Again a solution with regex engine. Sorry :)

1 files changed, 40 insertions, 0 deletions

diff --git a/challenge-172/brxfork/perl/ch-1.pl b/challenge-172/brxfork/perl/ch-1.pl
new file mode 100755
index 0000000000..7dd52a28d9
--- /dev/null
+++ b/challenge-172/brxfork/perl/ch-1.pl
@@ -0,0 +1,40 @@
+#!/usr/bin/env perl 
+use re 'eval';
+
+# 1) change the integer ($m) in string. For example 7 becomes 'NNNNNNN'
+# 2) make a big regex. The regex depends of the number of primes to sum ($n)
+# 3) use the regex to find all solutions
+
+# example : the regex will match (NN)(NNNNN) on "NNNNNNN" : $1 == "NN" , $2 == "NNNNN" => 7=2+5
+
+# regex pieces 
+my $firstnumber='(NN+?)'; # find a number > 2 
+my $isprime='(??{$+=~/^(NN+)\1+$/})'; # produce "1" if number is not prime and will fail with backtracking
+my $otherbiggernumber='(\g{-2}N+?)'; #\g{-2} is the capture of the precedent number, so current number is bigger
+my $storage='(?{push @sol,[map length, @{^CAPTURE}] })(?!)'; # store solution and force bactracking with '(?!)' --  https://perldoc.pl/perlvar#@%7B%5ECAPTURE%7D
+#my $storage='(?{push @sol,[map { $+[$_] - $-[$_] } 1..$n]})' # @+ et @- https://perldoc.pl/perlvar#@LAST_MATCH_END https://perldoc.pl/perlvar#@LAST_MATCH_START 
+
+
+#for my $input ([18,2],[19,3],[7,2],[119,3],[119,4],[119,5],[119,6],[119,7],[119,8],[119,9],[341,3],[341,4],[341,5],[341,6],map [$_,3],4..99) {
+for my $input ([18,2],[19,3],[7,2],[119,8],[341,3]) {
+
+    my $m = $input->[0];
+    my $n = $input->[1];
+    my $number = "N" x $m; 
+    my @sol;
+    my $regex =
+       '^' .
+       $firstnumber . $isprime .
+       ( ($otherbiggernumber . $isprime) x ($n - 1)) .
+       '$' .
+       $storage;
+
+    $number =~ $regex ;
+    print "regex : \n" , $regex , "\n";  #
+    print "$m ($n primes) = \n";
+    for my $s (@sol) {
+        print "\t", join ' + ',@$s;
+        print "\n";
+    }
+    print "---\n" ;
+}