--- trunk/run.pl 2006/06/26 16:39:51 536 +++ trunk/run.pl 2006/07/01 20:29:09 556 @@ -4,7 +4,6 @@ use Cwd qw/abs_path/; use File::Temp qw/tempdir/; -use Data::Dumper; use lib './lib'; use WebPAC::Common 0.02; @@ -19,6 +18,9 @@ use File::Path; use Time::HiRes qw/time/; use File::Slurp; +use MARC::Record 2.0; # need 2.0 for utf-8 encoding see marcpm.sf.net +use MARC::Lint; +use Data::Dump qw/dump/; =head1 NAME @@ -62,6 +64,19 @@ turn on extra validation of imput records, see L +=item --marc-normalize conf/normalize/mapping.pl + +This option specifies normalisation file for MARC creation + +=item --marc-output out/marc/test.marc + +Optional path to output file + +=item --marc-lint + +By default turned on if C<--marc-normalize> is used. You can disable lint +messages with C<--no-arc-lint>. + =back =cut @@ -75,6 +90,8 @@ my $only_filter; my $stats = 0; my $validate_path; +my ($marc_normalize, $marc_output); +my $marc_lint = 1; GetOptions( "limit=i" => \$limit, @@ -86,11 +103,14 @@ "debug" => \$debug, "stats" => \$stats, "validate=s" => \$validate_path, + "marc-normalize=s" => \$marc_normalize, + "marc-output=s" => \$marc_output, + "marc-lint!" => \$marc_lint, ); $config = LoadFile($config); -print "config = ",Dumper($config) if ($debug); +print "config = ",dump($config) if ($debug); die "no databases in config file!\n" unless ($config->{databases}); @@ -110,12 +130,17 @@ $log->info("using $use_indexer indexing engine..."); } +# disable indexing when creating marc +$use_indexer = undef if ($marc_normalize); + my $total_rows = 0; my $start_t = time(); my @links; my $indexer; +my $lint = new MARC::Lint if ($marc_lint); + while (my ($database, $db_config) = each %{ $config->{databases} }) { my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter); @@ -220,56 +245,107 @@ %{ $input }, ); - my $rules; - my $normalize_path = $input->{normalize}->{path}; + my @norm_array = ref($input->{normalize}) eq 'ARRAY' ? + @{ $input->{normalize} } : ( $input->{normalize} ); + + if ($marc_normalize) { + @norm_array = ( { + path => $marc_normalize, + output => $marc_output || 'out/marc/' . $database . '-' . $input->{name} . '.marc', + } ); + } + + foreach my $normalize (@norm_array) { - $log->logdie("Found '$normalize_path' as normalization file which isn't supported any more!") unless ( $normalize_path =~ m!\.pl$!i ); + my $normalize_path = $normalize->{path} || $log->logdie("can't find normalize path in config"); - my $rules = read_file( $normalize_path ) or die "can't open $normalize_path: $!"; + $log->logdie("Found '$normalize_path' as normalization file which isn't supported any more!") unless ( $normalize_path =~ m!\.pl$!i ); - foreach my $pos ( 0 ... $input_db->size ) { + my $rules = read_file( $normalize_path ) or die "can't open $normalize_path: $!"; - my $row = $input_db->fetch || next; + $log->info("Using $normalize_path for normalization..."); - my $mfn = $row->{'000'}->[0]; + my $marc_fh; + if (my $path = $normalize->{output}) { + open($marc_fh, '>', $path) || + $log->logdie("can't open MARC output $path: $!"); - if (! $mfn || $mfn !~ m#^\d+$#) { - $log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos"); - $mfn = $pos; - push @{ $row->{'000'} }, $pos; + $log->info("Creating MARC export file $path", $marc_lint ? ' (with lint)' : '', "\n"); } + # reset position in database + $input_db->seek(1); + + foreach my $pos ( 0 ... $input_db->size ) { + + my $row = $input_db->fetch || next; + + my $mfn = $row->{'000'}->[0]; - if ($validate) { - my @errors = $validate->validate_errors( $row ); - $log->error( "MFN $mfn validation errors:\n", join("\n", @errors) ) if (@errors); + if (! $mfn || $mfn !~ m#^\d+$#) { + $log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos"); + $mfn = $pos; + push @{ $row->{'000'} }, $pos; + } + + + if ($validate) { + my @errors = $validate->validate_errors( $row ); + $log->error( "MFN $mfn validation errors:\n", join("\n", @errors) ) if (@errors); + } + + + my $ds = WebPAC::Normalize::data_structure( + row => $row, + rules => $rules, + lookup => $lookup ? $lookup->lookup_hash : undef, + marc_encoding => 'utf-8', + ); + + $db->save_ds( + id => $mfn, + ds => $ds, + prefix => $input->{name}, + ) if ($ds && !$stats); + + $indexer->add( + id => $input->{name} . "/" . $mfn, + ds => $ds, + type => $config->{$use_indexer}->{type}, + ) if ($indexer && $ds); + + if ($marc_fh) { + my $marc = new MARC::Record; + $marc->encoding( 'utf-8' ); + my @marc_fields = WebPAC::Normalize::_get_marc_fields(); + if (! @marc_fields) { + $log->warn("MARC record $mfn is empty, skipping"); + } else { + $marc->add_fields( @marc_fields ); + if ($marc_lint) { + $lint->check_record( $marc ); + my $err = join( "\n", $lint->warnings ); + $log->error("MARC lint detected warning on MFN $mfn\n", + "Original imput row: ",dump($row), "\n", + "Normalized MARC row: ",dump(@marc_fields), "\n", + "MARC lint warnings: ",$err,"\n" + ) if ($err); + } + print $marc_fh $marc->as_usmarc; + } + } + + $total_rows++; } - - my $ds = WebPAC::Normalize::data_structure( - row => $row, - rules => $rules, - lookup => $lookup ? $lookup->lookup_hash : undef, - ); - - $db->save_ds( - id => $mfn, - ds => $ds, - prefix => $input->{name}, - ) if ($ds && !$stats); - - $indexer->add( - id => $input->{name} . "/" . $mfn, - ds => $ds, - type => $config->{$use_indexer}->{type}, - ) if ($indexer); + $log->info("statistics of fields usage:\n", $input_db->stats) if ($stats); - $total_rows++; - } + # close MARC file + close($marc_fh) if ($marc_fh); - $log->info("statistics of fields usage:\n", $input_db->stats) if ($stats); + } - }; + } eval { $indexer->finish } if ($indexer && $indexer->can('finish')); @@ -283,7 +359,7 @@ # # add Hyper Estraier links to other databases # - if (ref($db_config->{links}) eq 'ARRAY') { + if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) { foreach my $link (@{ $db_config->{links} }) { if ($use_indexer eq 'hyperestraier') { $log->info("saving link $database -> $link->{to} [$link->{credit}]");