--- trunk/lib/WebPAC/Input.pm	2007/02/04 15:09:01	799
+++ trunk/lib/WebPAC/Input.pm	2007/05/27 19:10:43	860
@@ -16,11 +16,11 @@
 
 =head1 VERSION
 
-Version 0.17
+Version 0.18
 
 =cut
 
-our $VERSION = '0.17';
+our $VERSION = '0.18';
 
 =head1 SYNOPSIS
 
@@ -286,7 +286,9 @@
 #				return unless defined($l);
 #				return $l unless ($rec_regex && $f_nr);
 
-				warn "-=> $f_nr ## $l\n" if ($debug);
+				return unless ( defined($l) && defined($f_nr) );
+
+				warn "-=> $f_nr ## |$l|\n" if ($debug);
 				$log->debug("-=> $f_nr ## $l");
 
 				# codepage conversion and recode_regex
@@ -299,16 +301,21 @@
 					my $c = 0;
 					foreach my $r (@{ $rec_regex->{$f_nr} }) {
 						my $old_l = $l;
-						eval '$l =~ ' . $r;
+						$log->logconfess("expected regex in ", dump( $r )) unless defined($r->{regex});
+						eval '$l =~ ' . $r->{regex};
 						if ($old_l ne $l) {
-							$log->debug("REGEX on $f_nr eval \$l =~ $r\n## old l: [$old_l]\n## new l: [$l]");
-							warn "*** $r |$old_l| -> |$l|\n" if ($debug);
+							my $d = "|$old_l| -> |$l| "; # . $r->{regex};
+							$d .= ' +' . $r->{line} . ' ' . $r->{file} if defined($r->{line});
+							$d .= ' ' . $r->{debug} if defined($r->{debug});
+							$log->debug("MODIFY $d");
+							warn "*** $d\n" if ($debug);
+
 						}
 						$log->error("error applying regex: $r") if ($@);
 					}
 				}
 
-				$log->debug("<=- $f_nr ## $l");
+				$log->debug("<=- $f_nr ## |$l|");
 				warn "<=- $f_nr ## $l\n" if ($debug);
 				return $l;
 		});
@@ -512,7 +519,8 @@
 
 	my $out = join("\n",
 		map {
-			my $f = $_ || die "no field";
+			my $f = $_;
+			die "no field in ", dump( $s->{fld} ) unless defined( $f );
 			my $v = $s->{fld}->{$f} || die "no s->{fld}->{$f}";
 			$max_fld = $v if ($v > $max_fld);
 
@@ -532,7 +540,7 @@
 			}
 
 			$o;
-		} sort { $a cmp $b } keys %{ $s->{fld} }
+		} sort { $a <=> $b } keys %{ $s->{fld} }
 	);
 
 	$log->debug( sub { dump($s) } );
@@ -551,21 +559,22 @@
 
 	return unless $self->{ll_db};
 
-	if ($self->{ll_db}->can('dump_rec')) {
+	if ($self->{ll_db}->can('dump_ascii')) {
 		return $self->{ll_db}->dump_ascii( $self->{pos} );
 	} else {
 		return dump( $self->{ll_db}->fetch_rec( $self->{pos} ) );
 	}
 }
 
-=head2 modify_record_regexps
+=head2 _get_regex
 
-Generate hash with regexpes to be applied using L<filter>.
+Helper function called which create regexps to be execute on code.
 
-  my $regexpes = $input->modify_record_regexps(
-		900 => { '^a' => { ' : ' => '^b' } },
-		901 => { '*' => { '^b' => ' ; ' } },
-  );
+  _get_regex( 900, 'regex:[0-9]+' ,'numbers' );
+  _get_regex( 900, '^b', ' : ^b' );
+
+It supports perl regexps with C<regex:> prefix to from value and has
+additional logic to skip empty subfields.
 
 =cut
 
@@ -582,14 +591,30 @@
 		$from = '\Q' . $from . '\E';
 	}
 	if ($sf =~ /^\^/) {
+		my $need_subfield_data = '*';	# no
+		# if from is also subfield, require some data in between
+		# to correctly skip empty subfields
+		$need_subfield_data = '+' if ($from =~ m/^\\Q\^/);
 		return
-			's/\Q'. $sf .'\E([^\^]*?)'. $from .'([^\^]*?)/'. $sf .'$1'. $to .'$2/';
+			's/\Q'. $sf .'\E([^\^]' . $need_subfield_data . '?)'. $from .'([^\^]*?)/'. $sf .'$1'. $to .'$2/';
 	} else {
 		return
 			's/'. $from .'/'. $to .'/g';
 	}
 }
 
+
+=head2 modify_record_regexps
+
+Generate hash with regexpes to be applied using L<filter>.
+
+  my $regexpes = $input->modify_record_regexps(
+		900 => { '^a' => { ' : ' => '^b' } },
+		901 => { '*' => { '^b' => ' ; ' } },
+  );
+
+=cut
+
 sub modify_record_regexps {
 	my $self = shift;
 	my $modify_record = {@_};
@@ -607,10 +632,11 @@
 			foreach my $from (keys %{ $modify_record->{$f}->{$sf} }) {
 				my $to = $modify_record->{$f}->{$sf}->{$from};
 				#die "no field?" unless defined($to);
-				$log->debug("transform: |$from| -> |$to|");
+				my $d = "|$from| -> |$to|";
+				$log->debug("transform: $d");
 
 				my $regex = _get_regex($sf,$from,$to);
-				push @{ $regexpes->{$f} }, $regex;
+				push @{ $regexpes->{$f} }, { regex => $regex, debug => $d };
 				$log->debug("regex: $regex");
 			}
 		}
@@ -671,7 +697,11 @@
 			$log->debug("transform: |$from| -> |$to|");
 
 			my $regex = _get_regex($sf,$from,$to);
-			push @{ $regexpes->{$f} }, $regex;
+			push @{ $regexpes->{$f} }, {
+				regex => $regex,
+				file => $modify_path,
+				line => $.,
+			};
 			$log->debug("regex: $regex");
 		}
 	}