Re: Extracting code into a subroutine

[tobyink]

It seems you could factor out some repetition, yes. Given this...

my @diskline=`grep DISKBUSY $nmondir/$files_by_date{$date} | grep -v B
+usy`;
foreach my $line (@diskline) {
   chomp ($line);
   my @diskarray=split /,/, $line;
   shift @diskarray;
   shift @diskarray;
   for my $i (0..$#diskarray) {
      $disks_by_date{$date}[$i]=join ',', $disks_by_date{$date}[$i],$d
+iskarray[$i]
   }
}
my @memline=`grep ^MEM, $nmondir/$files_by_date{$date} | grep -v Memor
+y`;
foreach my $line (@memline) {
   chomp ($line);
   my @memarray=split /,/, $line;
   my $rused=100-$memarray[2];
   my $mused=100-$memarray[3];
   $mem_by_date{$date}[0]=join ',', $mem_by_date{$date}[0],$rused;
   $mem_by_date{$date}[1]=join ',', $mem_by_date{$date}[1],$mused;
}

You could factor out the execution of the command, looping through lines, chomping, splitting and shifting off fields like this...

sub process (&$;$) {
  my ($code, $cmd, $skipfields) = @_;
  my @lines = `$cmd`;
  for (@lines) {
    chomp;
    my @array = split /,/;
    if ($skipfields) {
      shift @array for 1..$skipfields;
    }
    local $_ = \@array;
    $code->();
  }
}

process {
   for my $i (0 .. $#$_) {
      $disks_by_date{$date}[$i] = join(
         q[,],
         $disks_by_date{$date}[$i],
         $_->[$i],
      );
   }
}
q{grep DISKBUSY $nmondir/$files_by_date{$date} | grep -v Busy}, 2;

process {
   for my $i (0 .. 1) {
      my $used = 100 - $_->[2 + $i];
      $mem_by_date{$date}[$i] = join(
         q[,],
         $mem_by_date{$date}[$i],
         $used
      );
   }
}
q{grep ^MEM, $nmondir/$files_by_date{$date} | grep -v Memory};

Actually, maybe we can go further and factor out those joins...

sub joiny {
  my ($ref, $thing) = @_;
  $$ref = join q[,], $$ref, $thing;
}

sub process (&$;$) {
  my ($code, $cmd, $skipfields) = @_;
  my @lines = `$cmd`;
  for (@lines) {
    chomp;
    my @array = split /,/;
    if ($skipfields) {
      shift @array for 1..$skipfields;
    }
    local $_ = \@array;
    $code->();
  }
}

process {
   for my $i (0 .. $#$_) {
      joiny \($disks_by_date{$date}[$i]), $_->[$i];
   }
}
q{grep DISKBUSY $nmondir/$files_by_date{$date} | grep -v Busy}, 2;

process {
   for my $i (0 .. 1) {
      my $used = 100 - $_->[2 + $i];
      joiny \($mem_by_date{$date}[$i]), $used;
   }
}
q{grep ^MEM, $nmondir/$files_by_date{$date} | grep -v Memory};

None of this is tested, as I don't have your source files to test against.