You didn't say what should happen if a coord triple is only in one file. And probably needs a lot of safety checks, if these are big files.

#!/usr/bin/env perl

# Store X,Y,Z coordinates from one file, along with a number and press
+ure.
# Find a matching coord in a second file, subtract pressure1 from pres
+sure2.
# Output with the number from the first file.

# Sample data:
# 0, 4.38336420e+00, -2.47429684e-01, 6.79128885e+00, 1.51447906e+05
# 1, 4.38336420e+00, -2.78662264e-01, 6.79128885e+00, 1.58727812e+05
# 2, 4.38336420e+00, -2.78662264e-01, 6.69878864e+00, 1.61132406e+05
# 3, 4.38336420e+00, -2.47429684e-01, 6.69878864e+00, 1.54500719e+05

# Assume there are 2 filenames on the command line, in the correct ord
+er.

use strict;
use warnings;

our $key_format = '%16.8f'; # Fixed decimals for use in keys
our $out_format = '%e'; # scientific notation

sub read_it {
    my $line = shift;
    chomp($line);
    my ($n, $x, $y, $z, $p) = split /,\s+/, $line;
    my $key = sprintf "$key_format,$key_format,$key_format", $x, $y, $
+z;    
    return($n, $x, $y, $z, $key, $p);
}

our %press;

# File 1
while (<>) {
    my ($n, $x, $y, $z, $key, $p) = read_it($_);
    $press{$key}{num} = $n;
    $press{$key}{press} = $p;
} continue {
    # If this is the end of the first file, close it and escape the lo
+op
    if (eof) { # "eof", not "eof()"!
        close ARGV;
        last;
    }
}

# File 2
while (<>) {
    my ($n, $x, $y, $z, $key, $p2) = read_it($_);
    if (exists($press{$key})) {
        my $pdiff = $p2 - $press{$key}{press};
        my $n1 = $press{$key}{n};
        print "%d, $out_format, $out_format, $out_format, $out_format\
+n", $n1, $x, $y, $z, $pdiff;
    }
}

exit;
[download]

$hash{"$x,$y,$z"}{no} = $no;
$hash{"$x,$y,$z"}{pressure} = $pressure;
[download]

Then with the coords from the second file, search in the hash for matching coords.

Problem: The coords have a lot of decimal places. These should be normalized before creating the hash entries, and before searching the hash keys. Decide how many significant digits are more than enough, and use sprintf with a fixed decimal ("f") descriptor.

Hello buchi2,

Regarding the part of your question And is it possible, to trim the blanks during reading automatically away, wihout the trim sub?.

Why you need to trim the string through a function and not onsite directly?

Sample of code:

#!/usr/bin/perl
use strict;
use warnings;
use feature 'say';

sub trim {
   my ($str) = @_;
   $str =~ s/^\s+|\s+$//g;
   return $str;
}

my $test = ' sample ';

say 'Before (trim):';
say $test;
say trim($test);

say 'Without (trim):';
say $test;
$test =~ s/^\s+|\s+$//g;
say $test;

__END__

$ perl test.pl
Before (trim):
 sample
sample
Without (trim):
 sample
sample
[download]

Just do it as in the function without call the function, but the only reason to use the function is to avoid retyping over and over and over again the same code.

I do not have time now, to check the rest of the code but maybe later on I will find some time (hopefully).

Hope this helps, BR.

Hello!

Thank you very much.

>Just do it as in the function without call the function, but the only reason to use the function is to avoid retyping over and over and over again the same code.

So I will keep the sub, I use it more times.

Regards, buchi