Just a one-shot, throwaway script I quickly concocted up yesterday, which I... don't feel like throwing away altogether: perhaps it will be instructive for others, although I warn in advance that for generic use it would require quite some massaging.

The problem: two big files (of the same actual size) which would definitely seem to be the same file, but indeed are not, as a checksum on each shows. I want to show the actual differences:

#!/usr/bin/perl

use strict;
use warnings;

$|++;

die "Usage: $0 file1 file2\n" unless @ARGV==2;

my ($f1, $f2) = map { open my $fh, '<:raw', $_ or 
      die "Can't open `$_': $!\n"; $fh } @ARGV;

$/ = \0x100_000;

while (my $s1=<$f1>) {
    defined +(my $s2=<$f2>) or last;
    printf "Block %04d: no differences\r", $.
      and next if $s1 eq $s2;
    my @l = map length, ($s1 ^ $s2) =~ /^(\0*)(.+?)(\0*)\z/;
    printf "\nBlock %04d: A=[0 x %d], B=[* x %d], C=[0 x %d]\n", $., @
+l;
    print "\tB=[@{[ unpack '(H2)*' => substr $_, $l[0], $l[1] ]}]\n"
      for $s1, $s2;
}

__END__
[download]

C:\temp>perl compare.pl foo bar
Block 0727: no differences
Block 0728: A=[0 x 759818], B=[* x 1], C=[0 x 288757]
  B=[e7]
  B=[ef]
Block 0973: no differences
[download]