my $cnt = 0;
...
  $cnt++;
...
print $cnt, "\n";
[download]

You don't really need the $cnt variable.    Just remove the first two lines and change the last line to:

print "$.\n";
[download]

my $filename = '$mainDir\test.txt';
[download]

Variables don't interpolate in single quoted strings so you should change that to a double quoted string:

my $filename = "$mainDir\test.txt";
[download]

my $mth_end = sprintf "%04d,%02d,%02d",$y,$m,Days_in_Month($y,$m);
[download]

You never use $mth_end anywhere.

printf OUT "                                  Placement    Release    
+  Tot    Chg\n";
printf OUT "                                  Date         Date       
+  Days   Days   Cost\n";
[download]

You should be using print instead of printf:

print OUT "                                  Placement    Release     
+ Tot    Chg\n";
print OUT "                                  Date         Date        
+ Days   Days   Cost\n";
[download]

  my @dates = split '\s+', $_;
[download]

Or just:

  my @dates = split;
[download]

  if ($free >= 0) {
     my $amt = $free;
     my $cost = ($amt * 100);
     my $pfmt = "%-14s %-5s %-8s %-3s %-12s %-12s %-6s %-6s %-6s\n";
     printf OUT $pfmt, $dates[0], $dates[1], $dates[2], $dates[3], fmt
+_mdy($dates[4]), fmt_mdy($dates[5]), $diff, $amt, fmt_curr($cost);
        }
   elsif ($free < 0) {
    my $amt = 0;
    my $cost = ($amt * 100);

    my $pfmt = "%-14s %-5s %-8s %-3s %-12s %-12s %-6s %-6s %-6s\n"; 
    printf OUT $pfmt, $dates[0], $dates[1], $dates[2], $dates[3], fmt_
+mdy($dates[4]), fmt_mdy($dates[5]), $diff, $amt, fmt_curr($cost);
   }
[download]

You have a lot of code duplication in there.    You could simplify that as:

  my $amt = $free < 0 ? 0 : $free;
  my $cost = $amt * 100;
  my $pfmt = "%-14s %-5s %-8s %-3s %-12s %-12s %-6s %-6s %-6s\n";
  printf OUT $pfmt, @dates[ 0, 1, 2, 3 ], fmt_mdy( $dates[ 4 ] ), fmt_
+mdy( $dates[ 5 ] ), $diff, $amt, fmt_curr( $cost );
[download]

  $ymd =~ s/ //g;
#  return ' 'x10 if $ymd eq $NULL_DATE;
  my ($y,$m,$d) = split /\D/,$ymd;
[download]

Why remove spaces when they match /\D/ which your split will remove anyways?    The problem is that split /\D/,$ymd will only remove single non-digit characters and you need to remove multiple non-digit characters:

  my ( $y, $m, $d ) = split /\D+/, $ymd;
[download]

Or just capture the digit characters:

  my ( $y, $m, $d ) = $ymd =~ /\d+/g;
[download]

while (<$file>) {
...
sub fmt_curr {
...
sub fmt_mdy {
...
}
[download]

Subroutine names are available globally so putting them inside a while loop does not effect their scope, but it might complicate the use of variables inside the subroutine.

open PREV,'<','prev.txt' or die "$!";
my %prev_days=();
#BIGCO-C  CMO 180304 L 6/30/13  2
while (<PREV>){
  chomp;
  my @f = split /\s+/,$_;
  $prev_days{$f[2]} = $f[5];
}
[download]

while (<$file>) {
  my @f = split '\s+', $_;
  $cnt++;
# fill empty dates with month end
  my @ymd1 = split ',',$f[4] //= $prev_mth_end;
  my @ymd2 = split ',',$f[5] //= $prev_mth_end;
  my $diff = Delta_Days(@ymd1, @ymd2);
  my $prev = $prev_days{$f[2]} //= 0;

  my $amt = ($diff + $prev > 3) ? $diff + $prev - 3 : 0;
  my $cost = ($amt * 100);
  my $pfmt = "%-14s %-5s %-8s %-3s %-12s %-12s %-6s %-6s %-6s %-6s\n";
  printf OUT $pfmt, @f[0..3], fmt_mdy($f[4]), fmt_mdy($f[5]), $diff, $
+amt, fmt_curr($cost), $prev;
}
[download]

Your code is incredibly long for something that looks pretty simple, if I understand it correctly.

