I thought of the function index first, which gets the position of a string (or character, in this case).

my $seq = "XX**XXX!!!X**AB*!C*DXX*!!!XXX**XXX";
# The code presumes that it's a correct sequence.
# If you're getting the sequence from somewhere else you
# should check it to make sure it contains the right letters.
print "#" . index($seq, "A") . "-#" . index($seq, "D") . "\n";
[download]

This is the easiest way; your method works too, but it would involve splitting up the string and operating on that, so using index is easier.

$_ = 'XX**XXX!!!X**AB*!C*DXX*!!!XXX**XXX';
/A[!\*]*B[!\*]*C[!\*]*D/g and print '#', pos() + 1 - length($&), '-#',
+ pos();
__OUTPUT__
#14-#20
[download]

Are you given ABCD or are you given the position and length (or start and end) of the substring? Can the * and !'s occur anywhere? You should give a before and after sample. Something like this:

__DATA__
XXXXXXABCDXXXXXXXX
XX**XXX!!!X**AB*!C*DXX*!!!XXX**XXX
[download]

Want to print:

10  characters added
XXXXXXABCD**!!!***!*XX*!!!XXX**XXX
[download]

The following code may be a good starting point:

use strict;
use warnings;

my $match = 'ABCD';

while (<DATA>)
  {
  my $org = $_;
  defined (my $mutated = <DATA>) or die "Missing edited line";

  my $segment = substr $mutated, 0, index ($mutated, substr $match, 3,
+ 1) + 1;
  my $suffix = substr $mutated, index ($mutated, substr $match, 3, 1) 
++ 1;
  (my $pInsert = $segment) =~ tr/*!//cd;
  (my $pSegment = $segment) =~ tr/*!//d;
  print length ($pInsert) . " characters added\n";
  print "$pSegment$pInsert$suffix\n";
  }

__DATA__
XXXXXXABCDXXXXXXXX
XX**XXX!!!X**AB*!C*DXX*!!!XXX**XXX
[download]

Note that some error checking should be added and that there are a few assumptions about the match string and what the X sequences can actually contain.