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>.
20 use C4::Biblio qw( TransformMarcToKoha GetMarcFromKohaField GetFrameworkCode GetAuthorisedValueDesc GetBiblioData );
21 use C4::Koha qw( getFacets GetVariationsOfISBN GetNormalizedUPC GetNormalizedEAN GetNormalizedOCLCNumber GetNormalizedISBN getitemtypeimagelocation );
24 use Koha::SearchEngine::QueryBuilder;
27 use C4::XSLT qw( XSLTParse4Display );
28 use C4::Reserves qw( GetReserveStatus );
29 use C4::Charset qw( SetUTF8Flag );
30 use Koha::AuthorisedValues;
36 use Koha::RecordProcessor;
37 use Koha::SearchFilters;
43 our (@ISA, @EXPORT_OK);
54 enabled_staff_search_views
63 C4::Search - Functions for searching the Koha catalog.
67 See opac/opac-search.pl or catalogue/search.pl for example of usage
71 This module provides searching functions for Koha's bibliographic databases
77 # make all your functions, whether exported or not;
81 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
83 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
89 my $dbh = C4::Context->dbh;
90 my $result = TransformMarcToKoha({ record => $record });
94 # search duplicate on ISBN, easy and fast..
96 if ( $result->{isbn} ) {
97 $result->{isbn} =~ s/\(.*$//;
98 $result->{isbn} =~ s/\s+$//;
99 $query = "isbn:$result->{isbn}";
103 my $titleindex = 'ti,ext';
104 my $authorindex = 'au,ext';
107 $result->{title} =~ s /\\//g;
108 $result->{title} =~ s /\"//g;
109 $result->{title} =~ s /\(//g;
110 $result->{title} =~ s /\)//g;
112 $query = "$titleindex:\"$result->{title}\"";
113 if ( $result->{author} ) {
114 $result->{author} =~ s /\\//g;
115 $result->{author} =~ s /\"//g;
116 $result->{author} =~ s /\(//g;
117 $result->{author} =~ s /\)//g;
119 $query .= " $op $authorindex:\"$result->{author}\"";
123 my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
124 my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50);
126 if (!defined $error) {
127 foreach my $possible_duplicate_record (@{$searchresults}) {
128 my $marcrecord = new_record_from_zebra(
130 $possible_duplicate_record
133 my $result = TransformMarcToKoha({ record => $marcrecord });
135 # FIXME :: why 2 $biblionumber ?
137 push @results, $result->{'biblionumber'};
138 push @results, $result->{'title'};
147 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
149 This function provides a simple search API on the bibliographic catalog
155 * $query can be a simple keyword or a complete CCL query
156 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
157 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
158 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
159 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
164 Returns an array consisting of three elements
165 * $error is undefined unless an error is detected
166 * $results is a reference to an array of records.
167 * $total_hits is the number of hits that would have been returned with no limit
169 If an error is returned the two other return elements are undefined. If error itself is undefined
170 the other two elements are always defined
172 =item C<usage in the script:>
176 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
178 if (defined $error) {
179 $template->param(query_error => $error);
180 warn "error: ".$error;
181 output_html_with_http_headers $input, $cookie, $template->output;
185 my $hits = @{$marcresults};
188 for my $r ( @{$marcresults} ) {
189 my $marcrecord = MARC::File::USMARC::decode($r);
190 my $biblio = TransformMarcToKoha({ record => $marcrecord });
192 #build the iarray of hashs for the template.
194 title => $biblio->{'title'},
195 subtitle => $biblio->{'subtitle'},
196 biblionumber => $biblio->{'biblionumber'},
197 author => $biblio->{'author'},
198 publishercode => $biblio->{'publishercode'},
199 publicationyear => $biblio->{'publicationyear'},
204 $template->param(result=>\@results);
209 my ( $query, $offset, $max_results, $servers, %options ) = @_;
211 return ( 'No query entered', undef, undef ) unless $query;
212 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
213 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
220 # Initialize & Search Zebra
221 for ( my $i = 0 ; $i < @servers ; $i++ ) {
223 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
224 $query =~ s/:/=/g unless $options{skip_normalize};
225 $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]);
226 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
230 $zconns[$i]->errmsg() . " ("
231 . $zconns[$i]->errcode() . ") "
232 . $zconns[$i]->addinfo() . " "
233 . $zconns[$i]->diagset();
235 return ( $error, undef, undef ) if $zconns[$i]->errcode();
239 # caught a ZOOM::Exception
243 . $@->addinfo() . " "
245 warn $error." for query: $query";
246 return ( $error, undef, undef );
255 my $first_record = defined($offset) ? $offset + 1 : 1;
256 my $hits = $tmpresults[ $i - 1 ]->size();
257 $total_hits += $hits;
258 my $last_record = $hits;
259 if ( defined $max_results && $offset + $max_results < $hits ) {
260 $last_record = $offset + $max_results;
263 for my $j ( $first_record .. $last_record ) {
265 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
268 push @{$results}, $record if defined $record;
273 foreach my $zoom_query (@zoom_queries) {
274 $zoom_query->destroy();
277 return ( undef, $results, $total_hits );
282 ( undef, $results_hashref, \@facets_loop ) = getRecords (
284 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
285 $results_per_page, $offset, $branches, $itemtypes,
286 $query_type, $scan, $opac
289 The all singing, all dancing, multi-server, asynchronous, scanning,
290 searching, record nabbing, facet-building
292 See verbose embedded documentation.
298 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
299 $results_per_page, $offset, $branches, $itemtypes,
300 $query_type, $scan, $opac
303 my @servers = @$servers_ref;
304 my @sort_by = @$sort_by_ref;
305 $offset = 0 if $offset < 0;
307 # Initialize variables for the ZOOM connection and results object
310 my $results_hashref = ();
312 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
313 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search->as_list };
315 # Initialize variables for the faceted results objects
316 my $facets_counter = {};
317 my $facets_info = {};
318 my $facets = getFacets();
320 my @facets_loop; # stores the ref to array of hashes for template facets loop
322 ### LOOP THROUGH THE SERVERS
323 for ( my $i = 0 ; $i < @servers ; $i++ ) {
324 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
326 # perform the search, create the results objects
327 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
328 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
330 Koha::Logger->get->debug($simple_query) if $scan;
332 # Check if we've got a query_type defined, if so, use it
335 if ($query_type =~ /^ccl/) {
336 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
337 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
338 } elsif ($query_type =~ /^cql/) {
339 $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i]));
340 } elsif ($query_type =~ /^pqf/) {
341 $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i]));
343 warn "Unknown query_type '$query_type'. Results undetermined.";
346 $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
348 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
352 warn "WARNING: query problem with $query_to_use " . $@;
355 # Concatenate the sort_by limits and pass them to the results object
356 # Note: sort will override rank
358 foreach my $sort (@sort_by) {
359 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
360 $sort_by .= "1=1003 <i ";
362 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
363 $sort_by .= "1=1003 >i ";
365 elsif ( $sort eq "popularity_asc" ) {
366 $sort_by .= "1=9003 <i ";
368 elsif ( $sort eq "popularity_dsc" ) {
369 $sort_by .= "1=9003 >i ";
371 elsif ( $sort eq "call_number_asc" ) {
372 $sort_by .= "1=8007 <i ";
374 elsif ( $sort eq "call_number_dsc" ) {
375 $sort_by .= "1=8007 >i ";
377 elsif ( $sort eq "pubdate_asc" ) {
378 $sort_by .= "1=31 <i ";
380 elsif ( $sort eq "pubdate_dsc" ) {
381 $sort_by .= "1=31 >i ";
383 elsif ( $sort eq "acqdate_asc" ) {
384 $sort_by .= "1=32 <i ";
386 elsif ( $sort eq "acqdate_dsc" ) {
387 $sort_by .= "1=32 >i ";
389 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
390 $sort_by .= "1=4 <i ";
392 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
393 $sort_by .= "1=4 >i ";
395 elsif ( $sort eq "biblionumber_az" || $sort eq "biblionumber_asc" ) {
396 $sort_by .= "1=12 <i ";
398 elsif ( $sort eq "biblionumber_za" || $sort eq "biblionumber_dsc" ) {
399 $sort_by .= "1=12 >i ";
402 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
405 if ( $sort_by && !$scan && $results[$i] ) {
406 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
407 warn "WARNING sort $sort_by failed";
410 } # finished looping through servers
412 # The big moment: asynchronously retrieve results from all servers
417 my ( $i, $size ) = @_;
420 # loop through the results
421 $results_hash->{'hits'} = $size;
423 if ( $offset + $results_per_page <= $size ) {
424 $times = $offset + $results_per_page;
430 for ( my $j = $offset ; $j < $times ; $j++ ) {
433 ## Check if it's an index scan
435 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
437 # here we create a minimal MARC record and hand it off to the
438 # template just like a normal result ... perhaps not ideal, but
440 my $tmprecord = MARC::Record->new();
441 $tmprecord->encoding('UTF-8');
445 # the minimal record in author/title (depending on MARC flavour)
446 if ( C4::Context->preference("marcflavour") eq
449 $tmptitle = MARC::Field->new(
454 $tmprecord->append_fields($tmptitle);
458 MARC::Field->new( '245', ' ', ' ', a => $term, );
460 MARC::Field->new( '100', ' ', ' ', a => $occ, );
461 $tmprecord->append_fields($tmptitle);
462 $tmprecord->append_fields($tmpauthor);
464 $results_hash->{'RECORDS'}[$j] =
465 $tmprecord->as_usmarc();
470 $record = $results[ $i - 1 ]->record($j)->raw();
471 # warn "RECORD $j:".$record;
472 $results_hash->{'RECORDS'}[$j] = $record;
476 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
478 # Fill the facets while we're looping, but only for the
479 # biblioserver and not for a scan
480 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
481 $facets_counter = GetFacets( $results[ $i - 1 ] );
482 $facets_info = _get_facets_info( $facets );
486 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
488 sort { $a cmp $b } keys %$facets_counter
491 my @this_facets_array;
494 $facets_counter->{$link_value}
495 ->{$b} <=> $facets_counter->{$link_value}
497 } keys %{ $facets_counter->{$link_value} }
500 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
501 my $facet_link_value = $one_facet;
502 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
504 # fix the length that will display in the label,
505 my $facet_label_value = $one_facet;
506 my $facet_max_length = C4::Context->preference(
507 'FacetLabelTruncationLength')
510 substr( $one_facet, 0, $facet_max_length )
512 if length($facet_label_value) >
515 # if it's a branch, label by the name, not the code,
516 if ( $link_value =~ /branch/ ) {
517 if ( defined $branches
518 && ref($branches) eq "HASH"
519 && defined $branches->{$one_facet}
520 && ref( $branches->{$one_facet} ) eq
524 $branches->{$one_facet}
528 $facet_label_value = "*";
532 # if it's a itemtype, label by the name, not the code,
533 if ( $link_value =~ /itype/ ) {
534 if ( defined $itemtypes
535 && ref($itemtypes) eq "HASH"
536 && defined $itemtypes->{$one_facet}
537 && ref( $itemtypes->{$one_facet} ) eq
541 $itemtypes->{$one_facet}
542 ->{translated_description};
546 # also, if it's a location code, use the name instead of the code
547 if ( $link_value =~ /location/ ) {
548 # TODO Retrieve all authorised values at once, instead of 1 query per entry
549 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
550 $facet_label_value = $av->count ? $av->next->opac_description : '';
553 # also, if it's a collection code, use the name instead of the code
554 if ( $link_value =~ /ccode/ ) {
555 # TODO Retrieve all authorised values at once, instead of 1 query per entry
556 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
557 $facet_label_value = $av->count ? $av->next->opac_description : '';
560 # but we're down with the whole label being in the link's title.
561 push @this_facets_array,
564 $facets_counter->{$link_value}
566 facet_label_value => $facet_label_value,
567 facet_title_value => $one_facet,
568 facet_link_value => $facet_link_value,
569 type_link_value => $link_value,
571 if ($facet_label_value);
576 type_link_value => $link_value,
577 type_id => $link_value . "_id",
579 . $facets_info->{$link_value}->{'label_value'} =>
581 facets => \@this_facets_array,
585 $facets_info->{$link_value}->{'label_value'} =~
588 and ( Koha::Libraries->search->count == 1 )
595 # This sorts the facets into alphabetical order
597 foreach my $f (@facets_loop) {
598 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
600 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
605 return ( undef, $results_hashref, \@facets_loop );
613 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
615 if ( $use_zebra_facets ) {
616 $facets = _get_facets_from_zebra( $rs );
618 $facets = _get_facets_from_records( $rs );
624 sub _get_facets_from_records {
628 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
629 my $facets_config = getFacets();
631 my $size = $rs->size();
632 my $jmax = $size > $facets_maxrecs
636 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
638 my $marc_record = new_record_from_zebra (
640 $rs->record( $j )->raw()
643 if ( ! defined $marc_record ) {
644 warn "ERROR DECODING RECORD - $@: " .
645 $rs->record( $j )->raw();
649 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
655 =head2 _get_facets_data_from_record
657 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
659 Internal function that extracts facets information from a MARC::Record object
660 and populates $facets_counter for using in getRecords.
662 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
667 sub _get_facets_data_from_record {
669 my ( $marc_record, $facets, $facets_counter ) = @_;
671 for my $facet (@$facets) {
675 foreach my $tag ( @{ $facet->{ tags } } ) {
677 # tag number is the first three digits
678 my $tag_num = substr( $tag, 0, 3 );
679 # subfields are the remainder
680 my $subfield_letters = substr( $tag, 3 );
682 my @fields = $marc_record->field( $tag_num );
683 foreach my $field (@fields) {
684 # If $field->indicator(1) eq 'z', it means it is a 'see from'
685 # field introduced because of IncludeSeeFromInSearches, so skip it
686 next if $field->indicator(1) eq 'z';
688 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
689 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
691 unless ( grep { $_ eq $data } @used_datas ) {
692 push @used_datas, $data;
693 $facets_counter->{ $facet->{ idx } }->{ $data }++;
700 =head2 _get_facets_from_zebra
702 my $facets = _get_facets_from_zebra( $result_set )
704 Retrieves facets for a specified result set. It loops through the facets defined
705 in C4::Koha::getFacets and returns a hash with the following structure:
715 sub _get_facets_from_zebra {
719 # save current elementSetName
720 my $elementSetName = $rs->option( 'elementSetName' );
722 my $facets_loop = getFacets();
723 my $facets_data = {};
724 # loop through defined facets and fill the facets hashref
725 foreach my $facet ( @$facets_loop ) {
727 my $idx = $facet->{ idx };
728 my $sep = $facet->{ sep };
729 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
730 if ( $facet_values ) {
731 # we've actually got a result
732 $facets_data->{ $idx } = $facet_values;
735 # set elementSetName to its previous value to avoid side effects
736 $rs->option( elementSetName => $elementSetName );
741 =head2 _get_facet_from_result_set
744 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
746 Internal function that extracts facet information for a specific index ($facet_idx) and
747 returns a hash containing facet values and count:
750 $facet_value => $count ,
754 Warning: this function has the side effect of changing the elementSetName for the result
755 set. It is a helper function for the main loop, which takes care of backing it up for
760 sub _get_facet_from_result_set {
762 my $facet_idx = shift;
766 my $internal_sep = '<*>';
767 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
769 return if ( ! defined $facet_idx || ! defined $rs );
770 # zebra's facet element, untokenized index
771 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
772 # configure zebra results for retrieving the desired facet
773 $rs->option( elementSetName => $facet_element );
774 # get the facet record from result set
775 my $facet = $rs->record( 0 )->raw;
776 # if the facet has no restuls...
777 return if !defined $facet;
778 # TODO: benchmark DOM vs. SAX performance
779 my $facet_dom = XML::LibXML->load_xml(
782 my @terms = $facet_dom->getElementsByTagName('term');
786 foreach my $term ( @terms ) {
787 my $facet_value = $term->textContent;
788 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
789 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
790 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
796 =head2 _get_facets_info
798 my $facets_info = C4::Search::_get_facets_info( $facets )
800 Internal function that extracts facets information and properly builds
801 the data structure needed to render facet labels.
