Here's a substution-based version that may be somewhat along the lines Learnlot1 was thinking of originally. It does not generalize well to handling more than two strings as the split-to-hash approaches do. It suffers from the vagueness of the implied specification as other approaches do. But it works (for some definition of 'works').
>perl -wMstrict -le "for my $ar_vector ( [ qw(ABCDE BCDEO) ], [ qw(ABCDE BBBBB) ], [ qw(ABCDEA BCDEO) ], ) { my ($s1, $s2) = @$ar_vector; printf qq{'$s1' '$s2' -> '%s' \n}, non_common($s1, $s2); } ;; sub non_common { my ($sa, $sb) = @_; ;; (my $a_sans_b = $sa) =~ s{ [\Q$sb\E]+ }{}xmsg; (my $b_sans_a = $sb) =~ s{ [\Q$sa\E]+ }{}xmsg; return $a_sans_b . $b_sans_a; } " 'ABCDE' 'BCDEO' -> 'AO' 'ABCDE' 'BBBBB' -> 'ACDE' 'ABCDEA' 'BCDEO' -> 'AAO'
In reply to Re: compare two strings and return only he unique values.
by AnomalousMonk
in thread compare two strings and return only he unique values.
by Learnlot1
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |