--- trunk/run.pl 2005/12/18 21:06:39 285 +++ trunk/run.pl 2006/05/15 17:23:38 509 @@ -9,40 +9,135 @@ use WebPAC::Common 0.02; use WebPAC::Lookup; -use WebPAC::Input::ISIS 0.02; +use WebPAC::Input 0.03; use WebPAC::Store 0.03; use WebPAC::Normalize::XML; +use WebPAC::Normalize::Set; use WebPAC::Output::TT; -use WebPAC::Output::Estraier 0.05; use YAML qw/LoadFile/; -use LWP::Simple; +use Getopt::Long; +use File::Path; +use Time::HiRes qw/time/; +use File::Slurp; -my $limit = shift @ARGV; +=head1 NAME -my $config = LoadFile('conf/config.yml'); +run.pl - start WebPAC indexing -print "config = ",Dumper($config); +B + +Options: + +=over 4 + +=item --offset 42 + +start loading (all) databases at offset 42 + +=item --limit 100 + +limit loading to 100 records + +=item --clean + +remove database and Hyper Estraier index before indexing + +=item --only=database_name + +reindex just single database (legacy name is --one) + +=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 + +=back + +=cut + +my $offset; +my $limit; + +my $clean = 0; +my $config = 'conf/config.yml'; +my $debug = 0; +my $only_db_name; +my $force_set = 0; +my $stats = 0; + +GetOptions( + "limit=i" => \$limit, + "offset=i" => \$offset, + "clean" => \$clean, + "one=s" => \$only_db_name, + "only=s" => \$only_db_name, + "config" => \$config, + "debug" => \$debug, + "force-set" => \$force_set, + "stats" => \$stats, +); + +$config = LoadFile($config); + +print "config = ",Dumper($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 $use_indexer = $config->{use_indexer} || 'hyperestraier'; +if ($stats) { + $log->debug("option --stats disables update of indexing engine..."); + $use_indexer = undef; +} else { + $log->info("using $use_indexer indexing engine..."); +} + my $total_rows = 0; +my $start_t = time(); while (my ($database, $db_config) = each %{ $config->{databases} }) { - my $log = _new WebPAC::Common()->_get_logger(); + next if ($only_db_name && $database !~ m/$only_db_name/i); - # - # open Hyper Estraier database - # + my $indexer; - my $est_config = $config->{hyperestraier} || $log->logdie("can't find 'hyperestraier' part in confguration"); - $est_config->{database} = $database; + if ($use_indexer) { + my $indexer_config = $config->{$use_indexer} || $log->logdie("can't find '$use_indexer' part in confguration"); + $indexer_config->{database} = $database; + $indexer_config->{clean} = $clean; + $indexer_config->{label} = $db_config->{name}; + + if ($use_indexer eq 'hyperestraier') { + + # open Hyper Estraier database + use WebPAC::Output::Estraier '0.10'; + $indexer = new WebPAC::Output::Estraier( %{ $indexer_config } ); + + } elsif ($use_indexer eq 'kinosearch') { + + # open KinoSearch + use WebPAC::Output::KinoSearch; + $indexer_config->{clean} = 1 unless (-e $indexer_config->{index_path}); + $indexer = new WebPAC::Output::KinoSearch( %{ $indexer_config } ); - $log->info("using HyperEstraier URL $est_config->{masterurl}"); + } else { + $log->logdie("unknown use_indexer: $use_indexer"); + } + + $log->logide("can't continue without valid indexer") unless ($indexer); + } - my $est = new WebPAC::Output::Estraier( - %{ $est_config }, - ); # # now WebPAC::Store @@ -52,12 +147,17 @@ my $db_path = $config->{webpac}->{db_path} . '/' . $database; - $log->info("working on $database in $db_path"); + if ($clean) { + $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"); + } my $db = new WebPAC::Store( path => $db_path, database => $database, - debug => 1, + debug => $debug, ); @@ -74,25 +174,34 @@ $log->info("database $database doesn't have inputs defined"); } + my @supported_inputs = keys %{ $config->{webpac}->{inputs} }; + foreach my $input (@inputs) { my $type = lc($input->{type}); - die "I know only how to handle input type isis, not '$type'!\n" unless ($type eq 'isis'); + 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}, ); - $log->info("working on input $input->{path} [$input->{type}]"); + my $input_module = $config->{webpac}->{inputs}->{$type}; + + $log->info("working on input '$input->{path}' [$input->{type}] using $input_module lookup '$input->{lookup}'"); - my $isis = new WebPAC::Input::ISIS( + my $input_db = new WebPAC::Input( + module => $input_module, code_page => $config->{webpac}->{webpac_encoding}, - limit_mfn => $input->{limit}, + limit => $limit || $input->{limit}, + offset => $offset, lookup => $lookup, + recode => $input->{recode}, + stats => $stats, ); + $log->logdie("can't create input using $input_module") unless ($input); - my $maxmfn = $isis->open( + my $maxmfn = $input_db->open( path => $input->{path}, code_page => $input->{encoding}, # database encoding ); @@ -105,52 +214,103 @@ prefix => $input->{name}, ); + my $rules; my $normalize_path = $input->{normalize}->{path}; + if ($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 $normalize_path"); + } + } + if ($normalize_path =~ m/\.xml$/i) { $n->open( tag => $input->{normalize}->{tag}, - xml_file => $input->{normalize}->{path}, + 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: $!"; } - for ( 0 ... $isis->size ) { + foreach my $pos ( 0 ... $input_db->size ) { - my $row = $isis->fetch || next; + my $row = $input_db->fetch || next; - my $mfn = $row->{'000'}->[0] || die "can't find MFN"; + my $mfn = $row->{'000'}->[0]; - my $ds = $n->data_structure($row); + if (! $mfn || $mfn !~ m#^\d+$#) { + $log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos"); + $mfn = $pos; + push @{ $row->{'000'} }, $pos; + } + + + my $ds; + if ($n) { + $ds = $n->data_structure($row); + } else { + $ds = WebPAC::Normalize::Set::data_structure( + row => $row, + rules => $rules, + lookup => $lookup->lookup_hash, + ); + + $db->save_ds( + id => $mfn, + ds => $ds, + prefix => $input->{name}, + ) if ($ds && !$stats); + } - $est->add( - id => $input->{name} . "#" . $mfn, + $indexer->add( + id => $input->{name} . "/" . $mfn, ds => $ds, - type => $config->{hyperestraier}->{type}, - ); + type => $config->{$use_indexer}->{type}, + ) if ($indexer); $total_rows++; } + $log->info("statistics of fields usage:\n", $input_db->stats) if ($stats); + }; - $log->info("$total_rows records indexed"); + 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]", + $dt, ($total_rows / $dt) + ) + ); # # add Hyper Estraier links to other databases # if (ref($db_config->{links}) eq 'ARRAY') { foreach my $link (@{ $db_config->{links} }) { - $log->info("adding link $database -> $link->{to} [$link->{credit}]"); - $est->add_link( - from => $database, - to => $link->{to}, - credit => $link->{credit}, - ); + if ($use_indexer eq 'hyperestraier') { + $log->info("adding link $database -> $link->{to} [$link->{credit}]"); + $indexer->add_link( + from => $database, + to => $link->{to}, + credit => $link->{credit}, + ); + } else { + $log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]"); + } } }