805 sub _get_facets_info {
809 my $facets_info = {};
811 for my $facet ( @$facets ) {
812 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
819 sub _detect_truncation {
820 my ( $operand, $index ) = @_;
821 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
824 my @wordlist = split( /\s/, $operand );
825 foreach my $word (@wordlist) {
826 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
827 push @rightlefttruncated, $word;
829 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
830 push @lefttruncated, $word;
832 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
833 push @righttruncated, $word;
835 elsif ( index( $word, "*" ) < 0 ) {
836 push @nontruncated, $word;
839 push @regexpr, $word;
843 \@nontruncated, \@righttruncated, \@lefttruncated,
844 \@rightlefttruncated, \@regexpr
849 sub _build_stemmed_operand {
850 my ($operand,$lang) = @_;
851 require Lingua::Stem::Snowball ;
852 my $stemmed_operand=q{};
854 # Stemmer needs language
855 return $operand unless $lang;
857 # If operand contains a digit, it is almost certainly an identifier, and should
858 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
859 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
860 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
861 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
862 return $operand if $operand =~ /\d/;
864 # FIXME: the locale should be set based on the user's language and/or search choice
866 # Make sure we only use the first two letters from the language code
867 $lang = lc(substr($lang, 0, 2));
868 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
869 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
870 if ($lang eq 'nb' || $lang eq 'nn') {
873 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
874 encoding => "UTF-8" );
876 my @words = split( / /, $operand );
877 my @stems = $stemmer->stem(\@words);
878 for my $stem (@stems) {
879 $stemmed_operand .= "$stem";
880 $stemmed_operand .= "?"
881 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
882 $stemmed_operand .= " ";
885 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
886 return $stemmed_operand;
890 sub _build_weighted_query {
892 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
893 # pretty well but could work much better if we had a smarter query parser
894 my ( $operand, $stemmed_operand, $index ) = @_;
895 my $stemming = C4::Context->preference("QueryStemming") || 0;
896 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
897 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
898 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
900 my $weighted_query = "(rk=("; # Specifies that we're applying rank
902 # Keyword, or, no index specified
903 if ( ( $index eq 'kw' ) || ( !$index ) ) {
905 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
906 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
907 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
908 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
909 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
910 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
911 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
912 if $fuzzy_enabled; # add fuzzy, word list
913 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
914 if ( $stemming and $stemmed_operand )
915 ; # add stemming, right truncation
916 $weighted_query .= " or wrdl,r9=\"$operand\"";
918 # embedded sorting: 0 a-z; 1 z-a
919 # $weighted_query .= ") or (sort1,aut=1";
922 # Barcode searches should skip this process
923 elsif ( $index eq 'bc' ) {
924 $weighted_query .= "bc=\"$operand\"";
927 # Authority-number searches should skip this process
928 elsif ( $index eq 'an' ) {
929 $weighted_query .= "an=\"$operand\"";
932 # If the index is numeric, don't autoquote it.
933 elsif ( $index =~ /,st-numeric$/ ) {
934 $weighted_query .= " $index=$operand";
937 # If the index already has more than one qualifier, wrap the operand
938 # in quotes and pass it back (assumption is that the user knows what they
939 # are doing and won't appreciate us mucking up their query
940 elsif ( $index =~ ',' ) {
941 $weighted_query .= " $index=\"$operand\"";
944 #TODO: build better cases based on specific search indexes
946 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
947 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
948 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
949 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
950 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
951 if $fuzzy_enabled; # add fuzzy, word list
952 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
953 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
956 $weighted_query .= "))"; # close rank specification
957 return $weighted_query;
962 Return an array with available indexes.
986 'Author-personal-bibliography',
996 'Chronological-subdivision',
1006 'Conference-name-heading',
1007 'Conference-name-see',
1008 'Conference-name-seealso',
1011 'Control-number-identifier',
1015 'Corporate-name-heading',
1016 'Corporate-name-see',
1017 'Corporate-name-seealso',
1018 'Country-publication',
1021 'date-entered-on-file',
1022 'Date-of-acquisition',
1023 'Date-of-publication',
1024 'Date-time-last-modified',
1025 'Dewey-classification',
1026 'Dissertation-information',
1035 'Geographic-subdivision',
1038 'Heading-use-main-or-added-entry',
1039 'Heading-use-series-added-entry ',
1040 'Heading-use-subject-added-entry',
1044 'Identifier-standard',
1045 'Illustration-code',
1047 'Index-term-uncontrolled',
1048 'Interest-age-level',
1049 'Interest-grade-level',
1059 'language-original',
1069 'Local-classification',
1072 'Match-heading-see-from',
1077 'Multipart-resource-level',
1082 'Name-geographic-heading',
1083 'Name-geographic-see',
1084 'Name-geographic-seealso',
1090 'Other-control-number',
1093 'Personal-name-heading',
1094 'Personal-name-see',
1095 'Personal-name-seealso',
1097 'Place-publication',
1104 'Reading-grade-level',
1105 'Record-control-number',
1116 'Subject-heading-thesaurus',
1117 'Subject-name-personal',
1118 'Subject-subdivision',
1127 'Term-genre-form-heading',
1128 'Term-genre-form-see',
1129 'Term-genre-form-seealso',
1135 'Title-uniform-heading',
1136 'Title-uniform-see',
1137 'Title-uniform-seealso',
1147 'classification-source',
1149 'coded-location-qualifier',
1160 'Local-classification',
1163 'materials-specified',
1168 'Number-local-acquisition',
1173 'replacementpricedate',
1191 $simple_query, $query_cgi,
1192 $query_desc, $limit,
1193 $limit_cgi, $limit_desc,
1194 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1196 Build queries and limits in CCL, CGI, Human,
1197 handle truncation, stemming, field weighting, fuzziness, etc.
