Did the author run it on Unix ?. I'm guessing the line $json =~ /^.+\r\n\r\n:?(.+)$/s; is removing the header but on windows should be \n\n. However, if you remove the -i the curl download doesn't include the header. Using -o you can write the result to a file.

Try this revised script (untested).

#!perl
use strict;
my $base_url= 'https://';
my $user    = '';
my $pass    =  '';
my $out_dir = 'pages';
my $format  = 'html';

my $out= `curl -u $user:$pass -i $base_url/file/root/tree?format=ids`;

if ($out =~ /.+\s*([\d,]+)$/){
  my @ids = split /,/, $1;
  foreach my $id (@ids) {
    print "$id\n";
    my $cmd = "curl -u $user:$pass $base_url/files/$id/contents?format
+=$format -o $out_dir/$id.$format";
    my $status = system($cmd);
    if ($status) {
      die "system error: $?"
    }
  }
} else {
  print "Error - No ids found";
}
[download]