--- trunk/lib/WebPAC/Parser.pm	2006/09/25 13:24:09	703
+++ trunk/lib/WebPAC/Parser.pm	2009/07/28 13:55:44	1259
@@ -9,24 +9,27 @@
 use Data::Dump qw/dump/;
 use File::Slurp;
 
-use base qw/WebPAC::Common WebPAC::Normalize/;
+use base qw/WebPAC::Common/;
 
 =head1 NAME
 
-WebPAC::Parser - parse perl normalization configuration files and mungle it
+WebPAC::Parser - parse perl normalization configuration files (rules) and mungle it
 
 =head1 VERSION
 
-Version 0.05
+Version 0.08
 
 =cut
 
-our $VERSION = '0.05';
+our $VERSION = '0.08';
 
 =head1 SYNOPSIS
 
 This module will parse L<WebPAC::Normalize/lookup> directives and generate source
-to produce lookups and normalization.
+to produce lookups and normalization. It will also parse other parts of
+source to produce some of DWIM (I<Do What I Mean>) magic
+(like producing MARC oputput using L<WebPAC::Output::MARC> if there are C<marc_*>
+rules in normalisation).
 
 It's written using L<PPI>, pure-perl parser for perl and heavily influenced by
 reading about LISP. It might be a bit over-the board, but at least it removed
@@ -46,6 +49,7 @@
   my $parser = new WebPAC::Parser(
   	config => new WebPAC::Config(),
 	base_path => '/optional/path/to/conf',
+	only_database => $only
   );
 
 =cut
@@ -66,18 +70,6 @@
 	$self ? return $self : return undef;
 }
 
-=head2 lookup_create_rules
-
-  my $source = $parser->lookup_create_rules($database, $input);
-
-=cut
-
-sub lookup_create_rules {
-	my $self = shift;
-	my ($database,$input) = @_;
-	return $self->{_lookup_create}->{ _q($database) }->{ _q($input) };
-}
-
 =head2 valid_database
 
   my $ok = $parse->valid_database('key');
@@ -100,8 +92,8 @@
 
 sub valid_database_input {
 	my $self = shift;
-
 	my ($database,$input) = @_;
+	$input = _input_name($input);
 	return defined($self->{valid_inputs}->{ _q($database) }->{ _q($input) });
 }
 
@@ -116,6 +108,7 @@
 sub depends {
 	my $self = shift;
 	my ($database,$input) = @_;
+	$input = _input_name($input);
 	$self->_get_logger->debug("depends($database,$input)");
 	return unless (
 		defined( $self->{depends}->{ _q($database) } ) &&
@@ -124,6 +117,109 @@
 	return $self->{depends}->{ _q($database) }->{ _q($input) };
 }
 
+=head2 have_lookup_create
+
+  my @keys = $parser->have_lookup_create($database, $input);
+
+=cut
+
+sub have_lookup_create {
+	my $self = shift;
+	my ($database,$input) = @_;
+	$input = _input_name($input);
+	return unless (
+		defined( $self->{_lookup_create_key}->{ _q($database) } ) &&
+		defined( $self->{_lookup_create_key}->{ _q($database) }->{ _q($input) } )
+	);
+	return keys %{ $self->{_lookup_create_key}->{ _q($database) }->{ _q($input) } };
+}
+
+
+=head2 lookup_create_rules
+
+  my $source = $parser->lookup_create_rules($database, $input);
+
+=cut
+
+sub lookup_create_rules {
+	my $self = shift;
+	my ($database,$input) = @_;
+	$input = _input_name($input);
+	return unless (
+		defined( $self->{_lookup_create}->{ _q($database) } ) &&
+		defined( $self->{_lookup_create}->{ _q($database) }->{ _q($input) } )
+	);
+	return $self->{_lookup_create}->{ _q($database) }->{ _q($input) };
+}
+
+=head2 normalize_rules
+
+  my $source = $parser->normalize_rules($database, $input);
+
+=cut
+
+sub normalize_rules {
+	my $self = shift;
+	my ($database,$input) = @_;
+	$input = _input_name($input);
+	return unless (
+		defined( $self->{_normalize_source}->{ _q($database) } ) &&
+		defined( $self->{_normalize_source}->{ _q($database) }->{ _q($input) } )
+	);
+	return $self->{_normalize_source}->{ _q($database) }->{ _q($input) };
+}
+
+
+=head2 have_rules
+
+  my $do_marc = $parser->have_rules('marc', $database, $input);
+  my $do_index = $parser->have_rules('search', $database);
+
+This function will return hash containing count of all found C<marc_*> or
+C<search> directives. Input name is optional.
+
+=cut
+
+sub have_rules {
+	my $self = shift;
+
+	my $log = $self->_get_logger();
+	my $type = shift @_ || $log->logconfess("need at least type");
+	my $database = shift @_ || $log->logconfess("database is required");
+	my $input = shift @_;
+
+	$input = _input_name($input);
+
+
+	return unless defined( $self->{_have_rules}->{ _q($database) } );
+
+	my $database_rules = $self->{_have_rules}->{ _q($database ) };
+
+	if (defined($input)) {
+
+		return unless (
+			defined( $database_rules->{ _q($input) } ) &&
+			defined( $database_rules->{ _q($input) }->{ $type } )
+		);
+
+		return $database_rules->{ _q($input) }->{ $type };
+	}
+
+	my $usage;
+
+	foreach my $i (keys %{ $database_rules }) {
+		next unless defined( $database_rules->{$i}->{$type} );
+
+		foreach my $t (keys %{ $database_rules->{ $i }->{$type} }) {
+			$usage->{ $t } += $database_rules->{ $i }->{ $t };
+		}
+	}
+
+	return $usage;
+
+}
+
+
 =head1 PRIVATE
 
 =head2 _read_sources
@@ -141,11 +237,19 @@
 
 	my $nr = 0;
 
-	my @lookups;
+	my @sources;
+
+	my $lookup_src_cache;
+
+	my $only_database = $self->{only_database};
+	my $only_input = $self->{only_input};
 
 	$self->{config}->iterate_inputs( sub {
 		my ($input, $database) = @_;
 
+		return if ( $only_database && $database !~ m/$only_database/i );
+		return if ( $only_input && $input->{name} !~ m/$only_input/i );
+
 		$log->debug("database: $database input = ", dump($input));
 
 		foreach my $normalize (@{ $input->{normalize} }) {
@@ -164,8 +268,9 @@
 
 			$self->{valid_inputs}->{$database}->{$input_name}++;
 
-			push @lookups, sub {
-				$self->_parse_lookups( $database, $input_name, $full, $s );
+			push @sources, sub {
+				#warn "### $database $input_name, $full ###\n";
+				$self->_parse_source( $database, $input_name, $full, $s );
 			};
 
 			$nr++;
@@ -174,23 +279,23 @@
 
 	$log->debug("found $nr source files");
 
-	# parse all lookups
-	$_->() foreach (@lookups);
+	# parse all sources
+	$_->() foreach (@sources);
 
 	return $nr;
 }
 
-=head2 _parse_lookups
+=head2 _parse_source
 
-  $parser->_parse_lookups($database,$input,$path,$source);
+  $parser->_parse_source($database,$input,$path,$source);
 
-Called for each normalize source (rules) in each input by L</read_sources>
+Called for each normalize source (rules) in each input by L</_read_sources>
 
 It will report invalid databases and inputs in error log after parsing.
 
 =cut
 
-sub _parse_lookups {
+sub _parse_source {
 	my $self = shift;
 	my ($database, $input, $path, $source) = @_;
 
@@ -208,6 +313,7 @@
 	my $Document = PPI::Document->new( \$source ) || $log->logdie("can't parse source:\n", $self->{source});
 
 	$Document->prune('PPI::Token::Whitespace');
+	$Document->prune('PPI::Token::Comment');
 	#$Document->prune('PPI::Token::Operator');
 
 	# Find all the named subroutines
@@ -226,6 +332,13 @@
 			my ($Document,$Element) = @_;
 
 			$Element->isa('PPI::Token::Word') or return '';
+
+			if ( $Element->content eq 'sub' ) {
+				# repair demage done by prune of whitespace
+				$Element->insert_after( PPI::Token::Whitespace->new(' ') );
+				return '';
+			}
+
 			$Element->content eq 'lookup' or return '';
 
 			$log->debug("expansion: ", $Element->snext_sibling);
@@ -272,21 +385,18 @@
 
 			$log->debug("key = $key");
 
-			my $create = ' 
-				$coderef = ' . $e[7] . $e[8] . ';
-				foreach my $v ($coderef->()) {
-					next unless (defined($v) && $v ne \'\');
-					push @{ $lookup->{\'' . $key . '\'}->{$v} }, $mfn;
-				}
-			';
-
-			$log->debug("create: $create");
-
 			return $self->_lookup_error("invalid database $e[3] in $path" ) unless $self->valid_database( $e[3] );
 			return $self->_lookup_error("invalid input $e[5] of database $e[3] in $path", ) unless $self->valid_database_input( $e[3], $e[5] );
 
+			my $create = qq{
+				save_into_lookup($e[3],$e[5],'$key', $e[7] $e[8] );
+			};
+
+			$log->debug("create: $create");
+
 			# save code to create this lookup
-			$self->{_lookup_create}->{ _q($e[3]) }->{ _q($e[5]) }->{ _q($key) } .= $create;
+			$self->{_lookup_create}->{ _q($e[3]) }->{ _q($e[5]) } .= $create;
+			$self->{_lookup_create_key}->{ _q($e[3]) }->{ _q($e[5]) }->{ _q($key) }++;
 
 
 			if (defined( $self->{depends}->{ $database }->{ $input }->{ _q($e[3]) }->{ _q($e[5]) }->{ $key } )) {
@@ -303,7 +413,7 @@
 			}
 
 			$e[7]->remove;
-			$e[8]->insert_before( new PPI::Token::Quote::Single( "'$key'" ) );
+			$e[8]->insert_before( PPI::Token::Quote::Single->new( "'$key'" ) );
 			$e[8]->remove;
 
 
@@ -314,7 +424,7 @@
 	$log->debug("create: ", dump($self->{_lookup_create}) );
 	$log->debug("normalize: $normalize_source");
 
-	$self->{_normalize_source}->{$database}->{$input} = $normalize_source;
+	$self->{_normalize_source}->{$database}->{$input} .= $normalize_source;
 
 	if ($self->{debug}) {
 		my $Dumper = PPI::Dumper->new( $Document );
@@ -323,6 +433,19 @@
 
 	$log->error("Parser errors:\n", join("\n",@{ $self->{_lookup_errors} }) ) if ($self->{_lookup_errors});
 
+	$Document->find( sub {
+			my ($Document,$Element) = @_;
+
+			$Element->isa('PPI::Token::Word') or return '';
+			if ($Element->content =~ m/^(marc|search)/) {
+				my $what = $1;
+				$log->debug("found $what rules in $database/$input");
+				$self->{_have_rules}->{ $database }->{ $input }->{ $what }->{ $Element->content }++;
+			} else {
+				return '';
+			}
+	});
+
 	return 1;
 }