in ([start..stop,increment]){ print(n," ") } }
catch{ print(__exception) }
println();
}
println("start stop increment");
T( T(-2,2,1),T(-2,2,0),T(-2,2,-1),T(-2,2,10),T( 2,-2,1),
T( 2,2,1),T( 2,2,-1),T( 2,2,0),T( 0,0,0),
T(0.0, (0.0).pi, 0.7853981633974483), T("a","e",1), T("e","a",1) )
.apply2(looper); // apply2 is apply (map) without saving results
Output:
start stop  increment
  -2    2	 1 --> -2 -1 0 1 2 
  -2    2	 0 --> ValueError(range: step == 0)
  -2    2	-1 --> 
  -2    2	10 --> -2 
   2   -2	 1 --> 
   2    2	 1 --> 2 
   2    2	-1 --> 2 
   2    2	 0 --> ValueError(range: step == 0)
   0    0	 0 --> ValueError(range: step == 0)
   0  3.14159	 0 --> 0 0.785398 1.5708 2.35619 3.14159 
   a    e	 1 --> a b c d e 
   e    a	 1 -->