in reply to RegEx + vs. {1,}

Because you are trying to match string which occurs twice and 'abcdefg' is first correct candidate. I can think only about such approach for your problem: 
$ perl -le 'print $x = "abcdefgxxabcdefgzzabcdsjfhkdfab"; print $1 if 
+$x =~ /(\w{2,})(.*?\1){2}/;'
abcdefgxxabcdefgzzabcdsjfhkdfab
abcd

$ perl -le 'print $x = "abcdefgxxabcdefgzzabcdsjfhkdfab"; print $1 if 
+$x =~ /(\w{2,})(.*?\1){3}/;'
abcdefgxxabcdefgzzabcdsjfhkdfab
ab

$ perl -le 'print $x = "abcdefgxxabcdefgzzabcdsjfhkdfab"; print $1 if 
+$x =~ /(\w{2,})(.*?\1){4}/;'
abcdefgxxabcdefgzzabcdsjfhkdfab

[download]
I.e. The problem is you have to say exactly how many occurences you want (there is no 'greediness' in this case == you can't say "I want as many occurences as possible", only lowest possible number of occurences will be chosen).

Replies are listed 'Best First'.
Re^2: RegEx + vs. {1,}
by Sewi (Friar) on Oct 10, 2012 at 13:01 UTC
[reply]
      So if that's acceptable for you - use while loop to determine max amount of occurences. There will be no more than length / 2 occurences, so start with this max value and decrease it while trying to match: 
      $x = "abcdefgxxabcdefgzzabcdsjfhkdfab"; $len=int(length($x)/2);
while($x !~ /(\w{2,})(.*?\1){$len}/)
  { $len-- };
$x =~ /(\w{2,})(.*?\1){$len}/; # 'strange line'
print $1

      [download]
      (to self: do not know why I have to add 'strange line', without it nothing is printed, but $len is correctly set to 4)

      I tried to generate the list and include it in one regexp: 

      $ perl -le '$x = "abcdefgxxabcdefgzzabcdsjfhkdfab"; $len=int(length($x
+)/2); $restring = join"|", map {"(?:.*?\\1){$_}"} reverse(1..$len); p
+rint $restring; print $1 if $x =~ /(\w{2,})($restring)/;'

(?:.*?\1){15}|(?:.*?\1){14}|(?:.*?\1){13}|(?:.*?\1){12}|(?:.*?\1){11}|
+(?:.*?\1){10}|(?:.*?\1){9}|(?:.*?\1){8}|(?:.*?\1){7}|(?:.*?\1){6}|(?:
+.*?\1){5}|(?:.*?\1){4}|(?:.*?\1){3}|(?:.*?\1){2}|(?:.*?\1){1}
abcdefg

      [download]
      but it does not work as expected (probably some stupid mistake, maybe someone else can tell what's wrong with it).
[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom