Re^3: Data manipulation on a file

[SuicideJunkie]

Loop over for my $hostname (keys %hosts), or if you like, over for my $hostname (sort keys %hosts) so the host names come out in asciibetical order.

Between $hostname and join ':', @{ $hosts{$hostname} }, you've then got everything you need to concatenate onto your output string inside the loop. Make sure your output variable is defined outside the loop so you're not just throwing the results away when the loop ends.

%hosts might be more clear if named %mountPoints or %hostToMountPoints if you don't mind the typing.

[sstruthe]

Thanks for the replies, I am just about there with what I am needing. The print comes out nicely to STD out however I am a little confused as to how I assign the output to a scalar $output = . In the for loop what syntax is needed to assign the result to a variable concatinating each time. Once the loop of the hash is assigned to a $output then I can take it from there. Many thanks and sorry for the silly quesitons

#!/usr/bin/perl
use strict;
use warnings;

my $filer;
my %filer_hash;

for (qx(mount -t nfs | awk -F/ '{print \$1,\$3}' | sed -r 's/(blah.*:)
+|(bblah.*:)//g' |sort)) {
    chomp;
    my ($host, $mp) = split;
    push @{ $filer_hash{$host} }, $mp;
}
foreach $filer ( sort { ${filer_hash{$b}} <=> ${filer_hash{$a}} } keys
+ %filer_hash ) {
   print "$filer:@{$filer_hash{$filer}},"
}

[Corion]

Instead of printing it, append it to a variable:

my $output = '';
for ( ... ) {
    $output .= "$file:$@$filer_hash{$filer}},";
}
print "Now have: $output\n";

[sstruthe]

Wow another great bit of help, maybe one day I will get to grips with perl. For just now I thank you once again it all works great now. ta

[sstruthe]

Small problem now that the output when run on a large amount of linux hosts is producing duplicate mount points for different shares etc. This is due to the many mounts on the same mountpoint.

like this $ouputstring: host1: mountpoint1 mountpoint1 mountpoint1 mountpoint2 host25: mountpoint5 mountpoint5 mounpoint5 mountpoint6 m......78.

it would be really nice to remove the duplicates but this would mean logic before updating the array elements or logic coming out of the dump. As I am just dumping the full hash into a $scalar this could get tricky. Anyone got any suggestions on removing the duplicate mount points once the hash has been dumped to the scalar or should I put logic in the dump or even into the push. Perl is so powerfully but still a newbie here so any help appreciated. Many thanks

#!/usr/bin/perl
use strict;
use warnings;

my $filer;
my %filer_hash;

for (qx(mount -t nfs | awk -F/ '{print \$1,\$3}' | sed -r 's/(blah.*:)
+|(bblah.*:)//g' |sort)) {
    chomp;
    my ($host, $mp) = split;
    push @{ $filer_hash{$host} }, $mp;
}
foreach $filer ( sort { ${filer_hash{$b}} <=> ${filer_hash{$a}} } keys
+ %filer_hash ) {
   $outputstring .= "$filer:@{$filer_hash{$filer}},"
}
print "  This is my output : $outputstring

[CountZero]

Have a look at the uniq and distinct functions from List::MoreUtils.

For example:

use Modern::Perl qw/2015/;
use List::MoreUtils qw/uniq/;
my @mountpoints = qw/one two three three two four one one five six/;

print join ' ', uniq sort @mountpoints;

Output: five four one six three two

[sstruthe]

Thanks for the wisdom, however unfortunately if I use a module I would have to make sure that module was installed on every host that I want to use this on. Sadly that itself is a huge problem. It really has to be code from core perl I do not have the luxury of most of the CPAN modules. Unless of course "uniq" and "distinct" come with core perl. Any ideas are most welcome