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

use Data::Dumper;
$Data::Dumper::Sortkeys = 1;

my $in1 = 'data1.txt';
my $in2 = 'data2.txt';

# fill hashes (quick and dirty!!!)
my %hash1 = map { chomp; split m{;}, $_, 2 } grep { ! /^Kuerzel/ } do { open my $fh, '<', $in1; <$fh> };
my %hash2 = map { chomp; split m{;}, $_, 2 } grep { ! /^Kuerzel/ } do { open my $fh, '<', $in2; <$fh> };

# fill result hash; entries of hash2 OVERWRITE those from hash1
my %result = ( %hash1, %hash2 );

# check data
print Dumper( \%hash1, \%hash2, \%result );



__END__
Content of data files

* data1.txt:
Kuerzel;Gesetz
ABC;abc
DEF;def
GHI;ghi

* data2.txt
Kuerzel;Gesetz
ABC;123
JKL;jkl

Script's result (with comment of the source):

$VAR1 = {
          'ABC' => 'abc',
          'DEF' => 'def',
          'GHI' => 'ghi'
        };
$VAR2 = {
          'ABC' => '123',
          'JKL' => 'jkl'
        };
$VAR3 = {
          'ABC' => '123',       # overwritten by data from data2
          'DEF' => 'def',       # from data1
          'JKL' => 'jkl',       # from data2
          'GHI' => 'ghi'        # from data1
        };