#!/usr/bin/perl
use strict;
use warnings;

my %node;
my $folder = "P:\\Prod-Operations\\Maint-Eng\\Maintenance Projects\\IN
+F\\DB\\IFDB\\TAGHIERARCHY\\PerlScripts";
my $resultfile = "resultset.txt";
my $interset01 = "interset01.txt";
my $interset02 = "interset02.txt";

open (DATA, "< $folder\\$resultfile") || die "could not open file: $!"
+;
while (<DATA>) {
    my ( $c, $p ) = split /\|/;
    if ( $c eq $p ) {  # these are easy, so finish them first
        print;
        next;
    }

    if ( exists( $node{$c}{child_of} )) {
        warn "$.: bad record: $c is child of both $p and $node{$c}{chi
+ld_of}\n";
        next;
    }
    $node{$c}{child_of} = $p;
    $node{$p}{parent_of}{$c} = undef;
}

# begin the sorted output by looping over values that do not have pare
+nts:
# open (INT01,">$folder\\$interset01") or die "Can not open file $fold
+er\\$interset01 for writing, quitting\n";

for my $parent ( grep {!exists( $node{$_}{child_of} ) } keys %node ) {
    my $children = $node{$parent}{parent_of};  # ref to hash of child 
+values
    trace_down( $children, \%node );
}

sub trace_down
{
    my ( $kids, $tree ) = @_;
    for my $kid ( keys %$kids ) {
        # print INT01 "$kid|$$tree{$kid}{child_of}";
        print "$kid|$$tree{$kid}{child_of}\n";
        if ( exists( $$tree{$kid}{parent_of} )) {
            trace_down( $$tree{$kid}{parent_of}, $tree );
        }
    }
}
[download]

Apart from that, when I put your five lines of sample data (thanks for that) into my original code, it came out in this order:

A1-4100|A1
A1-4100-YZX-002|A1-4100
A1-4100-YZX-002-01|A1-4100-YZX-002
A1-4200|A1
A1-4200-ABC-001|A1-4200
[download]

A1-4200|A1
A1-4200-ABC-001|A1-4200
A1-4100|A1
A1-4100-YZX-002|A1-4100
A1-4100-YZX-002-01|A1-4100-YZX-002
[download]

As for the initial "TOP" record, are you sure you have that worked out fully? Would it be the case that every top-level parent (identified as such in my script) needs its own "top-level-string|TOP" record? If so, it would be easy to modify the code to make sure this is done for each of the top-level parents -- just add a print statement  print "$parent|TOP\n"; as the first thing in the main "for" loop.

I went ahead and did that on my own copy of the script, and ran it on the huge input sample that you posted above; adding the "TOP" lines like that actually makes it easier to inspect the output for correctness, as follows:

Use the unix "grep -n" command (there are perl versions posted at the Monastery and elsewhere -- e.g. here's mine: grepp -- Perl version of grep) to get the line numbers containing "TOP". Then, for any of those "top-level children", check the line numbers containing "\|top-level-string$". All the latter line numbers should be higher/later in the file than the corresponding TOP line.

Hello
I am now beginning to wonder at my own sanity as when I run your original code, replacing your data with the following


A1-4100-YZX-002|A1-4100
A1-4100|A1
A1-4200-ABC-001|A1-4200
A1-4200|A1
A1-4100-YZX-002-01|A1-4100-YZX-002
[download]

I get the following output


A1-4100|A1
A1-4200-ABC-001|A1-4200
A1-4100-YZX-002-01|A1-4100-YZX-002
A1-4100-YZX-002|A1-4100
A1-4200|A1
[download]

Which is very different to what your seeing. The full code I am running is..........


#!/usr/bin/perl

use strict;
use warnings;

my %node;

while (<DATA>) {
    my ( $c, $p ) = split;
    if ( $c eq $p ) {  # these are easy, so finish them first
        print;
        next;
    }

    if ( exists( $node{$c}{child_of} )) {
        warn "$.: bad record: $c is child of both $p and $node{$c}{chi
+ld_of}\n";
        next;
    }
    $node{$c}{child_of} = $p;
    $node{$p}{parent_of}{$c} = undef;
}

# begin the sorted output by looping over values that do not have pare
+nts:

for my $parent ( grep { !exists( $node{$_}{child_of} ) } keys %node ) 
+{
    my $children = $node{$parent}{parent_of};  # ref to hash of child 
+values
    trace_down( $children, \%node );
}

sub trace_down
{
    my ( $kids, $tree ) = @_;
    for my $kid ( keys %$kids ) {
        print "$kid $$tree{$kid}{child_of}\n";
        if ( exists( $$tree{$kid}{parent_of} )) {
            trace_down( $$tree{$kid}{parent_of}, $tree );
        }
    }
}

__DATA__
A1-4100-YZX-002|A1-4100
A1-4100|A1
A1-4200-ABC-001|A1-4200
A1-4200|A1
A1-4100-YZX-002-01|A1-4100-YZX-002
[download]

On the question of multiple "TOPS".... There will only ever be one "TOP" which is A1. The reason for this is that there are around 12 (I can't remember the exact number) "children" of A1 which are A1-4100, A1-4200, A1-4300 etc etc etc. Then you have (for example) "A1-4200-ABC-001". I already tried using "grep -w" (match only whole words) and using "sort" (I worked for a number of years on UNIX and still work with Linux) which I thought may fix things, however, the parent-to-child relationship does not follow strict numbering conventions, i.e. the sequence "A1-4200-ABC-001" may be a child of "A1-405-ABC-001-FF".
The top two tiers of the hierarchy follow a numbering convention (A1, A1-4600 etc) and can be followed quite easily, however, the rest has been built by a person and not by any "logical" numbering system so using partial string matching will not work, unfortunately.

I realize I am taking up a lot of your time with this and I do appreciate your input and patience with my coding inadequacies.
SlackyB