1199 See verbose embedded documentation.
1205 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1210 my @operators = $operators ? @$operators : ();
1211 my @indexes = $indexes ? @$indexes : ();
1212 my @operands = $operands ? @$operands : ();
1213 my @limits = $limits ? @$limits : ();
1214 my @sort_by = $sort_by ? @$sort_by : ();
1216 my $stemming = C4::Context->preference("QueryStemming") || 0;
1217 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1218 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1219 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1221 my $query = $operands[0] // "";
1222 my $simple_query = $operands[0];
1224 # initialize the variables we're passing back
1233 my $cclindexes = getIndexes();
1234 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1235 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1237 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1239 $query = "ccl=$query" if $cclq;
1243 my %group_OR_limits;
1244 my $availability_limit;
1245 foreach my $this_limit (@limits) {
1246 next unless $this_limit;
1247 if ( $this_limit =~ /available/ ) {
1249 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1251 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1252 $availability_limit .=
1253 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1254 $limit_cgi .= "&limit=available";
1258 # group_OR_limits, prefixed by mc-
1259 # OR every member of the group
1260 elsif ( $this_limit =~ /mc/ ) {
1261 my ($k,$v) = split(/:/, $this_limit,2);
1262 if ( $k !~ /mc-i(tem)?type/ ) {
1263 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1264 $this_limit =~ tr/"//d;
1265 $this_limit = $k.':"'.$v.'"';
1268 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1269 $limit_desc .= " or " if $group_OR_limits{$k};
1270 $group_OR_limits{$k} .= "$this_limit";
1271 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1272 $limit_desc .= " $this_limit";
1274 elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) {
1275 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1276 $limit .= " and " if $limit || $query;
1277 my $branchfield = C4::Context->preference('SearchLimitLibrary');
1279 if( $this_limit =~ '^multibranchlimit:' ){
1280 my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ );
1281 my $search_group = Koha::Library::Groups->find( $group_id );
1282 @branchcodes = map { $_->branchcode } $search_group->all_libraries;
1283 @branchcodes = sort { $a cmp $b } @branchcodes;
1285 @branchcodes = ( $this_limit =~ /^branch:(.*)$/ );
1289 if ( $branchfield eq "homebranch" ) {
1290 $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes;
1292 elsif ( $branchfield eq "holdingbranch" ) {
1293 $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes;
1296 $this_limit = sprintf "(%s or %s)",
1297 join( " or ", map { 'homebranch: ' . $_ } @branchcodes ),
1298 join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes );
1301 $limit .= "$this_limit";
1302 $limit_desc .= " $this_limit";
1303 } elsif ( $this_limit =~ '^search_filter:' ) {
1304 # Here we will get the query as a string, append to the limits, and pass through buildQuery
1305 # again to clean the terms and handle nested filters
1306 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1307 my ($filter_id) = ( $this_limit =~ /^search_filter:(.*)$/ );
1308 my $search_filter = Koha::SearchFilters->find( $filter_id );
1309 next unless $search_filter;
1310 my ($expanded_lim, $query_lim) = $search_filter->expand_filter;
1311 push @$expanded_lim, $query_lim;
1312 my ( $error, undef, undef, undef, undef, $fixed_limit, undef, undef, undef ) = buildQuery ( undef, undef, undef, $expanded_lim, undef, undef, $lang);
1313 $limit .= " and " if $limit || $query;
1314 $limit .= "$fixed_limit";
1315 $limit_desc .= " $limit";
1318 # Regular old limits
1320 $limit .= " and " if $limit || $query;
1321 $limit .= "$this_limit";
1322 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1323 $limit_desc .= " $this_limit";
1326 foreach my $k (keys (%group_OR_limits)) {
1327 $limit .= " and " if ( $query || $limit );
1328 $limit .= "($group_OR_limits{$k})";
1330 if ($availability_limit) {
1331 $limit .= " and " if ( $query || $limit );
1332 $limit .= "($availability_limit)";
1335 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1337 if ( $query =~ /^ccl=/ ) {
1339 # This is needed otherwise ccl= and &limit won't work together, and
1340 # this happens when selecting a subject on the opac-detail page
1341 my $original_q = $q; # without available part
1342 $q .= $limit if $limit;
1343 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1345 if ( $query =~ /^cql=/ ) {
1346 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1348 if ( $query =~ /^pqf=/ ) {
1350 $query_cgi = "q=pqf=".uri_escape_utf8($');
1351 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1354 # pass nested queries directly
1355 # FIXME: need better handling of some of these variables in this case
1356 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1357 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1358 # if ( $query =~ /(\(|\))/ ) {
1360 # undef, $query, $simple_query, $query_cgi,
1361 # $query, $limit, $limit_cgi, $limit_desc,
1366 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1367 # query operands and indexes and add stemming, truncation, field weighting, etc.
1368 # Once we do so, we'll end up with a value in $query, just like if we had an
1369 # incoming $query from the user
1372 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1373 my $previous_operand
1374 ; # a flag used to keep track if there was a previous query
1375 # if there was, we can apply the current operator
1377 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1379 # COMBINE OPERANDS, INDEXES AND OPERATORS
1380 if ( ($operands[$i] // '') ne '' ) {
1381 $operands[$i]=~s/^\s+//;
1383 # A flag to determine whether or not to add the index to the query
1386 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1387 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1391 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1393 my $operand = $operands[$i];
1394 my $index = $indexes[$i] || 'kw';
1396 # Add index-specific attributes
1398 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1399 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1400 #which is processed higher up in this sub. Other than that, year searches are typically
1401 #handled as limits which are not processed her either.
