• I can't figure out how to get the grammar to properly differentiate between the Association, KeyList, and Scalar assignments within that clause.

  You shouldn't have to. The parser will try them all until it finds one that succeeds. The problem is that you wrote it so the first production of Attr (AttrScalar(s?)) will always succeed, so it'll never get to try the 2nd and 3rd productions (AttrKeyList(s?) and AttrAssociation(s?)).

  Instead of telling the parser to search for
  0 or more of {one of {0 or more AttrScalar} or {0 or more AttrKeyList} or {0 or more AttrAssociation}}
  you should be asking for
  0 or more of {one of AttrScalar or AttrKeyList or AttrAssociation}

  In other words,
  Attr: AttrScalar(s?) | AttrKeyList(s?) | AttrAssociation(s?)
  should be
  Attr: AttrScalar | AttrKeyList | AttrAssociation

• Your second problem is that "5" in "CounterInterval = 5" doesn't match "string".

• You never check if you've reached the end of the string. That's why it returned a parse tree even though it was incomplete.
  VCSConfig: statement(s)
  should be
  VCSConfig: statement(s) /\Z/

• Are you sure that identifiers can't be one character long?
  /[a-zA-Z][\w@]+/
  should be
  /[a-zA-Z][\w@]*/
  and
  /[a-zA-Z]\w+/
  should be
  /[a-zA-Z]\w*/

• It's bad to separate a token into multiple rules. It causes characters to be removed. (See <skip.)
  pathname:  dquote(?) /([^"]+)/ dquote(?) { $item[1] }
  should be
  pathname: /"([^"]+)"/ { dequote($item[1]) }
          | /([^"]+)/ { $item[1] }

• Attr: AttrScalar | AttrKeyList | AttrAssociation
  is *very* inefficient because all three subrules start with "attribute '='".

• "cluster" name
  will see the following string as valid
  clusterpeanut
  You normally want to force a space in there. One way is to match any identifier, than require the identifier to be "cluster".
  This problem occurs in a few other places too.

• It's very useful to uppercase tokens and keep them separate. They look similar to other rules, but you'll find that you'll be treating them a little special.

• I find it much more readable to line up the : and the | of all the rules.

make_parser.pl, generates the parser. Run it to create VCSConfigParser.pm.

use strict;
use warnings;

use Parse::RecDescent qw( );

my $grammar = <<'EOG';
   <autotree>
  
   {
      # These affect the entire parser.
      use strict;
      use warnings;
      
      sub dequote {
         my $s = $_[0];
         $s =~ s/^"//;
         $s =~ s/"\z//;
         return $s;
      }
   }

   parse      : stmt(s) /\Z/

   stmt       : clause
              | def

   clause     : "include" pathname
 
   # Pathname may or may not be surrounded by double quotes
   pathname   : STRING
              | BAREWORD

   def        : IDENT def_[ $item[1] ] { $item[2] }
   def_       : { $arg[0] eq "cluster" ?1:0 } IDENT "(" attr(s?) ")"
              | { $arg[0] eq "system"  ?1:0 } IDENT "(" attr(s?) ")"

   attr       : ATTRNAME '=' attr_val
   
   attr_val   : ident
              | string
              | number
              | key_list
              | assoc_list
              
   val        : ident
              | string
              | number

   # These aren't inlined because of <autotree>
   ident      : IDENT 
   string     : STRING
   number     : NUMBER
   key_list   : '{' <leftop: IDENT /[,;]/ IDENT> '}'
   assoc_list : '{' <leftop: key_value /[,;]/ key_value> '}'

   key_value  : IDENT '=' val


   # === Tokens ===
   
   IDENT    : /[a-zA-Z]\w*/    { $item[1] }
   ATTRNAME : /[a-zA-Z][\w@]*/ { $item[1] }
   STRING   : /"(?:[^"]+)"/    { dequote($item[1]) }
   NUMBER   : /\d+/            { $item[1] }      # Need work.
   BAREWORD : /(?:[^"]+)/      { $item[1] }

EOG

Parse::RecDescent->Precompile($grammar, 'VCSConfigParser')
   or die("Bad grammar\n");
[download]

test.pl, a sample program that uses the parser.

use strict;
use warnings;

use VCSConfigParser qw( );
use Data::Dumper    qw( Dumper );

#$::RD_TRACE = '';

my $vcs_parser = VCSConfigParser->new();
my $vcs_config = do { local $/; <DATA> };

my $tree = $vcs_parser->parse( $vcs_config );
print Dumper $tree;

__DATA__
include "types.cf"
include "LBSybase.cf"
include "OracleTypes.cf"

cluster vcs (
        UserNames = { vcs = X1Nh6WIWs6ATQ }
        Administrators = { vcs }
        CounterInterval = abc
        )

system njengsunvcs1 (
        )

system njengsunvcs2 (
        )
[download]

Notes:

• def + def_ is an optimization of

  def : IDENT { $item[1] eq "cluster" ?1:0 } IDENT "(" attr(s?) ")"
      | IDENT { $item[1] eq "system"  ?1:0 } IDENT "(" attr(s?) ")"
  [download]

  By eliminating the common prefix of the productions, the parser is sped up.

• If you want to allow nesting, change val to attr_val.

---

---

Posts are HTML formatted. Put <p> </p> tags around your paragraphs. Put <code> </code> tags around your code and data!

Titles consisting of a single word are discouraged, and in most cases are disallowed outright.

Read Where should I post X? if you're not absolutely sure you're posting in the right place.

Please read these before you post! —

• How do I compose an effective node title?
• How do I post a question effectively?
• Markup in the Monastery

Posts may use any of the Perl Monks Approved HTML tags:

a, abbr, b, big, blockquote, br, caption, center, col, colgroup, dd, del, details, div, dl, dt, em, font, h1, h2, h3, h4, h5, h6, hr, i, ins, li, ol, p, pre, readmore, small, span, spoiler, strike, strong, sub, summary, sup, table, tbody, td, tfoot, th, thead, tr, tt, u, ul, wbr

You may need to use entities for some characters, as follows. (Exception: Within code tags, you can put the characters literally.)

	For:		Use:
	&		&
	<		&lt;
	>		&gt;
	[		&#91;
	]		&#93;

Link using PerlMonks shortcuts! What shortcuts can I use for linking?

See Writeup Formatting Tips and other pages linked from there for more info.