3 # This file is part of Koha.
5 # Koha is free software; you can redistribute it and/or modify it
6 # under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3 of the License, or
8 # (at your option) any later version.
10 # Koha is distributed in the hope that it will be useful, but
11 # WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
15 # You should have received a copy of the GNU General Public License
16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
19 #use warnings; FIXME - Bug 2505
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
27 use C4::Search::PazPar2;
30 use C4::Reserves; # GetReserveStatus
33 use Koha::AuthorisedValues;
37 use Koha::RecordProcessor;
43 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
46 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
51 C4::Search - Functions for searching the Koha catalog.
55 See opac/opac-search.pl or catalogue/search.pl for example of usage
59 This module provides searching functions for Koha's bibliographic databases
73 &enabled_staff_search_views
76 # make all your functions, whether exported or not;
80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
88 my $dbh = C4::Context->dbh;
89 my $result = TransformMarcToKoha( $record, '' );
94 my ( $biblionumber, $title );
96 # search duplicate on ISBN, easy and fast..
98 if ( $result->{isbn} ) {
99 $result->{isbn} =~ s/\(.*$//;
100 $result->{isbn} =~ s/\s+$//;
101 $query = "isbn:$result->{isbn}";
105 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
111 $titleindex = 'title|exact';
112 $authorindex = 'author|exact';
114 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
116 $titleindex = 'ti,ext';
117 $authorindex = 'au,ext';
121 $result->{title} =~ s /\\//g;
122 $result->{title} =~ s /\"//g;
123 $result->{title} =~ s /\(//g;
124 $result->{title} =~ s /\)//g;
126 # FIXME: instead of removing operators, could just do
127 # quotes around the value
128 $result->{title} =~ s/(and|or|not)//g;
129 $query = "$titleindex:\"$result->{title}\"";
130 if ( $result->{author} ) {
131 $result->{author} =~ s /\\//g;
132 $result->{author} =~ s /\"//g;
133 $result->{author} =~ s /\(//g;
134 $result->{author} =~ s /\)//g;
136 # remove valid operators
137 $result->{author} =~ s/(and|or|not)//g;
138 $query .= " $op $authorindex:\"$result->{author}\"";
142 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
144 if (!defined $error) {
145 foreach my $possible_duplicate_record (@{$searchresults}) {
146 my $marcrecord = new_record_from_zebra(
148 $possible_duplicate_record
151 my $result = TransformMarcToKoha( $marcrecord, '' );
153 # FIXME :: why 2 $biblionumber ?
155 push @results, $result->{'biblionumber'};
156 push @results, $result->{'title'};
165 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
167 This function provides a simple search API on the bibliographic catalog
173 * $query can be a simple keyword or a complete CCL query
174 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
175 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
176 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
177 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
182 Returns an array consisting of three elements
183 * $error is undefined unless an error is detected
184 * $results is a reference to an array of records.
185 * $total_hits is the number of hits that would have been returned with no limit
187 If an error is returned the two other return elements are undefined. If error itself is undefined
188 the other two elements are always defined
190 =item C<usage in the script:>
194 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
196 if (defined $error) {
197 $template->param(query_error => $error);
198 warn "error: ".$error;
199 output_html_with_http_headers $input, $cookie, $template->output;
203 my $hits = @{$marcresults};
206 for my $r ( @{$marcresults} ) {
207 my $marcrecord = MARC::File::USMARC::decode($r);
208 my $biblio = TransformMarcToKoha($marcrecord,q{});
210 #build the iarray of hashs for the template.
212 title => $biblio->{'title'},
213 subtitle => $biblio->{'subtitle'},
214 biblionumber => $biblio->{'biblionumber'},
215 author => $biblio->{'author'},
216 publishercode => $biblio->{'publishercode'},
217 publicationyear => $biblio->{'publicationyear'},
222 $template->param(result=>\@results);
227 my ( $query, $offset, $max_results, $servers, %options ) = @_;
229 return ( 'No query entered', undef, undef ) unless $query;
230 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
231 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
239 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
241 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
244 # Initialize & Search Zebra
245 for ( my $i = 0 ; $i < @servers ; $i++ ) {
247 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
249 $query =~ s/=/:/g unless $options{skip_normalize};
250 $QParser->parse( $query );
251 $query = $QParser->target_syntax($servers[$i]);
252 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
254 $query =~ s/:/=/g unless $options{skip_normalize};
255 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
257 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
261 $zconns[$i]->errmsg() . " ("
262 . $zconns[$i]->errcode() . ") "
263 . $zconns[$i]->addinfo() . " "
264 . $zconns[$i]->diagset();
266 return ( $error, undef, undef ) if $zconns[$i]->errcode();
270 # caught a ZOOM::Exception
274 . $@->addinfo() . " "
276 warn $error." for query: $query";
277 return ( $error, undef, undef );
286 my $first_record = defined($offset) ? $offset + 1 : 1;
287 my $hits = $tmpresults[ $i - 1 ]->size();
288 $total_hits += $hits;
289 my $last_record = $hits;
290 if ( defined $max_results && $offset + $max_results < $hits ) {
291 $last_record = $offset + $max_results;
294 for my $j ( $first_record .. $last_record ) {
296 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
299 push @{$results}, $record if defined $record;
304 foreach my $zoom_query (@zoom_queries) {
305 $zoom_query->destroy();
308 return ( undef, $results, $total_hits );
313 ( undef, $results_hashref, \@facets_loop ) = getRecords (
315 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
316 $results_per_page, $offset, $branches, $itemtypes,
317 $query_type, $scan, $opac
320 The all singing, all dancing, multi-server, asynchronous, scanning,
321 searching, record nabbing, facet-building
323 See verbose embedded documentation.
329 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
330 $results_per_page, $offset, $branches, $itemtypes,
331 $query_type, $scan, $opac
334 my @servers = @$servers_ref;
335 my @sort_by = @$sort_by_ref;
336 $offset = 0 if $offset < 0;
338 # Initialize variables for the ZOOM connection and results object
342 my $results_hashref = ();
344 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
345 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
347 # Initialize variables for the faceted results objects
348 my $facets_counter = {};
349 my $facets_info = {};
350 my $facets = getFacets();
352 my @facets_loop; # stores the ref to array of hashes for template facets loop
354 ### LOOP THROUGH THE SERVERS
355 for ( my $i = 0 ; $i < @servers ; $i++ ) {
356 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
358 # perform the search, create the results objects
359 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
360 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
362 #$query_to_use = $simple_query if $scan;
363 warn $simple_query if ( $scan and $DEBUG );
365 # Check if we've got a query_type defined, if so, use it
368 if ($query_type =~ /^ccl/) {
369 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
370 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
371 } elsif ($query_type =~ /^cql/) {
372 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
373 } elsif ($query_type =~ /^pqf/) {
374 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
376 warn "Unknown query_type '$query_type'. Results undetermined.";
379 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
381 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
385 warn "WARNING: query problem with $query_to_use " . $@;
388 # Concatenate the sort_by limits and pass them to the results object
389 # Note: sort will override rank
391 foreach my $sort (@sort_by) {
392 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
393 $sort_by .= "1=1003 <i ";
395 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
396 $sort_by .= "1=1003 >i ";
398 elsif ( $sort eq "popularity_asc" ) {
399 $sort_by .= "1=9003 <i ";
401 elsif ( $sort eq "popularity_dsc" ) {
402 $sort_by .= "1=9003 >i ";
404 elsif ( $sort eq "call_number_asc" ) {
405 $sort_by .= "1=8007 <i ";
407 elsif ( $sort eq "call_number_dsc" ) {
408 $sort_by .= "1=8007 >i ";
410 elsif ( $sort eq "pubdate_asc" ) {
411 $sort_by .= "1=31 <i ";
413 elsif ( $sort eq "pubdate_dsc" ) {
414 $sort_by .= "1=31 >i ";
416 elsif ( $sort eq "acqdate_asc" ) {
417 $sort_by .= "1=32 <i ";
419 elsif ( $sort eq "acqdate_dsc" ) {
420 $sort_by .= "1=32 >i ";
422 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
423 $sort_by .= "1=4 <i ";
425 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
426 $sort_by .= "1=4 >i ";
429 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
432 if ( $sort_by && !$scan && $results[$i] ) {
433 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
434 warn "WARNING sort $sort_by failed";
437 } # finished looping through servers
439 # The big moment: asynchronously retrieve results from all servers
444 my ( $i, $size ) = @_;
447 # loop through the results
448 $results_hash->{'hits'} = $size;
450 if ( $offset + $results_per_page <= $size ) {
451 $times = $offset + $results_per_page;
457 for ( my $j = $offset ; $j < $times ; $j++ ) {
461 ## Check if it's an index scan
463 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
465 # here we create a minimal MARC record and hand it off to the
466 # template just like a normal result ... perhaps not ideal, but
468 my $tmprecord = MARC::Record->new();
469 $tmprecord->encoding('UTF-8');
473 # the minimal record in author/title (depending on MARC flavour)
474 if ( C4::Context->preference("marcflavour") eq
477 $tmptitle = MARC::Field->new(
482 $tmprecord->append_fields($tmptitle);
486 MARC::Field->new( '245', ' ', ' ', a => $term, );
488 MARC::Field->new( '100', ' ', ' ', a => $occ, );
489 $tmprecord->append_fields($tmptitle);
490 $tmprecord->append_fields($tmpauthor);
492 $results_hash->{'RECORDS'}[$j] =
493 $tmprecord->as_usmarc();
498 $record = $results[ $i - 1 ]->record($j)->raw();
499 # warn "RECORD $j:".$record;
500 $results_hash->{'RECORDS'}[$j] = $record;
504 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
506 # Fill the facets while we're looping, but only for the
507 # biblioserver and not for a scan
508 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
509 $facets_counter = GetFacets( $results[ $i - 1 ] );
510 $facets_info = _get_facets_info( $facets );
514 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
516 sort { $a cmp $b } keys %$facets_counter
519 my @this_facets_array;
522 $facets_counter->{$link_value}
523 ->{$b} <=> $facets_counter->{$link_value}
525 } keys %{ $facets_counter->{$link_value} }
528 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
529 my $facet_link_value = $one_facet;
530 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
532 # fix the length that will display in the label,
533 my $facet_label_value = $one_facet;
534 my $facet_max_length = C4::Context->preference(
535 'FacetLabelTruncationLength')
538 substr( $one_facet, 0, $facet_max_length )
540 if length($facet_label_value) >
543 # if it's a branch, label by the name, not the code,
544 if ( $link_value =~ /branch/ ) {
545 if ( defined $branches
546 && ref($branches) eq "HASH"
547 && defined $branches->{$one_facet}
548 && ref( $branches->{$one_facet} ) eq
552 $branches->{$one_facet}
556 $facet_label_value = "*";
560 # if it's a itemtype, label by the name, not the code,
561 if ( $link_value =~ /itype/ ) {
562 if ( defined $itemtypes
563 && ref($itemtypes) eq "HASH"
564 && defined $itemtypes->{$one_facet}
565 && ref( $itemtypes->{$one_facet} ) eq
569 $itemtypes->{$one_facet}
570 ->{translated_description};
574 # also, if it's a location code, use the name instead of the code
575 if ( $link_value =~ /location/ ) {
576 # TODO Retrieve all authorised values at once, instead of 1 query per entry
577 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
578 $facet_label_value = $av->count ? $av->next->opac_description : '';
581 # also, if it's a collection code, use the name instead of the code
582 if ( $link_value =~ /ccode/ ) {
583 # TODO Retrieve all authorised values at once, instead of 1 query per entry
584 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
585 $facet_label_value = $av->count ? $av->next->opac_description : '';
588 # but we're down with the whole label being in the link's title.
589 push @this_facets_array,
592 $facets_counter->{$link_value}
594 facet_label_value => $facet_label_value,
595 facet_title_value => $one_facet,
596 facet_link_value => $facet_link_value,
597 type_link_value => $link_value,
599 if ($facet_label_value);
604 type_link_value => $link_value,
605 type_id => $link_value . "_id",
607 . $facets_info->{$link_value}->{'label_value'} =>
609 facets => \@this_facets_array,
613 $facets_info->{$link_value}->{'label_value'} =~
616 and ( Koha::Libraries->search->count == 1 )
623 # This sorts the facets into alphabetical order
625 foreach my $f (@facets_loop) {
626 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
628 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
631 return ( undef, $results_hashref, \@facets_loop );
639 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
641 if ( $use_zebra_facets ) {
642 $facets = _get_facets_from_zebra( $rs );
644 $facets = _get_facets_from_records( $rs );
650 sub _get_facets_from_records {
654 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
655 my $facets_config = getFacets();
657 my $size = $rs->size();
658 my $jmax = $size > $facets_maxrecs
662 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
664 my $marc_record = new_record_from_zebra (
666 $rs->record( $j )->raw()
669 if ( ! defined $marc_record ) {
670 warn "ERROR DECODING RECORD - $@: " .
671 $rs->record( $j )->raw();
675 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
681 =head2 _get_facets_data_from_record
683 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
685 Internal function that extracts facets information from a MARC::Record object
686 and populates $facets_counter for using in getRecords.
688 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
693 sub _get_facets_data_from_record {
695 my ( $marc_record, $facets, $facets_counter ) = @_;
697 for my $facet (@$facets) {
701 foreach my $tag ( @{ $facet->{ tags } } ) {
703 # tag number is the first three digits
704 my $tag_num = substr( $tag, 0, 3 );
705 # subfields are the remainder
706 my $subfield_letters = substr( $tag, 3 );
708 my @fields = $marc_record->field( $tag_num );
709 foreach my $field (@fields) {
710 # If $field->indicator(1) eq 'z', it means it is a 'see from'
711 # field introduced because of IncludeSeeFromInSearches, so skip it
712 next if $field->indicator(1) eq 'z';
714 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
716 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
717 push @used_datas, $data;
718 $facets_counter->{ $facet->{ idx } }->{ $data }++;
725 =head2 _get_facets_from_zebra
727 my $facets = _get_facets_from_zebra( $result_set )
729 Retrieves facets for a specified result set. It loops through the facets defined
730 in C4::Koha::getFacets and returns a hash with the following structure:
740 sub _get_facets_from_zebra {
744 # save current elementSetName
745 my $elementSetName = $rs->option( 'elementSetName' );
747 my $facets_loop = getFacets();
748 my $facets_data = {};
749 # loop through defined facets and fill the facets hashref
750 foreach my $facet ( @$facets_loop ) {
752 my $idx = $facet->{ idx };
753 my $sep = $facet->{ sep };
754 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
755 if ( $facet_values ) {
756 # we've actually got a result
757 $facets_data->{ $idx } = $facet_values;
760 # set elementSetName to its previous value to avoid side effects
761 $rs->option( elementSetName => $elementSetName );
766 =head2 _get_facet_from_result_set
769 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
771 Internal function that extracts facet information for a specific index ($facet_idx) and
772 returns a hash containing facet values and count:
775 $facet_value => $count ,
779 Warning: this function has the side effect of changing the elementSetName for the result
780 set. It is a helper function for the main loop, which takes care of backing it up for
785 sub _get_facet_from_result_set {
787 my $facet_idx = shift;
791 my $internal_sep = '<*>';
792 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
794 return if ( ! defined $facet_idx || ! defined $rs );
795 # zebra's facet element, untokenized index
796 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
797 # configure zebra results for retrieving the desired facet
798 $rs->option( elementSetName => $facet_element );
799 # get the facet record from result set
800 my $facet = $rs->record( 0 )->raw;
801 # if the facet has no restuls...
802 return if !defined $facet;
803 # TODO: benchmark DOM vs. SAX performance
804 my $facet_dom = XML::LibXML->load_xml(
807 my @terms = $facet_dom->getElementsByTagName('term');
811 foreach my $term ( @terms ) {
812 my $facet_value = $term->textContent;
813 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
814 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
820 =head2 _get_facets_info
822 my $facets_info = C4::Search::_get_facets_info( $facets )
824 Internal function that extracts facets information and properly builds
825 the data structure needed to render facet labels.
829 sub _get_facets_info {
833 my $facets_info = {};
835 for my $facet ( @$facets ) {
836 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
844 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
845 $results_per_page, $offset, $branches, $query_type,
848 #NOTE: Parameter $branches is not used here !
850 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
852 $paz->search($simple_query);
853 sleep 1; # FIXME: WHY?
856 my $results_hashref = {};
857 my $stats = XMLin($paz->stat);
858 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
860 # for a grouped search result, the number of hits
861 # is the number of groups returned; 'bib_hits' will have
862 # the total number of bibs.
863 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
864 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
866 HIT: foreach my $hit (@{ $results->{'hit'} }) {
867 my $recid = $hit->{recid}->[0];
869 my $work_title = $hit->{'md-work-title'}->[0];
871 if (exists $hit->{'md-work-author'}) {
872 $work_author = $hit->{'md-work-author'}->[0];
874 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
876 my $result_group = {};
877 $result_group->{'group_label'} = $group_label;
878 $result_group->{'group_merge_key'} = $recid;
881 if (exists $hit->{count}) {
882 $count = $hit->{count}->[0];
884 $result_group->{'group_count'} = $count;
886 for (my $i = 0; $i < $count; $i++) {
887 # FIXME -- may need to worry about diacritics here
888 my $rec = $paz->record($recid, $i);
889 push @{ $result_group->{'RECORDS'} }, $rec;
892 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
895 # pass through facets
896 my $termlist_xml = $paz->termlist('author,subject');
897 my $terms = XMLin($termlist_xml, forcearray => 1);
898 my @facets_loop = ();
899 #die Dumper($results);
900 # foreach my $list (sort keys %{ $terms->{'list'} }) {
902 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
904 # facet_label_value => $facet->{'name'}->[0],
907 # push @facets_loop, ( {
908 # type_label => $list,
909 # facets => \@facets,
913 return ( undef, $results_hashref, \@facets_loop );
917 sub _detect_truncation {
918 my ( $operand, $index ) = @_;
919 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
922 my @wordlist = split( /\s/, $operand );
923 foreach my $word (@wordlist) {
924 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
925 push @rightlefttruncated, $word;
927 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
928 push @lefttruncated, $word;
930 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
931 push @righttruncated, $word;
933 elsif ( index( $word, "*" ) < 0 ) {
934 push @nontruncated, $word;
937 push @regexpr, $word;
941 \@nontruncated, \@righttruncated, \@lefttruncated,
942 \@rightlefttruncated, \@regexpr
947 sub _build_stemmed_operand {
948 my ($operand,$lang) = @_;
949 require Lingua::Stem::Snowball ;
950 my $stemmed_operand=q{};
952 # Stemmer needs language
953 return $operand unless $lang;
955 # If operand contains a digit, it is almost certainly an identifier, and should
956 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
957 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
958 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
959 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
960 return $operand if $operand =~ /\d/;
962 # FIXME: the locale should be set based on the user's language and/or search choice
964 # Make sure we only use the first two letters from the language code
965 $lang = lc(substr($lang, 0, 2));
966 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
967 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
968 if ($lang eq 'nb' || $lang eq 'nn') {
971 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
972 encoding => "UTF-8" );
974 my @words = split( / /, $operand );
975 my @stems = $stemmer->stem(\@words);
976 for my $stem (@stems) {
977 $stemmed_operand .= "$stem";
978 $stemmed_operand .= "?"
979 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
980 $stemmed_operand .= " ";
982 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
983 return $stemmed_operand;
987 sub _build_weighted_query {
989 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
990 # pretty well but could work much better if we had a smarter query parser
991 my ( $operand, $stemmed_operand, $index ) = @_;
992 my $stemming = C4::Context->preference("QueryStemming") || 0;
993 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
994 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
995 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
997 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
999 # Keyword, or, no index specified
1000 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1002 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1003 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1004 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1005 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1006 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1007 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1008 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1009 if $fuzzy_enabled; # add fuzzy, word list
1010 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1011 if ( $stemming and $stemmed_operand )
1012 ; # add stemming, right truncation
1013 $weighted_query .= " or wrdl,r9=\"$operand\"";
1015 # embedded sorting: 0 a-z; 1 z-a
1016 # $weighted_query .= ") or (sort1,aut=1";
1019 # Barcode searches should skip this process
1020 elsif ( $index eq 'bc' ) {
1021 $weighted_query .= "bc=\"$operand\"";
1024 # Authority-number searches should skip this process
1025 elsif ( $index eq 'an' ) {
1026 $weighted_query .= "an=\"$operand\"";
1029 # If the index is numeric, don't autoquote it.
1030 elsif ( $index =~ /,st-numeric$/ ) {
1031 $weighted_query .= " $index=$operand";
1034 # If the index already has more than one qualifier, wrap the operand
1035 # in quotes and pass it back (assumption is that the user knows what they
1036 # are doing and won't appreciate us mucking up their query
1037 elsif ( $index =~ ',' ) {
1038 $weighted_query .= " $index=\"$operand\"";
1041 #TODO: build better cases based on specific search indexes
1043 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1044 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1045 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1046 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1047 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1048 if $fuzzy_enabled; # add fuzzy, word list
1049 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1050 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1053 $weighted_query .= "))"; # close rank specification
1054 return $weighted_query;
1059 Return an array with available indexes.
1083 'Author-personal-bibliography',
1093 'Chronological-subdivision',
1103 'Conference-name-heading',
1104 'Conference-name-see',
1105 'Conference-name-seealso',
1110 'Corporate-name-heading',
1111 'Corporate-name-see',
1112 'Corporate-name-seealso',
1113 'Country-publication',
1116 'date-entered-on-file',
1117 'Date-of-acquisition',
1118 'Date-of-publication',
1119 'Date-time-last-modified',
1120 'Dewey-classification',
1121 'Dissertation-information',
1130 'Geographic-subdivision',
1133 'Heading-use-main-or-added-entry',
1134 'Heading-use-series-added-entry ',
1135 'Heading-use-subject-added-entry',
1139 'Identifier-standard',
1140 'Illustration-code',
1142 'Index-term-uncontrolled',
1143 'Interest-age-level',
1144 'Interest-grade-level',
1154 'language-original',
1164 'Local-classification',
1167 'Match-heading-see-from',
1175 'Name-geographic-heading',
1176 'Name-geographic-see',
1177 'Name-geographic-seealso',
1183 'Other-control-number',
1186 'Personal-name-heading',
1187 'Personal-name-see',
1188 'Personal-name-seealso',
1190 'Place-publication',
1197 'Reading-grade-level',
1198 'Record-control-number',
1209 'Subject-heading-thesaurus',
1210 'Subject-name-personal',
1211 'Subject-subdivision',
1220 'Term-genre-form-heading',
1221 'Term-genre-form-see',
1222 'Term-genre-form-seealso',
1228 'Title-uniform-heading',
1229 'Title-uniform-see',
1230 'Title-uniform-seealso',
1240 'classification-source',
1242 'coded-location-qualifier',
1253 'Local-classification',
1256 'materials-specified',
1261 'Number-local-acquisition',
1266 'replacementpricedate',
1281 =head2 _handle_exploding_index
1283 my $query = _handle_exploding_index($index, $term)
1285 Callback routine to generate the search for "exploding" indexes (i.e.
1286 those indexes which are turned into multiple or-connected searches based
1291 sub _handle_exploding_index {
1292 my ($QParser, $filter, $params, $negate, $server) = @_;
1293 my $index = $filter;
1294 my $term = join(' ', @$params);
1296 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1298 my $marcflavour = C4::Context->preference('marcflavour');
1300 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1301 my $wantedcodes = '';
1302 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1303 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1304 foreach my $auth (@$results) {
1305 my $record = MARC::Record->new_from_usmarc($auth);
1306 my @references = $record->field('5..');
1308 if ($index eq 'su-br') {
1310 } elsif ($index eq 'su-na') {
1312 } elsif ($index eq 'su-rl') {
1315 foreach my $reference (@references) {
1316 my $codes = $reference->subfield($codesubfield);
1317 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1321 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1327 ( $operators, $operands, $indexes, $limits,
1328 $sort_by, $scan, $lang ) =
1329 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1331 Shim function to ease the transition from buildQuery to a new QueryParser.
1332 This function is called at the beginning of buildQuery, and modifies
1333 buildQuery's input. If it can handle the input, it returns a query that
1334 buildQuery will not try to parse.
1339 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1341 my @operators = $operators ? @$operators : ();
1342 my @indexes = $indexes ? @$indexes : ();
1343 my @operands = $operands ? @$operands : ();
1344 my @limits = $limits ? @$limits : ();
1345 my @sort_by = $sort_by ? @$sort_by : ();
1347 my $query = $operands[0];
1353 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1354 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1355 undef $QParser if (scalar @limits > 0);
1359 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1361 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1362 next unless $operands[$ii];
1363 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1365 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1366 $query .= $operands[$ii];
1368 elsif ( $indexes[$ii] =~ m/su-/ ) {
1369 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1373 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1376 foreach my $limit (@limits) {
1378 if ( scalar(@sort_by) > 0 ) {
1380 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1381 $query =~ s/$modifier_re//g;
1382 foreach my $modifier (@sort_by) {
1383 $query .= " #$modifier";
1387 $query_desc = $query;
1388 $query_desc =~ s/\s+/ /g;
1389 if ( C4::Context->preference("QueryWeightFields") ) {
1391 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1392 { 'target_syntax_callback' => \&_handle_exploding_index } );
1393 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1394 { 'target_syntax_callback' => \&_handle_exploding_index } );
1395 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1396 { 'target_syntax_callback' => \&_handle_exploding_index } );
1397 $QParser->parse($query);
1398 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1401 require Koha::QueryParser::Driver::PQF;
1402 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1403 s/$modifier_re//g for @operands;
1406 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1412 $simple_query, $query_cgi,
1413 $query_desc, $limit,
1414 $limit_cgi, $limit_desc,
1415 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1417 Build queries and limits in CCL, CGI, Human,
1418 handle truncation, stemming, field weighting, fuzziness, etc.
1420 See verbose embedded documentation.
1426 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1428 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1431 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1434 my @operators = $operators ? @$operators : ();
1435 my @indexes = $indexes ? @$indexes : ();
1436 my @operands = $operands ? @$operands : ();
1437 my @limits = $limits ? @$limits : ();
1438 my @sort_by = $sort_by ? @$sort_by : ();
1440 my $stemming = C4::Context->preference("QueryStemming") || 0;
1441 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1442 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1443 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1445 my $query = $operands[0];
1446 my $simple_query = $operands[0];
1448 # initialize the variables we're passing back
1457 my $cclindexes = getIndexes();
1458 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1459 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1461 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1463 $query = "ccl=$query" if $cclq;
1466 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1468 if ( $query =~ /^ccl=/ ) {
1470 # This is needed otherwise ccl= and &limit won't work together, and
1471 # this happens when selecting a subject on the opac-detail page
1472 @limits = grep {!/^$/} @limits;
1473 my $original_q = $q; # without available part
1474 unless ( grep { /^available$/ } @limits ) {
1475 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1479 if ( grep { /^available$/ } @limits ) {
1480 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1481 delete $limits['available'];
1483 $q .= ' and '.join(' and ', @limits) if @limits;
1485 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1487 if ( $query =~ /^cql=/ ) {
1488 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1490 if ( $query =~ /^pqf=/ ) {
1492 $query_cgi = "q=".uri_escape_utf8($query_desc);
1495 $query_cgi = "q=pqf=".uri_escape_utf8($');
1497 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1500 # pass nested queries directly
1501 # FIXME: need better handling of some of these variables in this case
1502 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1503 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1504 # if ( $query =~ /(\(|\))/ ) {
1506 # undef, $query, $simple_query, $query_cgi,
1507 # $query, $limit, $limit_cgi, $limit_desc,
1512 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1513 # query operands and indexes and add stemming, truncation, field weighting, etc.
1514 # Once we do so, we'll end up with a value in $query, just like if we had an
1515 # incoming $query from the user
1518 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1519 my $previous_operand
1520 ; # a flag used to keep track if there was a previous query
1521 # if there was, we can apply the current operator
1523 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1525 # COMBINE OPERANDS, INDEXES AND OPERATORS
1526 if ( ($operands[$i] // '') ne '' ) {
1527 $operands[$i]=~s/^\s+//;
1529 # A flag to determine whether or not to add the index to the query
1532 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1533 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1537 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1539 my $operand = $operands[$i];
1540 my $index = $indexes[$i];
1542 # Add index-specific attributes
1544 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1545 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1546 #which is processed higher up in this sub. Other than that, year searches are typically
1547 #handled as limits which are not processed her either.
1549 # Search ranges: Date of Publication, st-numeric
1550 if ( $index =~ /(yr|st-numeric)/ ) {
1551 #weight_fields/relevance search causes errors with date ranges
1552 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1553 #In the case of YYYY-YYYY, it will return no results
1554 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1557 # Date of Acquisition
1558 elsif ( $index =~ /acqdate/ ) {
1559 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1560 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1561 #top of the results just because they have lots of item records matching that date.
1562 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1564 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1566 # ISBN,ISSN,Standard Number, don't need special treatment
1567 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1569 $stemming, $auto_truncation,
1570 $weight_fields, $fuzzy_enabled
1573 if ( $index eq 'nb' ) {
1574 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1575 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1576 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1577 $indexes[$i] = $index = '';
1586 # Set default structure attribute (word list)
1587 my $struct_attr = q{};
1588 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1589 $struct_attr = ",wrdl";
1592 # Some helpful index variants
1593 my $index_plus = $index . $struct_attr . ':';
1594 my $index_plus_comma = $index . $struct_attr . ',';
1596 if ($auto_truncation){
1597 unless ( $index =~ /,(st-|phr|ext)/ ) {
1598 #FIXME only valid with LTR scripts
1599 $operand=join(" ",map{
1600 (index($_,"*")>0?"$_":"$_*")
1601 }split (/\s+/,$operand));
1602 warn $operand if $DEBUG;
1607 my $truncated_operand;
1608 my( $nontruncated, $righttruncated, $lefttruncated,
1609 $rightlefttruncated, $regexpr
1610 ) = _detect_truncation( $operand, $index );
1612 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1617 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1618 scalar(@$rightlefttruncated) > 0 )
1621 # Don't field weight or add the index to the query, we do it here
1623 undef $weight_fields;
1624 my $previous_truncation_operand;
1625 if (scalar @$nontruncated) {
1626 $truncated_operand .= "$index_plus @$nontruncated ";
1627 $previous_truncation_operand = 1;
1629 if (scalar @$righttruncated) {
1630 $truncated_operand .= "and " if $previous_truncation_operand;
1631 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1632 $previous_truncation_operand = 1;
1634 if (scalar @$lefttruncated) {
1635 $truncated_operand .= "and " if $previous_truncation_operand;
1636 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1637 $previous_truncation_operand = 1;
1639 if (scalar @$rightlefttruncated) {
1640 $truncated_operand .= "and " if $previous_truncation_operand;
1641 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1642 $previous_truncation_operand = 1;
1645 $operand = $truncated_operand if $truncated_operand;
1646 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1649 my $stemmed_operand;
1650 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1653 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1655 # Handle Field Weighting
1656 my $weighted_operand;
1657 if ($weight_fields) {
1658 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1659 $operand = $weighted_operand;
1663 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1665 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1667 query_cgi => $query_cgi,
1668 query_desc => $query_desc,
1669 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1670 parsed_operand => $operand,
1671 original_operand => $operands[$i] // '',
1673 index_plus => $index_plus,
1674 indexes_set => $indexes_set,
1675 previous_operand => $previous_operand,
1681 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1684 my %group_OR_limits;
1685 my $availability_limit;
1686 foreach my $this_limit (@limits) {
1687 next unless $this_limit;
1688 if ( $this_limit =~ /available/ ) {
1690 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1692 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1693 $availability_limit .=
1694 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1695 $limit_cgi .= "&limit=available";
1699 # group_OR_limits, prefixed by mc-
1700 # OR every member of the group
1701 elsif ( $this_limit =~ /mc/ ) {
1702 my ($k,$v) = split(/:/, $this_limit,2);
1703 if ( $k !~ /mc-i(tem)?type/ ) {
1704 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1705 $this_limit =~ tr/"//d;
1706 $this_limit = $k.":'".$v."'";
1709 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1710 $limit_desc .= " or " if $group_OR_limits{$k};
1711 $group_OR_limits{$k} .= "$this_limit";
1712 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1713 $limit_desc .= " $this_limit";
1716 # Regular old limits
1718 $limit .= " and " if $limit || $query;
1719 $limit .= "$this_limit";
1720 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1721 if ($this_limit =~ /^branch:(.+)/) {
1722 my $branchcode = $1;
1723 my $library = Koha::Libraries->find( $branchcode );
1724 if (defined $library) {
1725 $limit_desc .= " branch:" . $library->branchname;
1727 $limit_desc .= " $this_limit";
1730 $limit_desc .= " $this_limit";
1734 foreach my $k (keys (%group_OR_limits)) {
1735 $limit .= " and " if ( $query || $limit );
1736 $limit .= "($group_OR_limits{$k})";
1738 if ($availability_limit) {
1739 $limit .= " and " if ( $query || $limit );
1740 $limit .= "($availability_limit)";
1743 # Normalize the query and limit strings
1744 # This is flawed , means we can't search anything with : in it
1745 # if user wants to do ccl or cql, start the query with that
1746 # $query =~ s/:/=/g;
1747 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1748 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1749 $query =~ s/(?<=(wrdl)):/=/g;
1750 $query =~ s/(?<=(trn|phr)):/=/g;
1751 $query =~ s/(?<=(st-numeric)):/=/g;
1752 $query =~ s/(?<=(st-year)):/=/g;
1753 $query =~ s/(?<=(st-date-normalized)):/=/g;
1755 for ( $query, $query_desc, $limit, $limit_desc ) {
1756 s/ +/ /g; # remove extra spaces
1757 s/^ //g; # remove any beginning spaces
1758 s/ $//g; # remove any ending spaces
1759 s/==/=/g; # remove double == from query
1761 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1763 for ($query_cgi,$simple_query) {
1766 # append the limit to the query
1767 $query .= " " . $limit;
1771 warn "QUERY:" . $query;
1772 warn "QUERY CGI:" . $query_cgi;
1773 warn "QUERY DESC:" . $query_desc;
1774 warn "LIMIT:" . $limit;
1775 warn "LIMIT CGI:" . $limit_cgi;
1776 warn "LIMIT DESC:" . $limit_desc;
1777 warn "---------\nLeave buildQuery\n---------";
1781 undef, $query, $simple_query, $query_cgi,
1782 $query_desc, $limit, $limit_cgi, $limit_desc,
1787 =head2 _build_initial_query
1789 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1791 Build a section of the initial query containing indexes, operators, and operands.
1795 sub _build_initial_query {
1799 if ($params->{previous_operand}){
1800 #If there is a previous operand, add a supplied operator or the default 'and'
1801 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1804 #NOTE: indexes_set is typically set when doing truncation or field weighting
1805 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1807 #e.g. "kw,wrdl:test"
1808 #e.g. " and kw,wrdl:test"
1809 $params->{query} .= $operator . $operand;
1811 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1812 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1813 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1815 #e.g. " and kw,wrdl: test"
1816 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1818 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1820 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1823 =head2 searchResults
1825 my @search_results = searchResults($search_context, $searchdesc, $hits,
1826 $results_per_page, $offset, $scan,
1829 Format results in a form suitable for passing to the template
1833 # IMO this subroutine is pretty messy still -- it's responsible for
1834 # building the HTML output for the template
1836 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1837 my $dbh = C4::Context->dbh;
1842 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1843 my ($is_opac, $hidelostitems);
1844 if ($search_context->{'interface'} eq 'opac') {
1845 $hidelostitems = C4::Context->preference('hidelostitems');
1849 my $record_processor = Koha::RecordProcessor->new({
1850 filters => 'ViewPolicy'
1853 #Build branchnames hash
1854 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1856 # FIXME - We build an authorised values hash here, using the default framework
1857 # though it is possible to have different authvals for different fws.
1859 my $shelflocations =
1860 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1862 # get notforloan authorised value list (see $shelflocations FIXME)
1863 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1864 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1867 my $itemtypes = Koha::ItemTypes->search_with_localization;
1868 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1870 #search item field code
1871 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1873 ## find column names of items related to MARC
1874 my %subfieldstosearch;
1875 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1876 for my $column ( @columns ) {
1877 my ( $tagfield, $tagsubfield ) =
1878 &GetMarcFromKohaField( "items." . $column, "" );
1879 if ( defined $tagsubfield ) {
1880 $subfieldstosearch{$column} = $tagsubfield;
1884 # handle which records to actually retrieve
1886 if ( $hits && $offset + $results_per_page <= $hits ) {
1887 $times = $offset + $results_per_page;
1890 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1893 my $marcflavour = C4::Context->preference("marcflavour");
1894 # We get the biblionumber position in MARC
1895 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1897 # set stuff for XSLT processing here once, not later again for every record we retrieved
1901 $xslsyspref = "OPACXSLTResultsDisplay";
1902 $xslfile = C4::Context->preference( $xslsyspref );
1904 $xslsyspref = "XSLTResultsDisplay";
1905 $xslfile = C4::Context->preference( $xslsyspref ) || "default";
1907 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1908 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1910 my $userenv = C4::Context->userenv;
1912 = ( defined $userenv and $userenv->{number} )
1913 ? Koha::Patrons->find( $userenv->{number} )
1915 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1917 # loop through all of the records we've retrieved
1918 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1922 # For Scan searches we built USMARC data
1923 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1925 # Normal search, render from Zebra's output
1926 $marcrecord = new_record_from_zebra(
1931 if ( ! defined $marcrecord ) {
1932 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1940 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1941 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1943 SetUTF8Flag($marcrecord);
1944 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1945 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1946 $oldbiblio->{result_number} = $i + 1;
1948 # add imageurl to itemtype if there is one
1949 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1951 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1952 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1953 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1954 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1955 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1957 # edition information, if any
1958 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1959 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1960 # Build summary if there is one (the summary is defined in the itemtypes table)
1961 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1962 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1963 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1964 my @fields = $marcrecord->fields();
1967 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1969 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1970 $tag =~ /(.{3})(.)/;
1971 if($marcrecord->field($1)){
1972 my @abc = $marcrecord->field($1)->subfield($2);
1973 $tags->{$tag} = $#abc + 1 ;
1977 # We catch how many times to repeat this line
1979 foreach my $tag (keys(%$tags)){
1980 $max = $tags->{$tag} if($tags->{$tag} > $max);
