in reply to Arranging multiple lines

Two approaches seem logical to me. First, similar to yours, based on regular expressions. I would slurp in the data in one go though and then replace superfluous pieces, like this: 

use strict;
use warnings;
my $data;
{
        local $/;
        $data=<DATA>;
}
$data =~ s/\n\s*\n(From: |To: |User-Agent: )/|/g;
$data =~ s/Arrival Time: ([^|]*\|[^|]*\|[^|\n]*)\n/$1|--\n/g;
print "Arrival Time|From|To|User-Agent\n\n";
print $data;

__DATA__
Arrival Time: May 2, 2013 10:37:50.813000000

From: <sip:639gwhuaping01-14@119.38.228.43>;tag=70c8b229-1c

To: <sip:639gwhuaping01-14@119.38.228.43>

Arrival Time: May 2, 2013 10:38:05.274000000

From: <sip:639gwhuaping01-01@119.38.228.43>;tag=70c8b229-2

To: <sip:639gwhuaping01-01@119.38.228.43>

Arrival Time: May 2, 2013 10:38:05.451000000

From: <sip:639gwhuaping01-11@119.38.228.43>;tag=70c8b229-16

To: <sip:639gwhuaping01-11@119.38.228.43>

User-Agent: Quintum/1.0.0 SN/0030E130409A SW/P108-09-10

[download]

Alternatively, and probably easier to maintain, is to read the data into a hash and print/empty whenever a dataset is complete, like this (with the same __DATA__ segment as above): 

use strict;
use warnings;

sub printdata {
        my $data = shift;
        $$data{"User-Agent"} //= "--";
        print join "|", @$data{ ( "Arrival Time", "From", "To", "User-
+Agent" ) };
        print "\n\n";
        %$data = ();
}

my %data;
my ($item, $value);
print "Arrival Time|From|To|User-Agent\n\n";
while(<DATA>){
        chomp;
        next unless ($item, $value) = /^(.*?): (.*)/;
        printdata \%data if $item eq "Arrival Time" and %data;
        $data{$item} = $value;
}
printdata \%data;

__DATA__
Arrival Time: May 2, 2013 10:37:50.813000000
...

[download]

Replies are listed 'Best First'.
Re^2: Arranging multiple lines
by r2ro (Initiate) on Jun 03, 2013 at 12:36 UTC

    Thank you very much hdb,

    but again you have to forgive my ignorance on this, I got lost in incorporating your code with input and output file arguments. :(

    Please bear with me, I just started studying perl 2 days ago.

    use strict;
use warnings;

my $infile = $ARGV[0];
my $outfile = $ARGV[1];

open my $in,   "<",  $infile       or die $!;
open my $out, ">", $outfile    or die $!;

sub printdata {
        my $data = shift;
        $$data{"User-Agent"} //= "--";
        print join "|", @$data{ ( "Arrival Time", "From", "To", "User-
+Agent" ) };
        print "\n\n";
        %$data = ();
}

my %data;
my ($item, $value);
print "Arrival Time|From|To|User-Agent\n\n";
while(<$in>){
        chomp;
        next unless ($item, $value) = /^(.*?): (.*)/;
        printdata \%data if $item eq "Arrival Time" and %data;
        $data{$item} = $value;
}
printdata \%data;

close $in;
close $out;

    [download]
[reply]
[d/l]

      One way to do it is to pass the $out handle into the printdata function. Beware of commata after $out in this example: 

      use strict;
use warnings;

sub printdata {
    my $out = shift;
    my $data = shift;
    $$data{"User-Agent"} //= "--";
    print $out join "|", @$data{ ( "Arrival Time", "From", "To", "User
+-Agent" ) };
    print $out "\n\n";
    %$data = ();
}

open my $out, ">", "tmp.txt";

my %data;
my ($item, $value);
print $out "Arrival Time|From|To|User-Agent\n\n";
while(<DATA>){
    chomp;
    next unless ($item, $value) = /^(.*?): (.*)/;
    printdata $out, \%data if $item eq "Arrival Time" and %data;
    $data{$item} = $value;
}
printdata $out, \%data;

      [download]

      Alternatively, you could assemble a string in printdata, return it from the sub and print later.

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom