1 package Koha::MetaSearcher;
3 # Copyright 2014 ByWater Solutions
5 # This file is part of Koha.
7 # Koha is free software; you can redistribute it and/or modify it
8 # under the terms of the GNU General Public License as published by
9 # the Free Software Foundation; either version 3 of the License, or
10 # (at your option) any later version.
12 # Koha is distributed in the hope that it will be useful, but
13 # WITHOUT ANY WARRANTY; without even the implied warranty of
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 # GNU General Public License for more details.
17 # You should have received a copy of the GNU General Public License
18 # along with Koha; if not, see <http://www.gnu.org/licenses>.
22 use base 'Class::Accessor';
24 use C4::Charset qw( MarcToUTF8Record SetUTF8Flag );
25 use C4::Search qw( new_record_from_zebra );
26 use DBIx::Class::ResultClass::HashRefInflator;
30 use Koha::MetadataRecord;
32 use Storable qw( fd_retrieve store_fd );
33 use Time::HiRes qw( clock_gettime CLOCK_MONOTONIC );
39 __PACKAGE__->mk_accessors( qw( fetch offset on_error resultset ) );
42 my ( $class, $options ) = @_;
44 my ( $uuid, $uuidstring );
45 UUID::generate($uuid);
46 UUID::unparse( $uuid, $uuidstring );
53 resultset => $uuidstring,
59 my ( $self, $index, $server, $marcrecord ) = @_;
61 my @kohafields = ('biblio.title','biblio.subtitle','biblio.seriestitle','biblio.author',
62 'biblioitems.isbn','biblioitems.issn','biblioitems.lccn','biblioitems.editionstatement',
63 'biblio.copyrightdate','biblioitems.publicationyear');
64 my $metadata = C4::Biblio::TransformMarcToKoha({ kohafields => \@kohafields, record => $marcrecord});
65 $metadata->{edition} = delete $metadata->{editionstatement};
66 $metadata->{date} = delete $metadata->{copyrightdate};
67 $metadata->{date} //= delete $metadata->{publicationyear};
69 push @{ $self->{results} }, {
72 record => $marcrecord,
73 metadata => $metadata,
78 my ( $self, $server_ids, $query ) = @_;
80 my $resultset_expiry = 300;
82 my $cache = Koha::Caches->get_instance();
83 my $schema = Koha::Database->new->schema;
86 map { $_ => 0 } @$server_ids
89 map { $_ => 0 } @$server_ids
94 my $start = clock_gettime( CLOCK_MONOTONIC );
95 my $select = IO::Select->new;
100 foreach my $server_id ( @$server_ids ) {
101 if ( $server_id =~ /^\d+$/ ) {
103 my $server = $schema->resultset('Z3950server')->find(
104 { id => $server_id },
105 { result_class => 'DBIx::Class::ResultClass::HashRefInflator' },
107 $server->{type} = 'z3950';
109 push @servers, $server;
110 } elsif ( $server_id =~ /(\w+)(?::(\w+))?/ ) {
117 servername => $server_id,
122 # HashRefInflator is used so that the information will survive into the fork
123 foreach my $server ( @servers ) {
125 my $set = $cache->get_from_cache( 'z3950-resultset-' . $self->resultset . '-' . $server->{id} );
126 if ( ref($set) eq 'HASH' ) {
127 $set->{server} = $server;
128 push @cached_sets, $set;
133 $select->add( $self->_start_worker( $server, $query ) );
136 # Handle these while the servers are searching
137 foreach my $set ( @cached_sets ) {
138 $self->_handle_hits( $stats, $set );
141 while ( $select->count ) {
142 foreach my $readfh ( $select->can_read() ) {
143 my $result = fd_retrieve( $readfh );
145 $select->remove( $readfh );
149 next if ( ref $result ne 'HASH' );
151 if ( $result->{error} ) {
152 $self->{on_error}->( $result->{server}, $result->{error} );
156 $self->_handle_hits( $stats, $result );
159 $cache->set_in_cache( 'z3950-resultset-' . $self->resultset . '-' . $result->{server}->{id}, {
160 hits => $result->{hits},
161 num_fetched => $result->{num_fetched},
162 num_hits => $result->{num_hits},
163 }, { expiry => $resultset_expiry } );
168 $stats->{time} = clock_gettime( CLOCK_MONOTONIC ) - $start;
174 my ( $self, $server, $query ) = @_;
175 pipe my $readfh, my $writefh;
177 # Accessing the cache or Koha database after the fork is risky, so get any resources we need
