1 package Koha::SearchEngine::Elasticsearch::Indexer;
3 # Copyright 2013 Catalyst IT
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>.
23 use List::Util qw(any);
24 use base qw(Koha::SearchEngine::Elasticsearch);
28 use Koha::SearchEngine::Zebra::Indexer;
34 Koha::SearchEngine::Elasticsearch::Indexer - handles adding new records to the index
38 my $indexer = Koha::SearchEngine::Elasticsearch::Indexer->new(
39 { index => Koha::SearchEngine::BIBLIOS_INDEX } );
40 $indexer->drop_index();
41 $indexer->update_index(\@biblionumbers, \@records);
48 =item C<Koha::SearchEngine::Elasticsearch::Indexer::INDEX_STATUS_OK>
50 Represents an index state where index is created and in a working state.
52 =item C<Koha::SearchEngine::Elasticsearch::Indexer::INDEX_STATUS_REINDEX_REQUIRED>
54 Not currently used, but could be useful later, for example if can detect when new field or mapping added.
56 =item C<Koha::SearchEngine::Elasticsearch::Indexer::INDEX_STATUS_RECREATE_REQUIRED>
58 Representings an index state where index needs to be recreated and is not in a working state.
66 INDEX_STATUS_REINDEX_REQUIRED => 1,
67 INDEX_STATUS_RECREATE_REQUIRED => 2,
72 =head2 update_index($biblionums, $records)
75 $self->update_index($biblionums, $records);
77 die("Something went wrong trying to update index:" . $_[0]);
80 Converts C<MARC::Records> C<$records> to Elasticsearch documents and performs
81 an update request for these records on the Elasticsearch index.
87 Arrayref of biblio numbers for the C<$records>, the order must be the same as
88 and match up with C<$records>.
92 Arrayref of C<MARC::Record>s.
99 my ($self, $biblionums, $records) = @_;
101 my $documents = $self->marc_records_to_documents($records);
104 for (my $i = 0; $i < scalar @$biblionums; $i++) {
105 my $id = $biblionums->[$i];
106 my $document = $documents->[$i];
112 push @body, $document;
116 my $elasticsearch = $self->get_elasticsearch();
117 $response = $elasticsearch->bulk(
118 index => $self->index_name,
119 type => 'data', # is just hard coded in Indexer.pm?
126 =head2 set_index_status_ok
128 Convenience method for setting index status to C<INDEX_STATUS_OK>.
132 sub set_index_status_ok {
134 $self->index_status(INDEX_STATUS_OK);
137 =head2 is_index_status_ok
139 Convenience method for checking if index status is C<INDEX_STATUS_OK>.
143 sub is_index_status_ok {
145 return $self->index_status == INDEX_STATUS_OK;
148 =head2 set_index_status_reindex_required
150 Convenience method for setting index status to C<INDEX_REINDEX_REQUIRED>.
154 sub set_index_status_reindex_required {
156 $self->index_status(INDEX_STATUS_REINDEX_REQUIRED);
159 =head2 is_index_status_reindex_required
161 Convenience method for checking if index status is C<INDEX_STATUS_REINDEX_REQUIRED>.
165 sub is_index_status_reindex_required {
167 return $self->index_status == INDEX_STATUS_REINDEX_REQUIRED;
170 =head2 set_index_status_recreate_required
172 Convenience method for setting index status to C<INDEX_STATUS_RECREATE_REQUIRED>.
176 sub set_index_status_recreate_required {
178 $self->index_status(INDEX_STATUS_RECREATE_REQUIRED);
181 =head2 is_index_status_recreate_required
183 Convenience method for checking if index status is C<INDEX_STATUS_RECREATE_REQUIRED>.
187 sub is_index_status_recreate_required {
189 return $self->index_status == INDEX_STATUS_RECREATE_REQUIRED;
192 =head2 index_status($status)
194 Will either set the current index status to C<$status> and return C<$status>,
195 or return the current index status if called with no arguments.
201 Optional argument. If passed will set current index status to C<$status> if C<$status> is
202 a valid status. See L</CONSTANTS>.
