--- trunk/run.pl 2006/05/15 17:23:38 509 +++ trunk/run.pl 2006/07/04 11:36:03 580 @@ -4,21 +4,22 @@ use Cwd qw/abs_path/; use File::Temp qw/tempdir/; -use Data::Dumper; use lib './lib'; use WebPAC::Common 0.02; use WebPAC::Lookup; use WebPAC::Input 0.03; use WebPAC::Store 0.03; -use WebPAC::Normalize::XML; -use WebPAC::Normalize::Set; +use WebPAC::Normalize; use WebPAC::Output::TT; +use WebPAC::Validate; +use WebPAC::Output::MARC; use YAML qw/LoadFile/; use Getopt::Long; use File::Path; use Time::HiRes qw/time/; use File::Slurp; +use Data::Dump qw/dump/; =head1 NAME @@ -42,24 +43,43 @@ remove database and Hyper Estraier index before indexing -=item --only=database_name +=item --only=database_name/input_filter reindex just single database (legacy name is --one) +C is optional part which can be C +or C from input + =item --config conf/config.yml path to YAML configuration file -=item --force-set - -force conversion C<< normalize->path >> in C from -C<.xml> to C<.pl> - =item --stats disable indexing and dump statistics about field and subfield usage for each input +=item --validate path/to/validation_file + +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-marc-lint>. + +=item --marc-dump + +Force dump or input and marc record for debugging. + =back =cut @@ -70,31 +90,43 @@ my $clean = 0; my $config = 'conf/config.yml'; my $debug = 0; -my $only_db_name; -my $force_set = 0; +my $only_filter; my $stats = 0; +my $validate_path; +my ($marc_normalize, $marc_output); +my $marc_lint = 1; +my $marc_dump = 0; GetOptions( "limit=i" => \$limit, "offset=i" => \$offset, "clean" => \$clean, - "one=s" => \$only_db_name, - "only=s" => \$only_db_name, + "one=s" => \$only_filter, + "only=s" => \$only_filter, "config" => \$config, - "debug" => \$debug, - "force-set" => \$force_set, + "debug+" => \$debug, "stats" => \$stats, + "validate=s" => \$validate_path, + "marc-normalize=s" => \$marc_normalize, + "marc-output=s" => \$marc_output, + "marc-lint!" => \$marc_lint, + "marc-dump!" => \$marc_dump, ); $config = LoadFile($config); -print "config = ",Dumper($config) if ($debug); +print "config = ",dump($config) if ($debug); die "no databases in config file!\n" unless ($config->{databases}); my $log = _new WebPAC::Common()->_get_logger(); $log->info( "-" x 79 ); +my $validate; +$validate = new WebPAC::Validate( + path => $validate_path, +) if ($validate_path); + my $use_indexer = $config->{use_indexer} || 'hyperestraier'; if ($stats) { $log->debug("option --stats disables update of indexing engine..."); @@ -103,14 +135,19 @@ $log->info("using $use_indexer indexing engine..."); } +# disable indexing when creating marc +$use_indexer = undef if ($marc_normalize); + my $total_rows = 0; my $start_t = time(); -while (my ($database, $db_config) = each %{ $config->{databases} }) { +my @links; +my $indexer; - next if ($only_db_name && $database !~ m/$only_db_name/i); +while (my ($database, $db_config) = each %{ $config->{databases} }) { - my $indexer; + my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter); + next if ($only_database && $database !~ m/$only_database/i); if ($use_indexer) { my $indexer_config = $config->{$use_indexer} || $log->logdie("can't find '$use_indexer' part in confguration"); @@ -148,10 +185,10 @@ my $db_path = $config->{webpac}->{db_path} . '/' . $database; if ($clean) { - $log->info("creating new database $database in $db_path"); + $log->info("creating new database '$database' in $db_path"); rmtree( $db_path ) || $log->warn("can't remove $db_path: $!"); } else { - $log->debug("working on $database in $db_path"); + $log->info("working on database '$database' in $db_path"); } my $db = new WebPAC::Store( @@ -178,17 +215,21 @@ foreach my $input (@inputs) { + next if ($only_input && ($input->{name} !~ m#$only_input#i && $input->{type} !~ m#$only_input#i)); + my $type = lc($input->{type}); die "I know only how to handle input types ", join(",", @supported_inputs), " not '$type'!\n" unless (grep(/$type/, @supported_inputs)); my $lookup = new WebPAC::Lookup( lookup_file => $input->{lookup}, - ); + ) if ($input->{lookup}); my $input_module = $config->{webpac}->{inputs}->{$type}; - $log->info("working on input '$input->{path}' [$input->{type}] using $input_module lookup '$input->{lookup}'"); + $log->info("working on input '$input->{name}' in $input->{path} [type: $input->{type}] using $input_module", + $input->{lookup} ? "lookup '$input->{lookup}'" : "" + ); my $input_db = new WebPAC::Input( module => $input_module, @@ -204,67 +245,61 @@ my $maxmfn = $input_db->open( path => $input->{path}, code_page => $input->{encoding}, # database encoding + %{ $input }, ); - my $n = new WebPAC::Normalize::XML( - # filter => { 'foo' => sub { shift } }, - db => $db, - lookup_regex => $lookup->regex, - lookup => $lookup, - prefix => $input->{name}, - ); - - my $rules; - my $normalize_path = $input->{normalize}->{path}; + my @norm_array = ref($input->{normalize}) eq 'ARRAY' ? + @{ $input->{normalize} } : ( $input->{normalize} ); - if ($force_set) { - my $new_norm_path = $normalize_path; - $new_norm_path =~ s/\.xml$/.pl/; - if (-e $new_norm_path) { - $log->debug("--force-set replaced $normalize_path with $new_norm_path"); - $normalize_path = $new_norm_path; - } else { - $log->debug("--force-set failed on $new_norm_path, fallback to $normalize_path"); - } + if ($marc_normalize) { + @norm_array = ( { + path => $marc_normalize, + output => $marc_output || 'out/marc/' . $database . '-' . $input->{name} . '.marc', + } ); } - if ($normalize_path =~ m/\.xml$/i) { - $n->open( - tag => $input->{normalize}->{tag}, - xml_file => $normalize_path, - ); - } elsif ($normalize_path =~ m/\.(?:yml|yaml)$/i) { - $n->open_yaml( - path => $normalize_path, - tag => $input->{normalize}->{tag}, - ); - } elsif ($normalize_path =~ m/\.(?:pl)$/i) { - $n = undef; - $log->info("using WebPAC::Normalize::Set to process $normalize_path"); - $rules = read_file( $normalize_path ) or die "can't open $normalize_path: $!"; - } + foreach my $normalize (@norm_array) { - foreach my $pos ( 0 ... $input_db->size ) { + my $normalize_path = $normalize->{path} || $log->logdie("can't find normalize path in config"); - my $row = $input_db->fetch || next; + $log->logdie("Found '$normalize_path' as normalization file which isn't supported any more!") unless ( $normalize_path =~ m!\.pl$!i ); - my $mfn = $row->{'000'}->[0]; + my $rules = read_file( $normalize_path ) or die "can't open $normalize_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("Using $normalize_path for normalization..."); - - my $ds; - if ($n) { - $ds = $n->data_structure($row); - } else { - $ds = WebPAC::Normalize::Set::data_structure( + my $marc = new WebPAC::Output::MARC( + path => $normalize->{output}, + lint => $marc_lint, + dump => $marc_dump, + ) if ($normalize->{output}); + + # 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 (! $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_hash, + lookup => $lookup ? $lookup->lookup_hash : undef, + marc_encoding => 'utf-8', ); $db->save_ds( @@ -272,26 +307,46 @@ 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) { + my $i = 0; + + while (my $fields = WebPAC::Normalize::_get_marc_fields( fetch_next => 1 ) ) { + $marc->add( + id => $mfn . ( $i ? "/$i" : '' ), + fields => $fields, + leader => WebPAC::Normalize::marc_leader(), + row => $row, + ); + $i++; + } + + $log->info("Created $i instances of MFN $mfn\n") if ($i > 1); + } + + $total_rows++; } - $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 + $marc->finish if ($marc); - $log->info("statistics of fields usage:\n", $input_db->stats) if ($stats); + } - }; + } eval { $indexer->finish } if ($indexer && $indexer->can('finish')); my $dt = time() - $start_t; - $log->info("$total_rows records indexed in " . - sprintf("%.2f sec [%.2f rec/sec]", + $log->info("$total_rows records ", $indexer ? "indexed " : "", + sprintf("in %.2f sec [%.2f rec/sec]", $dt, ($total_rows / $dt) ) ); @@ -299,15 +354,15 @@ # # add Hyper Estraier links to other databases # - if (ref($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("adding link $database -> $link->{to} [$link->{credit}]"); - $indexer->add_link( + $log->info("saving link $database -> $link->{to} [$link->{credit}]"); + push @links, { from => $database, to => $link->{to}, credit => $link->{credit}, - ); + }; } else { $log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]"); } @@ -316,3 +371,7 @@ } +foreach my $link (@links) { + $log->info("adding link $link->{from} -> $link->{to} [$link->{credit}]"); + $indexer->add_link( %{ $link } ); +}