180 my $marcflavour = C4::Context->preference('marcflavour');
182 if ( ( $pid = fork ) ) {
187 } elsif ( !defined $pid ) {
190 $self->{on_error}->( $server, 'Failed to fork' );
196 my ( $num_hits, $num_fetched, $hits, $results );
199 if ( $server->{type} eq 'z3950' ) {
200 my $zoptions = ZOOM::Options->new();
201 $zoptions->option( 'elementSetName', 'F' );
202 $zoptions->option( 'databaseName', $server->{db} );
203 $zoptions->option( 'user', $server->{userid} ) if $server->{userid};
204 $zoptions->option( 'password', $server->{password} ) if $server->{password};
205 $zoptions->option( 'preferredRecordSyntax', $server->{syntax} );
206 $zoptions->option( 'timeout', $server->{timeout} ) if $server->{timeout};
208 $connection = ZOOM::Connection->create($zoptions);
210 $connection->connect( $server->{host}, $server->{port} );
211 $results = $connection->search_pqf( $query ); # Starts the search
212 } elsif ( $server->{type} eq 'koha' ) {
213 $connection = C4::Context->Zconn( $server->{extra} );
214 $results = $connection->search_pqf( $query ); # Starts the search
215 } elsif ( $server->{type} eq 'batch' ) {
216 $server->{encoding} = 'utf-8';
221 error => $connection ? $connection->exception() : $@,
227 if ( $server->{type} eq 'batch' ) {
228 # TODO: actually handle PQF
229 $query =~ s/@\w+ (?:\d+=\d+ )?//g;
232 my $schema = Koha::Database->new->schema;
233 $schema->storage->debug(1);
234 my $match_condition = [ map +{ -like => '%' . $_ . '%' }, split( /\s+/, $query ) ];
235 $hits = [ $schema->resultset('ImportRecord')->search(
237 import_batch_id => $server->{extra},
239 { 'import_biblios.title' => $match_condition },
240 { 'import_biblios.author' => $match_condition },
241 { 'import_biblios.isbn' => $match_condition },
242 { 'import_biblios.issn' => $match_condition },
246 join => [ qw( import_biblios ) ],
247 rows => $self->{fetch},
249 )->get_column( 'marc' )->all ];
251 $num_hits = $num_fetched = scalar @$hits;
253 $num_hits = $results->size;
254 $num_fetched = ( $self->{offset} + $self->{fetch} ) < $num_hits ? $self->{fetch} : $num_hits;
256 $hits = [ map { $_->raw() } @{ $results->records( $self->{offset}, $num_fetched, 1 ) } ];
259 if ( !@$hits && $connection && $connection->exception() ) {
261 error => $connection->exception(),
267 if ( $server->{type} eq 'koha' ) {
268 $hits = [ map { C4::Search::new_record_from_zebra( $server->{extra}, $_ ) } @$hits ];
270 $hits = [ map { $self->_import_record( $_, $marcflavour, $server->{encoding} ? $server->{encoding} : "iso-5426" ) } @$hits ];
275 num_fetched => $num_fetched,
276 num_hits => $num_hits,
284 my ( $self, $raw, $marcflavour, $encoding ) = @_;
286 my ( $marcrecord ) = MarcToUTF8Record( $raw, $marcflavour, $encoding ); #ignores charset return values
288 SetUTF8Flag($marcrecord);
293 my ( $self, $stats, $set ) = @_;
295 my $server = $set->{server};
297 my $num_hits = $stats->{num_hits}->{ $server->{id} } = $set->{num_hits};
298 my $num_fetched = $stats->{num_fetched}->{ $server->{id} } = $set->{num_fetched};
300 $stats->{total_hits} += $num_hits;
301 $stats->{total_fetched} += $num_fetched;
303 foreach my $j ( 0..$#{ $set->{hits} } ) {
304 $self->handle_hit( $self->{offset} + $j, $server, $set->{hits}->[$j] );
309 my ( $self, $key, $direction ) = @_;
311 my $empty_flip = -1; # Determines the flip of ordering for records with empty sort keys.
313 foreach my $hit ( @{ $self->{results} } ) {
314 ( $hit->{sort_key} = $hit->{metadata}->{$key} || '' ) =~ s/\W//g;
317 $self->{results} = [ sort {
318 # Sort empty records at the end
319 return -$empty_flip unless $a->{sort_key};
320 return $empty_flip unless $b->{sort_key};
322 $direction * ( $a->{sort_key} cmp $b->{sort_key} );
323 } @{ $self->{results} } ];
327 my ( $self, $offset, $length ) = @_;
331 foreach my $i ( $offset..( $offset + $length - 1 ) ) {
332 push @subset, $self->{results}->[$i] if $self->{results}->[$i];