You could use PAR::Packer to make an executable containing all the modules you need.

Cheers,

Martin

Update: Nevermind!

Seems to have a lot of XS for being pure perl. :)

General approach I'd use: Use Time::Local and localtime to alternate between epoc time and a more standard time format. (Add days in epoc time, convert to standard, use, then convert back to epoc at a specific time of day and start over.)

I'll leave the actual code as an exercise. ;)

use Time::Local;

my $date_epoc = timelocal(0,0,12,23,11,2007);

my $end_date = timelocal(0,0,12,12,4,2008);

my @date_array;

push @date_array, '2008-11-23';

while ( $date_epoc < $end_date ) {  # Might want <= instead: I'd have 
+to check.
   $date_epoc = $date_epoc + 86400;
   my (undef, undef, undef, $day, $mon, $year) = localtime($date_epoc)
+;
   push @date_array, "$year-$mon-$day";
   $date_epoc = timelocal(0,0,12,$day,$mon,$year);
}
[download]

• You have an off-by-one error for the months.
• The last line of your loop is a no-op.
• @date_array is redundant.
• A C-style for loop works great here.
• strftime is most appropriate here.
• 86400 is better written as 24*60*60. The latter is self-documenting.
• I prefer to use timegm/gmtime instead of setting the time to noon.

use strict;
use warnings;

use POSIX       qw( strftime );
use Time::Local qw( timegm );

my ($y1,$m1,$d1) = (2007,23,11);
my ($y2,$m2,$d2) = (2007, 4,12);

my $date1 = timegm(0,0,0,$d1,$m1-1,$y1);
my $date2 = timegm(0,0,0,$d2,$m2-1,$y2);

my @dates;
for (my $date=$date1; $date<=$date2; $date+=24*60*60) {
   push @dates, strftime('%Y-%m-%d', gmtime($date));
}
[download]

1. Yes, I do. My Bad.
2. No, it's not. It's actually vitally important: It handles daylight savings time, leap seconds, and other time oddities, by ignoring them and resetting things.
3. Yes, it does. I just never think of them...
4. strftime is nice, but since you need to separate the info out anyway, I figured we might as well leave it off.
5. Good point.
6. Again: The setting to noon helps make sure you don't have to worry about the oddities in computing time. (It's less likely to make a difference at noon.) I don't care what timezone you are in, just make sure you aren't using midnight, as you'll get errors!

Thanks, I will look over this tonight. You both have been very helpful.

Thanks, this approach makes sense. Thanks!

This solution doesn't use modules at all.

use strict;
use warnings;

my $startDate = q{2007-11-23};
my $endDate   = q{2008-04-12};

my $leap = isLeap( ( split m{-}, $startDate )[ 0 ] );

my @dates = ( $startDate );
while( $startDate ne $endDate )
{
    $startDate = incByOneDay( $startDate );
    push @dates, $startDate;
}

print do{
   local $" = qq{\n};
   qq{@dates\n};
   };

sub incByOneDay
{
    my $raDaysInMonth = [
       [ 0, 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ],
       [ 0, 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 ],
       ];

    my( $y, $m, $d ) = split m{-}, $_[ 0 ];
    $d ++;
    $d = 1, $m ++
       if $d > $raDaysInMonth->[ $leap ]->[ $m ];
    $m = 1, $y ++, $leap = isLeap( $y )
       if $m > 12;

    return sprintf q{%04d-%02d-%02d}, $y, $m, $d;
}

sub isLeap
{
    my $year = shift;
    return 0 if $year % 4;
    return 1 if $year % 100;
    return 1 unless $year % 400;
    return 0;
}
[download]

I would probably use modules if starting from scratch but I had this code lying around anyway so I thought I'd adapt it.

Cheers,

JohnGG

while( $startDate lt $endDate ) would be more forgiving than while( $startDate ne $endDate ) in case $endDate happened to be something like 2009-02-29.

Well, not really, as we should be validating $startDate and $endDate before we even get to the stage of generating our date range array. However, I left that as an exercise for the reader :-)

Cheers,

JohnGG