1403 # Search ranges: Date of Publication, st-numeric
1404 if ( $index =~ /(yr|st-numeric)/ ) {
1405 #weight_fields/relevance search causes errors with date ranges
1406 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1407 #In the case of YYYY-YYYY, it will return no results
1408 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1411 # Date of Acquisition
1412 elsif ( $index =~ /acqdate/ ) {
1413 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1414 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1415 #top of the results just because they have lots of item records matching that date.
1416 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1418 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1420 # ISBN,ISSN,Standard Number, don't need special treatment
1421 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1423 $stemming, $auto_truncation,
1424 $weight_fields, $fuzzy_enabled
1427 if ( $index eq 'nb' ) {
1428 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1429 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1430 $operands[$i] = $operand = '(' . join( ' OR ', map { 'nb=' . $_ } @isbns ) . ')';
1431 $indexes[$i] = $index = 'kw';
1434 if ( $index eq 'ns' ) {
1435 if ( C4::Context->preference("SearchWithISSNVariations") ) {
1436 my @issns = C4::Koha::GetVariationsOfISSN( $operand );
1437 $operands[$i] = $operand = '(' . join( ' OR ', map { 'ns=' . $_ } @issns ) . ')';
1438 $indexes[$i] = $index = 'kw';
1443 # Set default structure attribute (word list)
1444 my $struct_attr = q{};
1445 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1446 $struct_attr = ",wrdl";
1449 # Some helpful index variants
1450 my $index_plus = $index . $struct_attr . ':';
1451 my $index_plus_comma = $index . $struct_attr . ',';
1453 if ($auto_truncation){
1454 unless ( $index =~ /,(st-|phr|ext)/ ) {
1455 #FIXME only valid with LTR scripts
1456 $operand=join(" ",map{
1457 (index($_,"*")>0?"$_":"$_*")
1458 }split (/\s+/,$operand));
1463 my $truncated_operand = q{};
1464 my( $nontruncated, $righttruncated, $lefttruncated,
1465 $rightlefttruncated, $regexpr
1466 ) = _detect_truncation( $operand, $index );
1468 Koha::Logger->get->debug(
1469 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1473 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1474 scalar(@$rightlefttruncated) > 0 )
1477 # Don't field weight or add the index to the query, we do it here
1479 undef $weight_fields;
1480 my $previous_truncation_operand;
1481 if (scalar @$nontruncated) {
1482 $truncated_operand .= "$index_plus @$nontruncated ";
1483 $previous_truncation_operand = 1;
1485 if (scalar @$righttruncated) {
1486 $truncated_operand .= "and " if $previous_truncation_operand;
1487 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1488 $previous_truncation_operand = 1;
1490 if (scalar @$lefttruncated) {
1491 $truncated_operand .= "and " if $previous_truncation_operand;
1492 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1493 $previous_truncation_operand = 1;
1495 if (scalar @$rightlefttruncated) {
1496 $truncated_operand .= "and " if $previous_truncation_operand;
1497 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1498 $previous_truncation_operand = 1;
1501 $operand = $truncated_operand if $truncated_operand;
1502 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1505 my $stemmed_operand = q{};
1506 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1509 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1511 # Handle Field Weighting
1512 my $weighted_operand = q{};
1513 if ($weight_fields) {
1514 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1515 $operand = $weighted_operand;
1519 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1521 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1523 #N.B. Truncation is mutually exclusive with Weighted Queries,
1524 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1525 #the need for this relevance wrapper.
1526 $operand = "(rk=($operand))" unless $weight_fields;
1528 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1530 query_cgi => $query_cgi,
1531 query_desc => $query_desc,
1532 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1533 parsed_operand => $operand,
1534 original_operand => $operands[$i] // '',
1536 index_plus => $index_plus,
1537 indexes_set => $indexes_set,
1538 previous_operand => $previous_operand,
1544 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1546 # Normalize the query and limit strings
1547 # This is flawed , means we can't search anything with : in it
1548 # if user wants to do ccl or cql, start the query with that
1549 # $query =~ s/:/=/g;
1550 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1551 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1552 $query =~ s/(?<=(wrdl)):/=/g;
1553 $query =~ s/(?<=(trn|phr)):/=/g;
1554 $query =~ s/(?<=(st-numeric)):/=/g;
1555 $query =~ s/(?<=(st-year)):/=/g;
1556 $query =~ s/(?<=(st-date-normalized)):/=/g;
1558 # Removing warnings for later substitutions
1560 $query_desc //= q{};
1563 $limit_desc //= q{};
1565 $simple_query //= q{};
1567 for ( $query, $query_desc, $limit, $limit_desc ) {
1568 s/ +/ /g; # remove extra spaces
1569 s/^ //g; # remove any beginning spaces
1570 s/ $//g; # remove any ending spaces
1571 s/==/=/g; # remove double == from query
1573 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1575 for ($query_cgi,$simple_query) {
1578 # append the limit to the query
1579 $query .= " " . $limit;
1581 Koha::Logger->get->debug(
1582 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1583 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1586 undef, $query, $simple_query, $query_cgi,
1587 $query_desc, $limit, $limit_cgi, $limit_desc,
1592 =head2 _build_initial_query
1594 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1596 Build a section of the initial query containing indexes, operators, and operands.
