alias.yaml

---
a:
  b: &1
    abcde: &2
    - fghi
    - jklm
    - &3
      nopq: rstu
c:
- *1
d: *2
e: *3
f:
  test: 1
  test2: 2
g:
  test: 1
  test2: 3

detect-same.pl

#!/usr/bin/env perl

# Usage: perl detect-same.pl filename
# WARNING: experimental and ugly!

use strict;
use warnings;
use 5.010;
use lib "lib";
use YAML::PP::Parser;
use YAML::PP::Emitter;
use YAML::PP::Writer;
use Data::Dumper;

my %seen;
my @events;
my @event_data;
my $writer = YAML::PP::Writer->new;
my $emitter = YAML::PP::Emitter->new();
$emitter->set_writer($writer);
my $parser = YAML::PP::Parser->new(
    receiver => sub {
        my ($self, @args) = @_;
        my $str = YAML::PP::Parser->event_to_test_suite(\@args);
        push @event_data, \@args;
        push @events, $str;
    },
);
my $yaml = do { open my $fh, $ARGV[0] or die $!; local $/; <$fh> };
$parser->parse($yaml);

my @emit;
my $anchor_index = 0;

my $i = -1;
while (my $event = shift @events) {
    $i++;
    my $original_event = $event_data[ $i ];
    push @emit, $original_event;
    if ($event !~ m/^[+]/ or $event =~ m/^\+STR/) {
        next;
    }
    my $exists = $seen{ $event } ? ($event) : ();
    my $list = $seen{ $event } ||= { index => {}, events => {} };
    $list->{index}->{ $i }++;
    my $duplicate_count = 0;
    my $last_index;

    my $ref = \$list->{events};
    my $level = 1;
    my $last_level = $level;
    for my $e (@events) {
        if ($e =~ m/^\+/) {
            $level++;
        }
        elsif ($e =~ m/^-/) {
            $level--;
        }
        if ($$ref->{ $e } and $exists) {
            if ($level >= 0) {
                my $test = $$ref->{ $e };
                $last_index = (keys %{ $$ref->{ $e }->{index} })[0];
                $last_level = $level;
                $duplicate_count++;
            }
        }
        else {
            $exists = 0;
        }
        $$ref->{ $e } ||= { index => {}, events => {} };
        $$ref->{ $e }->{index}->{ $i }++;
        $ref = \$$ref->{ $e }->{events};
    }
    if ($duplicate_count > 1 and $last_level == 0) {
        # We've found a duplicate, no need to analyze the following events
        splice @events, 0, $duplicate_count, ();
        $i += $duplicate_count;

        my $same_event = $event_data[ $last_index ];
        $same_event->[1]->{anchor} = ++$anchor_index;
        # we remove this event and put an alias at its place
        my $delete = pop @emit;
        push @emit, [ alias_event => { value => $anchor_index } ];
    }
}

$emitter->init;
for my $event (@emit) {
    my ($type, $info) = @$event;
#    my $str = YAML::PP::Parser->event_to_test_suite([$type, $info]);
    $emitter->$type($info);
}
my $out = $emitter->writer->output;
say $out;

duplicates.yaml

---
a:
  b:
    abcde:
    - fghi
    - jklm
    - nopq: rstu

c:
- abcde:
  - fghi
  - jklm
  - nopq: rstu

d:
- fghi
- jklm
- nopq: rstu

e:
  nopq: rstu

f:
  test: 1
  test2: 2

g:
  test: 1
  test2: 3