209 my ($self, $status) = @_;
210 my $key = 'ElasticsearchIndexStatus_' . $self->index;
212 if (defined $status) {
213 unless (any { $status == $_ } (
215 INDEX_STATUS_REINDEX_REQUIRED,
216 INDEX_STATUS_RECREATE_REQUIRED,
219 Koha::Exceptions::Exception->throw("Invalid index status: $status");
221 C4::Context->set_preference($key, $status);
225 return C4::Context->preference($key);
229 =head2 update_mappings
231 Generate Elasticsearch mappings from mappings stored in database and
232 perform a request to update Elasticsearch index mappings. Will throw an
233 error and set index status to C<INDEX_STATUS_RECREATE_REQUIRED> if update
238 sub update_mappings {
240 my $elasticsearch = $self->get_elasticsearch();
241 my $mappings = $self->get_elasticsearch_mappings();
243 foreach my $type (keys %{$mappings}) {
245 my $response = $elasticsearch->indices->put_mapping(
246 index => $self->index_name,
249 $type => $mappings->{$type}
253 $self->set_index_status_recreate_required();
254 my $reason = $_[0]->{vars}->{body}->{error}->{reason};
255 my $index_name = $self->index_name;
256 Koha::Exceptions::Exception->throw(
257 error => "Unable to update mappings for index \"$index_name\". Reason was: \"$reason\". Index needs to be recreated and reindexed",
261 $self->set_index_status_ok();
264 =head2 update_index_background($biblionums, $records)
266 This has exactly the same API as C<update_index> however it'll
267 return immediately. It'll start a background process that does the adding.
269 If it fails to add to Elasticsearch then it'll add to a queue that will cause
270 it to be updated by a regular index cron job in the future.
274 # TODO implement in the future - I don't know the best way of doing this yet.
275 # If fork: make sure process group is changed so apache doesn't wait for us.
277 sub update_index_background {
279 $self->update_index(@_);
284 This function takes an array of record numbers and fetches the records to send to update_index
287 If $records parameter is provided the records will be used as-is, this is only utilized for authorities
290 The other variables are used for parity with Zebra indexing calls. Currently the calls are passed through
296 my ( $self, $record_numbers, $op, $server, $records ) = @_;
297 $record_numbers = [$record_numbers] if ref $record_numbers ne 'ARRAY' && defined $record_numbers;
298 $records = [$records] if ref $records ne 'ARRAY' && defined $records;
299 if ( $op eq 'specialUpdate' ) {
300 my $index_record_numbers;
302 foreach my $record_number ( @$record_numbers ){
303 my $record = _get_record( $record_number, $server );
305 push @$records, $record;
306 push @$index_record_numbers, $record_number;
310 $self->update_index_background( $index_record_numbers, $records ) if $index_record_numbers && $records;
312 elsif ( $op eq 'recordDelete' ) {
313 $self->delete_index_background( $record_numbers );
315 #FIXME Current behaviour is to index Zebra when using ES, at some point we should stop
316 Koha::SearchEngine::Zebra::Indexer::index_records( $self, $record_numbers, $op, $server, undef );
320 my ( $id, $server ) = @_;
321 return $server eq 'biblioserver'
322 ? C4::Biblio::GetMarcBiblio({ biblionumber => $id, embed_items => 1 })
323 : C4::AuthoritiesMarc::GetAuthority($id);
326 =head2 delete_index($biblionums)
328 C<$biblionums> is an arrayref of biblionumbers to delete from the index.
333 my ($self, $biblionums) = @_;
335 my $elasticsearch = $self->get_elasticsearch();
336 my @body = map { { delete => { _id => "$_" } } } @{$biblionums};
337 my $result = $elasticsearch->bulk(
338 index => $self->index_name,
342 if ($result->{errors}) {
343 croak "An Elasticsearch error occurred during bulk delete";
347 =head2 delete_index_background($biblionums)
349 Identical to L</delete_index($biblionums)>
353 # TODO: Should be made async
354 sub delete_index_background {
356 $self->delete_index(@_);
361 Drops the index from the Elasticsearch server.
367 if ($self->index_exists) {
368 my $elasticsearch = $self->get_elasticsearch();
369 $elasticsearch->indices->delete(index => $self->index_name);
370 $self->set_index_status_recreate_required();
376 Creates the index (including mappings) on the Elasticsearch server.
382 my $settings = $self->get_elasticsearch_settings();
383 my $elasticsearch = $self->get_elasticsearch();
384 $elasticsearch->indices->create(
385 index => $self->index_name,
387 settings => $settings
390 $self->update_mappings();
395 Checks if index has been created on the Elasticsearch server. Returns C<1> or the
396 empty string to indicate whether index exists or not.
402 my $elasticsearch = $self->get_elasticsearch();
403 return $elasticsearch->indices->exists(
404 index => $self->index_name,
416 =item Chris Cormack C<< <chrisc@catalyst.net.nz> >>
418 =item Robin Sheat C<< <robin@catalyst.net.nz> >>