1983 # we replace, and repeat each line
1984 for (my $i = 0 ; $i < $max ; $i++){
1985 my $newline = $line;
1987 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1988 $tag =~ /(.{3})(.)/;
1990 if($marcrecord->field($1)){
1991 my @repl = $marcrecord->field($1)->subfield($2);
1992 my $subfieldvalue = $repl[$i];
1993 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1996 $newsummary .= "$newline\n";
2000 $newsummary =~ s/\[(.*?)]//g;
2001 $newsummary =~ s/\n/<br\/>/g;
2002 $oldbiblio->{summary} = $newsummary;
2005 # Pull out the items fields
2006 my @fields = $marcrecord->field($itemtag);
2007 my $marcflavor = C4::Context->preference("marcflavour");
2009 # adding linked items that belong to host records
2010 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
2011 my $analyticsfield = '773';
2012 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2013 $analyticsfield = '773';
2014 } elsif ($marcflavor eq 'UNIMARC') {
2015 $analyticsfield = '461';
2017 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2018 my $hostbiblionumber = $hostfield->subfield("0");
2019 my $linkeditemnumber = $hostfield->subfield("9");
2020 if( $hostbiblionumber ) {
2021 my $hostbiblio = GetMarcBiblio({
2022 biblionumber => $hostbiblionumber,
2023 embed_items => 1 });
2024 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2026 my @hostitems = $hostbiblio->field($itemfield);
2027 foreach my $hostitem (@hostitems){
2028 if ($hostitem->subfield("9") eq $linkeditemnumber){
2029 my $linkeditem =$hostitem;
2030 # append linked items if they exist
2031 push @fields, $linkeditem if $linkeditem;
2039 # Setting item statuses for display
2040 my @available_items_loop;
2041 my @onloan_items_loop;
2042 my @other_items_loop;
2044 my $available_items;
2048 my $ordered_count = 0;
2049 my $available_count = 0;
2050 my $onloan_count = 0;
2051 my $longoverdue_count = 0;
2052 my $other_count = 0;
2053 my $withdrawn_count = 0;
2054 my $itemlost_count = 0;
2055 my $hideatopac_count = 0;
2056 my $itembinding_count = 0;
2057 my $itemdamaged_count = 0;
2058 my $item_in_transit_count = 0;
2059 my $can_place_holds = 0;
2060 my $item_onhold_count = 0;
2061 my $notforloan_count = 0;
2062 my $items_count = scalar(@fields);
2063 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2064 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2065 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2067 # loop through every item
2068 foreach my $field (@fields) {
2071 # populate the items hash
2072 foreach my $code ( keys %subfieldstosearch ) {
2073 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2075 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2079 # hidden because lost
2080 if ($hidelostitems && $item->{itemlost}) {
2081 $hideatopac_count++;
2084 # hidden based on OpacHiddenItems syspref
2085 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
2087 push @hiddenitems, @hi;
2088 $hideatopac_count++;
2093 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2094 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2096 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2097 if ($item->{$hbranch}) {
2098 $item->{'branchname'} = $branches{$item->{$hbranch}};
2100 elsif ($item->{$otherbranch}) { # Last resort
2101 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2104 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2105 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2106 if ( $item->{onloan}
2108 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
2111 my $key = $prefix . $item->{onloan} . $item->{barcode};
2112 $onloan_items->{$key}->{due_date} = $item->{onloan};
2113 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2114 $onloan_items->{$key}->{branchname} = $item->{branchname};
2115 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2116 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2117 $onloan_items->{$key}->{description} = $item->{description};
2118 $onloan_items->{$key}->{imageurl} =
2119 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2121 # if something's checked out and lost, mark it as 'long overdue'
2122 if ( $item->{itemlost} ) {
2123 $onloan_items->{$key}->{longoverdue}++;
2124 $longoverdue_count++;
2126 else { # can place holds as long as item isn't lost
2127 $can_place_holds = 1;
2131 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2134 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2137 if ( $item->{notforloan} < 0 ) {
2139 } elsif ( $item->{notforloan} > 0 ) {
2140 $notforloan_count++;
2143 # is item in transit?
2144 my $transfertwhen = '';
2145 my ($transfertfrom, $transfertto);
2147 # is item on the reserve shelf?
2148 my $reservestatus = '';
2150 unless ($item->{withdrawn}
2151 || $item->{itemlost}
2153 || $item->{notforloan}
2154 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2155 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2157 # A couple heuristics to limit how many times
2158 # we query the database for item transfer information, sacrificing
2159 # accuracy in some cases for speed;
2161 # 1. don't query if item has one of the other statuses
2162 # 2. don't check transit status if the bib has
2163 # more than 20 items
2165 # FIXME: to avoid having the query the database like this, and to make
2166 # the in transit status count as unavailable for search limiting,
2167 # should map transit status to record indexed in Zebra.
2169 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2170 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2173 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2174 if ( $item->{withdrawn}
2175 || $item->{itemlost}
2177 || $item->{notforloan}
2178 || $reservestatus eq 'Waiting'
2179 || ($transfertwhen ne ''))
2181 $withdrawn_count++ if $item->{withdrawn};
2182 $itemlost_count++ if $item->{itemlost};
2183 $itemdamaged_count++ if $item->{damaged};
2184 $item_in_transit_count++ if $transfertwhen ne '';
2185 $item_onhold_count++ if $reservestatus eq 'Waiting';
2186 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2188 # can place a hold on a item if
2189 # not lost nor withdrawn
2190 # not damaged unless AllowHoldsOnDamagedItems is true
2191 # item is either for loan or on order (notforloan < 0)
2192 $can_place_holds = 1
2195 && !$item->{withdrawn}
2196 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2197 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2202 my $key = $prefix . $item->{status};
2203 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2204 $other_items->{$key}->{$_} = $item->{$_};
2206 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2207 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2208 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2209 $other_items->{$key}->{count}++ if $item->{$hbranch};
2210 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2211 $other_items->{$key}->{description} = $item->{description};
2212 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2216 $can_place_holds = 1;
2218 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2219 foreach (qw(branchname itemcallnumber description)) {
2220 $available_items->{$prefix}->{$_} = $item->{$_};
2222 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2223 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2226 } # notforloan, item level and biblioitem level
2228 # if all items are hidden, do not show the record
2229 if ($items_count > 0 && $hideatopac_count == $items_count) {
2233 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2234 for my $key ( sort keys %$onloan_items ) {
2235 (++$onloanitemscount > $maxitems) and last;
2236 push @onloan_items_loop, $onloan_items->{$key};
2238 for my $key ( sort keys %$other_items ) {
2239 (++$otheritemscount > $maxitems) and last;
2240 push @other_items_loop, $other_items->{$key};
2242 for my $key ( sort keys %$available_items ) {
2243 (++$availableitemscount > $maxitems) and last;
2244 push @available_items_loop, $available_items->{$key}
2247 # XSLT processing of some stuff
2248 # we fetched the sysprefs already before the loop through all retrieved record!
