Re^2: perl file formatting

[perlforsgs]

Tried something like this But couldnt get the required ,Any help APPRECIATED

 
#!/usr/bin/perl

my $isHeader = 1;
my $targetDate = 20161002;

foreach (<STDIN>){
    if (!$isHeader){
    my ($M12,$M11,$M481,$M55,$M989) = split(' ',$_);
    if ($M11 <= $targetDate){
        print $_;
    }
    }
    $isHeader = 0;
}

[1nickt]

Are your input data columns really separated by a space? In the sample above it looks like it may be tabs, in which case your regexp won't match. Maybe try:

split(/\s+/, $_);

Hope this helps!

[perlforsgs]

The separator is comma ,yes missed that in my snippet and I also need the date to be dynamically retrieved based on the current date of processing ,rather than hardcoded date

[GotToBTru]

tybalt89 showed you how to get the current date in the YYYYMMDD format you need.

use strict;
use warnings;

my $enddate = todate();

while(<DATA>) {
  last if ($. > 1 and $enddate < (split /,/)[1]);
  print $_;
}

sub todate {
  my ($year, $month, $day) = (localtime)[5,4,3];
  return sprintf "%04d%02d%02d",$year + 1900, $month + 1, $day;
}

__DATA__
F12,F11,F481,F55,F989
8204B,20161002,1045.469,test,Y
8204,20170112,1064.848,test,Y
8204A,20170113,1064.505,test,Y
8204B,20170114,1045.469,test,Y

$. is a built-in variable that counts lines of input. The split function implicitly operates on the current input line, producing a list (or maybe an array), and we take a slice of that to look at the only field we care about, the date in the second column.

H:\perl>perl 1178418.pl
F12,F11,F481,F55,F989
8204B,20161002,1045.469,test,Y
8204,20170112,1064.848,test,Y
8204A,20170113,1064.505,test,Y

Updated: reduced if () {last if()} to just last if().

[perlforsgs]

Actually each fields are separated by comma

[tybalt89]

#!/usr/bin/perl

# http://perlmonks.org/?node_id=1178418

use strict;
use warnings;

sub eightdigitdate
  {
  my @fields = localtime(shift);
  sprintf '%04d%02d%02d', $fields[5] + 1900, $fields[4] + 1, $fields[3
+];
  }

sub nextday
  {
  use Time::Local;
  shift() =~ /(\d{4})(\d\d)(\d\d)/;
  eightdigitdate( 60*60*24 + timelocal 0, 0, 12, $3, $2 - 1, $1 );
  }

#my $today = eightdigitdate(time);
my $today = '20161002';

print scalar <STDIN>;
print my @data = <STDIN>;

my $more = 1;
while($more)
  {
  for(@data)
    {
    my $date = (split)[1];
    if( $date < $today )
      {
      s/\b$date\b/ nextday($date) /e;
      print;
      }
    else
      {
      $more = 0;
      }
    }
  }

[perlforsgs]

F12 F11 F481 F55 F989 E1 E2 E3

8204 20160930 1064.848 test Y 10

8204A 20160930 1064.505 test Y 6 4

Thanks for all the help.Now things are getting lil more from previous.We have three more columns added E1,E2 ,E3.E3 doesnt have anything as off now .Just a placeholder. Now a new logic must be bulit for as below: F481 = IF (E2 NOT NULL) then (E1 - E2) / 2 ELSE (F481) So after computations the output must be as:

F12 F11 F481 F55 F989

8204 20160930 1064.848 test Y

8204A 20160930 1 test Y

The final output record format will be as per previous format .The first record meets the else part and hence remain unchanged . The second record gets computed as above.The new columns are never part of actual output