Re: Text::Balanced v Parse::RecDescent

Andrew Savige wrote:

> It definitely runs a lot faster.
> However, [it screws up for] the original test data:

Mea culpa. I'm a little distracted just at the moment and I
botched the upgrade to (?>...). Here's a solution that actually
works for both data sets (and, I devoutly hope, for all others too!):

-----cut----------cut----------cut----------cut----------cut-----

use re 'eval';

our $quoted = qr/ ' [^'\\]* (?> (?> \\. [^'\\]* )* ) '  # Match 'str'
                 | " [^"\\]* (?> (?> \\. [^"\\]* )* ) "  # Match 'str'
                 /x;

our $element = qr/ (?> (?: (?> [^'"{},]+ )      # Match non-special characters
                      | \\.                      # Or match escaped anything
                      | $quoted                  # Or match quoted anything
                      | (??{$nested})            # Or match {...,...,...}
                    )+ )	                        # ...as many times as possible
                  /xs;

our $nested  = qr/ [{]                          # Match {
                    (?> (?: $element , )* )      # Match list of subelements
                    $element?                    # Match last subelement
                    [}]                          # Match }
                  /x;


while (<DATA>) {
         @fields = m/\G ( $element ) ,? /gx;     # Capture elements repeatedly

         use Data::Dumper 'Dumper';
         print Dumper( @fields );

         print "=================\n";
}

__DATA__
abc, ',def'  "\"ab'c,}" xyz , fred IN { 1, "x}y",3 } x, 'z'
{1}, hello one two three four five six seven eight nine heat-death-of-the-universe

-----cut----------cut----------cut----------cut----------cut-----


> Will the new Perl 6 pattern matching be a vast improvement for these
> sort of parsing problems?

No. Because it still won't do our thinking for us. ;-)

However, syntactically it will be much cleaner, and it will
probably feature much better debugging tools. Both of which
*will* make a big difference.

Here's the same thing in Perl 6:

-----cut----------cut----------cut----------cut----------cut-----

grammar CSV::Nested {

     rule quoted  { ' <-['\\]>*  [ \\. <[^'\\]>* ]*: '  # Match 'str'
                  | " <-["\\]>*  [ \\. <[^"\\]>* ]*: "  # Match 'str'
                  }

     rule element { [ <-['"{},]>+:       # Match non-special characters
                    | \\.                # Or match escaped anything
                    | <quoted>           # Or match quoted anything
                    | <nested>           # Or match {...,...,...}
                    ]+:	                # ...as many times as possible
                  }

     rule nested  { \{                   # Match {
                    [ <element> , ]*:    # Match list of subelements
                    <element>?           # Match last subelement
                    \}                   # Match }
                  }
}


while <$*DATA> {
         @fields = m:ec/ ( <CSV::Nested.element> ) ,? /;     # Capture elements repeatedly

         use Data::Dumper 'Dumper';
         print Dumper( @fields );

         print "=================\n";
}

__DATA__
abc, ',def'  "\"ab'c,}" xyz , fred IN { 1, "x}y",3 } x, 'z'
{1}, hello one two three four five six seven eight nine heat-death-of-the-universe

-----cut----------cut----------cut----------cut----------cut-----


Damian