Re^5: Perl Parsing Based on Supplied Precedence


Perl: the Markov chain saw
	PerlMonks

Re^5: Perl Parsing Based on Supplied Precedence

by wirito (Acolyte)

on Nov 07, 2012 at 16:42 UTC ( [id://1002692]=note: print w/replies, xml )

Need Help??

in reply to Re^4: Perl Parsing Based on Supplied Precedence
in thread Perl Parsing Based on Supplied Precedence

Cool!
I was playing to add parenthesis too. But as I have a 'difficult' access to CPAN I didn't tried Text::Balanced. However I reach the regex monkey way to deal with it, just adding a two globals and a new loop on parse():

my $par;
my $par_count = 0;
sub parse{
  my ($regex,$input)=@_;
  $input =~ /^__PAREN(\d+)__$/ and return $par->{$1};

  $input=~s/\s//g;
  while( $input =~ /\(([^()]+)\)/ ) {
    my $sub = $1;
    $par->{$par_count} = parse($regex, $sub);
    my $tag = '__PAREN' . $par_count++ . '__';
    $input =~ s/\(([^()]+)\)/$tag/;
  }

  for(reverse @$regex){
    if($input=~m/(.+)($_)(.+)/) {
      my ($before,$op,$after,$node) = ($1, $2, $3);
      $node->{$op}=[parse($regex,$before),parse($regex,$after)];
      return $node;
    }
  }
  return $input;
}
[download]

Your variable handling has amazed me :)