You haven't described your files, but I assume that the three first columns, put together, define a unique identifier for the tasks you want to process.

You first need to read the first file and store in a hash the 3-day gratuity left-overs. Something like this (untested):

my %free_days_left;
while (<$file1>) {
     chomp;
     my @fields = split /\s+/, $_;
     my $key = join "-", @fields[0..2];
     $free_days_left{$key} = $fields[5];
}
[download]

This could be made more compact, but I prefer just to explain the process as clearly as possible.

Now, the %free_days_left hash contains the free days left. You just need to read through the second file, compute the number of days, subtract the free days and output the result (again, a general idea, not tested).

while (<$file2>) {
     chomp;
     my @fields = split /\s+/, $_;
     my $key = join "-", @fields[0..2];
     my ($start_date, $end_date) = @fields[5,6];
     my $start_day = (split /,/, $start_date)[2];
     my $end_day = (split /,/, $end_date)[2];
     my $duration = $end_day - $start_day + 1:
     my $corrected_duration = exists $free_days_left{$key}? $duration 
+- $free_days_left{$key} : $duration;
     print $file3 join " ", $_, $duration, $corrected_duration, $corre
+cted_duration * $unit_price, "\n";
}
[download]

OK, the formating of the output file may need a bit more work, but I guess you've got the basic idea.

C:\Steve\Dev\PerlMonks\P-2013-08-13@1518-Time-Days-Discount-Computatio
+n>doit1.pl
Possible unintended interpolation of @ymd2 in string at C:\Steve\Dev\P
+erlMonks\P-2013-08-13@1518-Time-Days-Discount-Computation\doit1.pl li
+ne 33.

Line 33:    my @ymd1 = split ',',$dates[4] //= $prev_mth_end;

Execution of C:\Steve\Dev\PerlMonks\P-2013-08-13@1518-Time-Days-Discou
+nt-Computation\doit1.pl aborted due to compilation errors.
[download]

So my question is, what the heck does //= do?

This is the relatively new "defined or" operator

$a //= $b;

means that $a is unchanged if it is defined, and is assigned to $b if not defined.

At least, that's my understanding, I have never used it, but the following test under the Perl debugger seems to confirm it:

$ perl -de 42

Loading DB routines from perl5db.pl version 1.33
Editor support available.

Enter h or `h h' for help, or `man perldebug' for more help.

main::(-e:1):   42
  DB<1> $a //=5;

  DB<2> p $a
5
  DB<3> $a //=6;

  DB<4> p $a
5
  DB<5>
[download]

This feature added in perl v5.10.0 Refer perl5100delta:

and the statement $c //= $d; can now be used instead of $c = $d unless defined $c;

Bill

I probably won't be the first one to get there, but I'm downloading and trying to figure out what you're trying to do.

:-)

WILYCO    CNW   437106 L 2013,07,08
WILYCO    MKT     4399 L 2013,07,03
WILYCO    MKT     4559 L 2013,07,23 2013,07,26
[download]

my @ymd1 = split ',',$dates[4] //= $prev_mth_end;
my @ymd2 = split ',',$dates[5] //= $prev_mth_end;
[download]

# work out end of current month
my ($y,$m,$d) = Today();
my $mth_end = sprintf "%04d,%02d,%02d",$y,$m,Days_in_Month($y,$m);
# ($year,$month,$day) = Add_Delta_Days($year,$month,$day,$Dd);
my $prev_mth_end = sprintf "%04d,%02d,%02d",
                   Add_Delta_Days($y,$m,1,-1);
[download]

I don't see where you are taking into account for the days from the previous month (June, file 1).

For example, a record from file 1 shows 2 days:

BIGCO-C  MKT   575 L 6/30/13  2

BIGCO-C   MKT      575 L 2013,07,01 2013,07,02

From that, subtracting the 3 free days would give 0 billing days? Are you calculating Demurrage?