1600 sub _build_initial_query {
1604 if ($params->{previous_operand}){
1605 #If there is a previous operand, add a supplied operator or the default 'and'
1606 $operator = ($params->{operator}) ? ($params->{operator}) : 'AND';
1609 #NOTE: indexes_set is typically set when doing truncation or field weighting
1610 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1612 #e.g. "kw,wrdl:test"
1613 #e.g. " and kw,wrdl:test"
1614 $params->{query} .= " " . $operator . " " . $operand;
1616 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1617 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1618 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if ( $params->{original_operand} ne '' );
1620 #e.g. " and kw,wrdl: test"
1621 $params->{query_desc} .= " " . $operator . " " . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1623 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1625 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1628 =head2 searchResults
1630 my @search_results = searchResults($search_context, $searchdesc, $hits,
1631 $results_per_page, $offset, $scan,
1634 Format results in a form suitable for passing to the template
1638 # IMO this subroutine is pretty messy still -- it's responsible for
1639 # building the HTML output for the template
1641 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1642 my $dbh = C4::Context->dbh;
1647 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1648 my ($is_opac, $hidelostitems);
1649 if ($search_context->{'interface'} eq 'opac') {
1650 $hidelostitems = C4::Context->preference('hidelostitems');
1654 my $record_processor = Koha::RecordProcessor->new({
1655 filters => 'ViewPolicy'
1658 #Build branchnames hash
1659 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' })->as_list;
1661 # FIXME - We build an authorised values hash here, using the default framework
1662 # though it is possible to have different authvals for different fws.
1664 my $shelflocations =
1665 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1667 # get notforloan authorised value list (see $shelflocations FIXME)
1668 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1669 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1672 my $itemtypes = Koha::ItemTypes->search_with_localization;
1673 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1675 #search item field code
1676 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1678 ## find column names of items related to MARC
1679 my %subfieldstosearch;
1680 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1681 for my $column ( @columns ) {
1682 my ( $tagfield, $tagsubfield ) =
1683 &GetMarcFromKohaField( "items." . $column );
1684 if ( defined $tagsubfield ) {
1685 $subfieldstosearch{$column} = $tagsubfield;
1689 # handle which records to actually retrieve
1690 my $times; # Times is which record to process up to
1691 if ( $hits && $offset + $results_per_page <= $hits ) {
1692 $times = $offset + $results_per_page;
1695 $times = $hits; # If less hits than results_per_page+offset we go to the end
1698 my $marcflavour = C4::Context->preference("marcflavour");
1699 # We get the biblionumber position in MARC
1700 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1702 # set stuff for XSLT processing here once, not later again for every record we retrieved
1704 my $userenv = C4::Context->userenv;
1706 = ( defined $userenv and $userenv->{number} )
1707 ? Koha::Patrons->find( $userenv->{number} )
1709 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1711 # loop through all of the records we've retrieved
1712 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1716 # For Scan searches we built USMARC data
1717 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1719 # Normal search, render from Zebra's output
1720 $marcrecord = new_record_from_zebra(
1725 if ( ! defined $marcrecord ) {
1726 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1734 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1735 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1737 SetUTF8Flag($marcrecord);
1738 my $oldbiblio = TransformMarcToKoha({ record => $marcrecord, limit_table => 'no_items' });
1739 $oldbiblio->{result_number} = $i + 1;
1741 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1742 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1743 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1744 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1745 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1747 # edition information, if any
1748 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1750 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1751 # add imageurl to itemtype if there is one
1752 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1753 # Build summary if there is one (the summary is defined in the itemtypes table)
1754 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1756 # Pull out the items fields
1757 my @fields = $marcrecord->field($itemtag);
1758 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1759 my $marcflavor = C4::Context->preference("marcflavour");
1761 # adding linked items that belong to host records
1762 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1763 my $analyticsfield = '773';
1764 if ($marcflavor eq 'MARC21') {
1765 $analyticsfield = '773';
1766 } elsif ($marcflavor eq 'UNIMARC') {
1767 $analyticsfield = '461';
1769 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1770 my $hostbiblionumber = $hostfield->subfield("0");
1771 my $linkeditemnumber = $hostfield->subfield("9");
1772 if( $hostbiblionumber ) {
1773 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1774 if ($linkeditemmarc) {
1775 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1776 if ($linkeditemfield) {
1777 push( @fields, $linkeditemfield );
1784 # Setting item statuses for display
1785 my @available_items_loop;
1786 my @onloan_items_loop;
1787 my @other_items_loop;
1789 my $available_items;
1793 my $ordered_count = 0;
1794 my $available_count = 0;
1795 my $onloan_count = 0;
1796 my $longoverdue_count = 0;
1797 my $other_count = 0;
1798 my $withdrawn_count = 0;
1799 my $itemlost_count = 0;
1800 my $hideatopac_count = 0;
1801 my $itembinding_count = 0;
1802 my $itemdamaged_count = 0;
1803 my $item_in_transit_count = 0;
1804 my $item_onhold_count = 0;
1805 my $notforloan_count = 0;
1806 my $item_recalled_count = 0;
1807 my $items_count = scalar(@fields);
1808 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1809 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1810 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1812 # loop through every item
1813 foreach my $field (@fields) {
1816 # populate the items hash
1817 foreach my $code ( keys %subfieldstosearch ) {
1818 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1821 unless ( $item->{itemnumber} ) {
1822 warn "MARC item without itemnumber retrieved for biblio ($oldbiblio->{biblionumber})";
1826 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1830 # hidden based on OpacHiddenItems syspref or because lost
1831 my $hi = Koha::Items->search( { itemnumber => $item->{itemnumber} } )
1832 ->filter_by_visible_in_opac({ patron => $search_context->{patron} });
1833 unless ( $hi->count ) {
1834 push @hiddenitems, $item->{itemnumber};
1835 $hideatopac_count++;
1840 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1841 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1843 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1844 if ($item->{$hbranch}) {
1845 $item->{'branchname'} = $branches{$item->{$hbranch}};
1847 elsif ($item->{$otherbranch}) { # Last resort
1848 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1852 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1853 . ( $item->{location} ? $item->{location} : q{} )
1854 . ( $item->{itype} ? $item->{itype} : q{} )
1855 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1856 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1857 if ( $item->{onloan}
1859 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1862 my $key = $prefix . $item->{onloan} . $item->{barcode};
1863 $onloan_items->{$key}->{due_date} = $item->{onloan};
1864 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1865 $onloan_items->{$key}->{branchname} = $item->{branchname};
1866 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1867 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1868 $onloan_items->{$key}->{description} = $item->{description};
1869 $onloan_items->{$key}->{imageurl} =
1870 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1872 # if something's checked out and lost, mark it as 'long overdue'
1873 if ( $item->{itemlost} ) {
1874 $onloan_items->{$key}->{longoverdue}++;
1875 $longoverdue_count++;
1879 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1882 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1883 $item->{notforloan} = 1 if !$item->{notforloan} &&
1884 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1887 if ( $item->{notforloan} < 0 ) {
1889 } elsif ( $item->{notforloan} > 0 ) {
1890 $notforloan_count++;
1893 # is item in transit?
