UPDATE:

Still not sure what the requested output is, but if it's only about eliminating the edges try

   DB<101> @list= ("a".."c","DBIC","A".."C","DANCER","a".."c")
 => ("a", "b", "c", "DBIC", "A", "B", "C", "DANCER", "a", "b", "c")
 
  DB<102> grep { /DBIC/ .. /DANCER/ and ! /DBIC/ and ! /DANCER/ } @lis
+t
 => ("A", "B", "C")
[download]

not DRY but effective.

That looks very readable and produces the output I said I need, but not for some of the edge cases that other monks thought of:

say for grep { /DBIC/ .. /Dancer/ and !/DBIC/ and !/Dancer/ } qw<eval 
+DBIC::3 eval DBIC::2 DBIC::1 MyApp::3 MyApp::2 MyApp::1 Dancer::3 Dan
+cer::2 Dancer::1>;
[download]

and

say for grep { /DBIC/ .. /Dancer/ and !/DBIC/ and !/Dancer/ } qw<MyApp
+::3 MyApp::2 MyApp::1 Dancer::3 Dancer::2 Dancer::1>;
[download]

I hadn't thought of these edge cases when I posted, but the first one is a distinct possibility. The example code with reverse before works ok for these.

Thanks for showing me a new way of using the flip-flop.



- Boldra

The code is not even close to what it should be, and it's totally irrelevant if it's simply the output that is close to what it should be.

close enough ...see update.

Cheers Rolf