I can clarify that intent. Consider this:

my $file2 = <DATA>; 
chomp $file2;
die $file2;
__DATA__
The quick
brown
fox
[download]

It yields:

The quick at data.pl line 5, <DATA> line 1.
[download]

where as

undef $/;
my $file2 = <DATA>; 
chomp $file2;
die $file2;
__DATA__
The quick
brown
fox
[download]

yields:

The quick
brown
fox
[download]

I understand your dillema with versions, and of course, you are free to do so.

As for the counting: I suggested to count all words in one file as positives, and all words in the other file as negative. Thus, if the word "the" has the same occurrance in both files, then the value for that word in the hash will be zero. And either positive or negative if it occurs more than n time in one of them.