Build array from an output

I am trying to create a script that will take output of a system command and I want to organize the data into an array.

The script is to discover VMs residing on a local ESXi server I have. I am using plink.exe to send the command to the server and then it returns a list of VMs that looks like this.

Vmid   Name                File                 Guest OS      Version   Annota开发者_如何转开发tion
128    NS01    [datastore2] NS01/NS01.vmx     ubuntu64Guest   vmx-07
144    NS02    [datastore2] NS02/NS02.vmx     ubuntu64Guest   vmx-07
208    MX01    [datastore2] MX01/MX01.vmx     ubuntu64Guest   vmx-07
224    SQL01   [datastore2] SQL01/SQL01.vmx   ubuntu64Guest   vmx-07
240    WS01    [datastore2] WS01/WS01.vmx     ubuntu64Guest   vmx-07

How would I take this and make an array out of it? The only columns that really matter are VMID, Name, File

The command I am using to get the output is this.

# Parse ESX\ESXi server for virtual machines that reside on it
system ("$plink \-batch \-pw $esx_password $esx_user\@$esx_host vim-cmdvmsvc/getallvms\n");

Any insight would be great.

---

Yes, use backquotes not system if you want to parse the output.

my $cmd 
    = "$plink -batch -pw $esx_password $esx_user\@$esx_host "
    . "vim-cmdvmsvc/getallvms"
    ;
my @lines = `$cmd`;

system will just use your standard out.

Not sure why you were escaping the dashes, though...

Once you have the output, you can do the following (note that I leveraged the fixed-length fields look):

foreach ( @line ) { 
    # this is the get-it-and-do-something-else-with-it version
    my ( $vmid, $name, $file )
        = substr( $_, 0, 47 ) =~ m/^ ( \d+ ) \s+ ( \S+ ) \s+ (.*\S) \s* $/x
        ;

    # OR the store-it-in-an-array-of-hashes version:
    @{ my $h = {}}{ qw<VMID Name File> }
        = substr( $_, 0, 47 ) =~ m/^ ( \d+ ) \s+ ( \S+ ) \s+ (.*\S) \s* $/x
        ;
    push @array, $h if %$h;
}

---

If none of the data columns will contain spaces, you could easily split(/\n/) the output into lines, iterate over those, and split(/\s+/) the rows into arrays. Then extract the columns you are interested in, by index.