Passing a 2D array from one Perl script to another - perl

Question

I have a Perl script called master.pl. I have a 2D array in that called #inputarray.
I need to pass the 2D array values from master.pl to another program called child.pl and access the data in the child.pl.
I have tried a lot but I'm not able to dereference the array in child.pl.
Can you please help me?
master.pl
system "start perl child.pl $appl $count #inputarray";
child.pl
($appl, $count, #inputarray) = #ARGV;
for (my $k = 0; $k < $count + 1; $k++) {
for (my $m = 0; $m < 6; $m++) {
print "$inputarray[$k][$m] ";
}
print "\n";
}

Answer 1

Method 1:
Have a look at the standard module Data::Dumper, it's ideal for what you want.
Save your data structure in a temporary file using Data::Dumper and then read it in your second script.
Method 2:
Using Storable to store array in first script and retrieve it from other.
Edit (after you provided your code):
See you can access the array like this
master.pl
#!/usr/local/bin/perl
use strict;
use warnings;
use Storable;
my #inputarray = ([1, 2, 3], [4, 5, 6], [7, 8, 9]);
store (\#inputarray, "/home/chankey/child.$$") or die "could not store";
system("perl", "child.pl", $$) == 0 or die "error";
child.pl
#/usr/local/bin/perl
use strict;
use warnings;
use Storable;
use Data::Dumper;
my $parentpid = shift;
my $ref = retrieve("/home/chankey/child.$parentpid") or die "coudn't retrieve";
print Dumper $ref;
print $$ref[0][0]; #prints 1
Output
$VAR1 = [
[
1,
2,
3
],
[
4,
5,
6
],
[
7,
8,
9
]
]; #from Dumper
1 #from print $$ref[0][0]
As you can see from dump, you've received the #inputarray in $ref. Now use it the way you want.

Answer 2

Check Storable for any kind of perl data serialization/deserialization.
For passing data between processes on POSIX systems I use named pipes, for better compatibility with Windows you could use temporary files, using File::Temp.

Answer 3

You could use an anonymous pipe, which works in the same way on UNIX and Windows (I assume you are using Windows because of the start.
Try this:
use strict;
use warnings;
my $appl = 'orange';
my #inputarray = ([0,1,2],[3,4,5],[6,7,8]);
We don't need $count, you can get the number of elements in the array using scalar context, or the highest index number using $#inputarray;
I have omitted the start because it makes it difficult to debug (the console window closes after the run).
my $cmd = 'perl child.pl';
open(my $pipe, '|-', $cmd) or
die "Unable to execte $cmd; $!";
Using Data::Dumper we can add eval statements and reduce whitespace generation:
use Data::Dumper;
local $Data::Dumper::Purity = 1;
local $Data::Dumper::Indent = 0;
my $dat = Data::Dumper->new([\$appl,\#inputarray],
[qw($appl $inputarray)]);
print $pipe $dat->Dump();
close ($pipe);
Now for the child, which reads the pipe (input stream):
use strict;
use warnings;
my ($inp) = <STDIN>;
my ($appl, $inputarray);
eval "$inp";
print "appl = $$appl\n";
Use of eval is usually frowned upon, and it can give a security hole, so use with care. I think it is justifed here.
Your loops are a tad complex, and smell of C. These are rather more Perlish:
for my $ref (#$inputarray) {
for my $ele (#$ref) {
print "$ele "
}
print "\n"
}
YAML is safer, because it does not require the eval, but needs to be installed.