2249 if (!$scan && $xslfile) {
2250 $record_processor->options({
2251 frameworkcode => $fw,
2252 interface => $search_context->{'interface'}
2255 $record_processor->process($marcrecord);
2256 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2259 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2260 if (!C4::Context->preference("item-level_itypes")) {
2261 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2262 $can_place_holds = 0;
2265 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2266 $oldbiblio->{items_count} = $items_count;
2267 $oldbiblio->{available_items_loop} = \@available_items_loop;
2268 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2269 $oldbiblio->{other_items_loop} = \@other_items_loop;
2270 $oldbiblio->{availablecount} = $available_count;
2271 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2272 $oldbiblio->{onloancount} = $onloan_count;
2273 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2274 $oldbiblio->{othercount} = $other_count;
2275 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2276 $oldbiblio->{withdrawncount} = $withdrawn_count;
2277 $oldbiblio->{itemlostcount} = $itemlost_count;
2278 $oldbiblio->{damagedcount} = $itemdamaged_count;
2279 $oldbiblio->{intransitcount} = $item_in_transit_count;
2280 $oldbiblio->{onholdcount} = $item_onhold_count;
2281 $oldbiblio->{orderedcount} = $ordered_count;
2282 $oldbiblio->{notforloancount} = $notforloan_count;
2284 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2285 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2286 my $subfields = substr $fieldspec, 3;
2287 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2288 my @alternateholdingsinfo = ();
2289 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2290 my $alternateholdingscount = 0;
2292 for my $field (@holdingsfields) {
2293 my %holding = ( holding => '' );
2294 my $havesubfield = 0;
2295 for my $subfield ($field->subfields()) {
2296 if ((index $subfields, $$subfield[0]) >= 0) {
2297 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2298 $holding{'holding'} .= $$subfield[1];
2302 if ($havesubfield) {
2303 push(@alternateholdingsinfo, \%holding);
2304 $alternateholdingscount++;
2308 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2309 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2312 push( @newresults, $oldbiblio );
2318 =head2 enabled_staff_search_views
2320 %hash = enabled_staff_search_views()
2322 This function returns a hash that contains three flags obtained from the system
2323 preferences, used to determine whether a particular staff search results view
2328 =item C<Output arg:>
2330 * $hash{can_view_MARC} is true only if the MARC view is enabled
2331 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2332 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2334 =item C<usage in the script:>
2338 $template->param ( C4::Search::enabled_staff_search_views );
2342 sub enabled_staff_search_views
2345 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2346 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2347 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2351 =head2 z3950_search_args
2353 $arrayref = z3950_search_args($matchpoints)
2355 This function returns an array reference that contains the search parameters to be
2356 passed to the Z39.50 search script (z3950_search.pl). The array elements
2357 are hash refs whose keys are name and value, and whose values are the
2358 name of a search parameter, the value of that search parameter and the URL encoded
2359 value of that parameter.
2361 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2363 The search parameter values are obtained from the bibliographic record whose
2364 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2366 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2367 a general purpose search argument. In this case, the returned array contains only
2368 entry: the key is 'title' and the value is derived from $matchpoints.
2370 If a search parameter value is undefined or empty, it is not included in the returned
2373 The returned array reference may be passed directly to the template parameters.
2377 =item C<Output arg:>
2379 * $array containing hash refs as described above
2381 =item C<usage in the script:>
2385 $data = Biblio::GetBiblioData($bibno);
2386 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2390 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2394 sub z3950_search_args {
2397 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2398 my $isbn = Business::ISBN->new( $isbn_string );
2400 if (defined $isbn && $isbn->is_valid)
2402 if ( ref($bibrec) ) {
2403 $bibrec->{isbn} = $isbn_string;
2404 $bibrec->{title} = undef;
2406 $bibrec = { isbn => $isbn_string };
2410 $bibrec = { title => $bibrec } if !ref $bibrec;
2413 for my $field (qw/ lccn isbn issn title author dewey subject /)
2415 push @$array, { name => $field, value => $bibrec->{$field} }
2416 if defined $bibrec->{$field};
2421 =head2 GetDistinctValues($field);
2423 C<$field> is a reference to the fields array
2427 sub GetDistinctValues {
2428 my ($fieldname,$string)=@_;
2429 # returns a reference to a hash of references to branches...
2430 if ($fieldname=~/\./){
2431 my ($table,$column)=split /\./, $fieldname;
2432 my $dbh = C4::Context->dbh;
2433 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2434 my $sth = $dbh->prepare("select DISTINCT($column) as value, count(*) as cnt from $table ".($string?" where $column like \"$string%\"":"")."group by value order by $column ");
2436 my $elements=$sth->fetchall_arrayref({});
2441 my @servers=qw<biblioserver authorityserver>;
2442 my (@zconns,@results);
2443 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2444 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2447 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2450 # The big moment: asynchronously retrieve results from all servers
2456 my ( $i, $size ) = @_;
2457 for ( my $j = 0 ; $j < $size ; $j++ ) {
2459 @hashscan{qw(value cnt)} =
2460 $results[ $i - 1 ]->display_term($j);
2461 push @elements, \%hashscan;
2469 =head2 _ZOOM_event_loop
2471 _ZOOM_event_loop(\@zconns, \@results, sub {
2472 my ( $i, $size ) = @_;
2476 Processes a ZOOM event loop and passes control to a closure for
2477 processing the results, and destroying the resultsets.
2481 sub _ZOOM_event_loop {
2482 my ($zconns, $results, $callback) = @_;
2483 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2484 my $ev = $zconns->[ $i - 1 ]->last_event();
2485 if ( $ev == ZOOM::Event::ZEND ) {
2486 next unless $results->[ $i - 1 ];
2487 my $size = $results->[ $i - 1 ]->size();
2489 $callback->($i, $size);
2494 foreach my $result (@$results) {
2499 =head2 new_record_from_zebra
2501 Given raw data from a searchengine result set, return a MARC::Record object
2503 This helper function is needed to take into account all the involved
2504 system preferences and configuration variables to properly create the
2505 MARC::Record object.
2507 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2508 data. If we are using DOM, then it has to be MARCXML.
2510 If we are using elasticsearch, it'll already be a MARC::Record and this
2511 function needs a new name.
2515 sub new_record_from_zebra {
2518 my $raw_data = shift;
2519 # Set the default indexing modes
2520 my $search_engine = C4::Context->preference("SearchEngine");
2521 if ($search_engine eq 'Elasticsearch') {
2522 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2524 my $index_mode = ( $server eq 'biblioserver' )
2525 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2526 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2528 my $marc_record = eval {
2529 if ( $index_mode eq 'dom' ) {
2530 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2532 MARC::Record->new_from_usmarc( $raw_data );
2539 return $marc_record;
2544 END { } # module clean-up code here (global destructor)
2551 Koha Development Team <http://koha-community.org/>