Well let me see if I can shed some light.

$ret = $1 if $g =~ /\n\s*\n((?m:^#.*\n)+)$id/;
[download]

/\n\s*\n        # This matches your blank line between comments.
 (              # Capture into $1.
   (?m:         # Don't capture, but match this group as multiline. Wh
+ich means match '^' and '$'
                # at the beginning or end of any line in the string.
      ^#        # Match the comment delimiter at the start of a line.
      .*\n      # Match everything up to, and including, the newline a
+t the end of a line.
   )+           # Match this group one or more times.
 )              # End of $1 capture.
 $id            # This is the key you're looking for.
/x
[download]

kelan

Yak it up with Fullscreen ChatBox

And since I'm getting excited about Perl6 pattern matching and want the practice, let's do this with a Perl6 rule, line for line with the above Perl5 regex.

# <Perl6>

/^^ \h* $$    # Match a line that only contains horizontal whitespace,
+ if anything (blank line).
 (            # Capture to $1.
   [          # Group, but don't capture.
     ^^ \#    # Comment delimiter at the start of a line.
     .* $$    # Everything until the end of the line, includes the new
+line.
   ]+         # One or more comment lines in a row.
 )            # End $1 capture.
 $id          # The key we're looking for. This matches as a literal s
+tring, not a pattern.
/

# </Perl6>
[download]

And here's the short version:

m/^^ \h* $$ $ret:=([^^ \# .* $$]+) $id/;
[download]

You can see I added a small bit ('$ret:=') just before the capture. This binds the value of the capture, $1 in this case, to $ret if the pattern matches, but not if it fails. Just what we want.

kelan

---

Yak it up with Fullscreen ChatBox