my ($before, $after) = $seq =~ /(..)$find(..)/;
[download]

... will find only one instance of a sub-sequence in a sequence. The following will find multiple instances of sub-sequences, either overlapping or non-overlapping:

>perl -wMstrict -le
"my $seq  = 'atggggtaggggat';
 my $find = 'ggg';
 print 'non-overlapping:';
 while ($seq =~ m{ (..) $find (..) }xmsg) {
   print qq{($1)$find($2)};
   }
 print 'overlapping:';
 while ($seq =~ m{ (?= (..) $find (..)) }xmsg) {
   print qq{($1)$find($2)};
   }
"
non-overlapping:
(at)ggg(gt)
(ag)ggg(at)
overlapping:
(at)ggg(gt)
(tg)ggg(ta)
(ta)ggg(ga)
(ag)ggg(at)
[download]