If you want to do something more sophisticated than that using regexes, you need the extra features provided by the Regexp::Grammars module. Have a look at the distribution's demo/ subdirectory for some examples of parsing and then manipulating arithmetic expressions (in particular: the various demo_calc*.pl files).

And, yes, it's unfortunate that Perl now has the equivalent of full grammar-based matching, without any facility for extracting useful information from such matches. I know that Damian Conway and Nicholas Clark were discussing that very limitation (and what might possibly be done about it in a future version of Perl) at YAPC::EU a few weeks ago, so maybe Perl will eventually get a native implementation of that missing feature as well.

But in the meantime, have a look at Regexp::Grammars (or Marpa or Parse::Yapp). Perl 5 regexes aren't (yet) full grammars and that seems to be what you want here. Any of those modules can give you that, and Regexp::Grammars can give it to you in (decidedly non-vanilla) regexes.

Thanks for the clarification and the pointers! I will have a look. Regarding building a parse tree with a recursive regexp, I found an inspiring example at ppp from demerphq (which seems a tiny bit hackish though). It uses a different technique with assertions that evaluate Perl code:

use strict;
use warnings;
use Data::Dumper;

$^R=undef;
my $re = qr/
    (?{[$^R]})                      # initialize a new node in the par
+setree
                                    #   preserving the old as the 0th 
+element
    \(                              # match a paren
    (?>                             # atomic match
        ([^()]+)                    # capture
        (?{
            push @{$^R},$1;         # push the capture string into the
+ current node
            $^R                     # propagate the current node
        })
    |
        (?0)                        # recurse
        (?{
            push @{$^R->[0]},$^R;   # push the current node into the o
+ld node
            shift @{$^R}            # and restore the old node to be t
+he current
                                    # (while removing it from the new)
        })
    )*
    \)
/x;

"(a(b)(c(d)e)(f)g)" =~ m/$re/x
           and
print "matched\n"; 
print Dumper $^R;
[download]

which gives (with perl 5.10.0):

matched
$VAR1 = [
          undef,
          'a',
          [
            'b'
          ],
          [
            'c',
            [
              'd'
            ],
            'e'
          ],
          [
            'f'
          ],
          'g'
        ];
[download]

But my current goal is not parsing, it is optimization of stringified output from parsed productions. It would have been cool to do optimizing transforms with recursive pattern substitutions. Then optimizations could be applied on the most nested level first, and propagate to higher levels.

Yes, Regexp::Grammars basically just automates the same technique as demerphq's example uses: injection of a distributed stack-based tree-constructor via embedded code blocks.

But you really want that automation if the regex gets much more complex than the above example. Apart from the obvious maintainabilty issues of sprinkling dozens of (?{...}) blocks through your regex, there's a much more fundamental problem: if the regex can ever backtrack a repetition, that will almost certainly screw up the synchronization between the stack and the parser state, unless you take special precautions to unwind the stack correctly (which is what the module does for you).

You may view the original node and the consideration vote tally.

You may view the original node and the consideration vote tally.