1894 my $transfertwhen = '';
1895 my ($transfertfrom, $transfertto);
1897 # is item on the reserve shelf?
1898 my $reservestatus = '';
1900 # is item a waiting recall?
1901 my $recallstatus = '';
1903 unless ($item->{withdrawn}
1904 || $item->{itemlost}
1906 || $item->{notforloan}
1907 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1908 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1910 # A couple heuristics to limit how many times
1911 # we query the database for item transfer information, sacrificing
1912 # accuracy in some cases for speed;
1914 # 1. don't query if item has one of the other statuses
1915 # 2. don't check transit status if the bib has
1916 # more than 20 items
1918 # FIXME: to avoid having the query the database like this, and to make
1919 # the in transit status count as unavailable for search limiting,
1920 # should map transit status to record indexed in Zebra.
1922 my $item_object = Koha::Items->find($item->{itemnumber});
1923 my $transfer = defined($item_object) ? $item_object->get_transfer : undef;
1924 ( $transfertwhen, $transfertfrom, $transfertto ) =
1927 $transfer->datesent, $transfer->frombranch,
1931 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1932 if ( C4::Context->preference('UseRecalls') ) {
1933 if ( Koha::Recalls->search({ item_id => $item->{itemnumber}, status => 'waiting' })->count ) {
1934 $recallstatus = 'Waiting';
1939 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1940 if ( $item->{withdrawn}
1941 || $item->{itemlost}
1943 || $item->{notforloan}
1944 || $reservestatus eq 'Waiting'
1945 || $recallstatus eq 'Waiting'
1946 || ($transfertwhen && $transfertwhen ne ''))
1948 $withdrawn_count++ if $item->{withdrawn};
1949 $itemlost_count++ if $item->{itemlost};
1950 $itemdamaged_count++ if $item->{damaged};
1951 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1952 $item_onhold_count++ if $reservestatus eq 'Waiting';
1953 $item_recalled_count++ if $recallstatus eq 'Waiting';
1954 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1958 my $key = $prefix . $item->{status};
1959 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1960 $other_items->{$key}->{$_} = $item->{$_};
1962 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1963 $other_items->{$key}->{recalled} = ($recallstatus) ? 1 : 0;
1964 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1965 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1966 $other_items->{$key}->{count}++ if $item->{$hbranch};
1967 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1968 $other_items->{$key}->{description} = $item->{description};
1969 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1974 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1975 foreach (qw(branchname itemcallnumber description)) {
1976 $available_items->{$prefix}->{$_} = $item->{$_};
1978 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1979 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1982 } # notforloan, item level and biblioitem level
1984 # if all items are hidden, do not show the record
1985 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1989 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1990 for my $key ( sort keys %$onloan_items ) {
1991 (++$onloanitemscount > $maxitems) and last;
1992 push @onloan_items_loop, $onloan_items->{$key};
1994 for my $key ( sort keys %$other_items ) {
1995 (++$otheritemscount > $maxitems) and last;
1996 push @other_items_loop, $other_items->{$key};
1998 for my $key ( sort keys %$available_items ) {
1999 (++$availableitemscount > $maxitems) and last;
2000 push @available_items_loop, $available_items->{$key}
2003 # XSLT processing of some stuff
2004 # we fetched the sysprefs already before the loop through all retrieved record!
2006 $record_processor->options({
2007 frameworkcode => $fw,
2008 interface => $search_context->{'interface'}
2011 $record_processor->process($marcrecord);
2013 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
2015 biblionumber => $oldbiblio->{biblionumber},
2016 record => $marcrecord,
2019 ? 'OPACXSLTResultsDisplay'
2020 : 'XSLTResultsDisplay'
2023 hidden_items => \@hiddenitems,
2024 xslt_variables => $xslt_variables,
2029 my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2030 $oldbiblio->{biblio_object} = $biblio_object;
2031 $oldbiblio->{coins} = eval { $biblio_object->get_coins }
2033 && C4::Context->preference('COinSinOPACResults')
2036 my $can_place_holds = 1;
2037 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2038 if (!C4::Context->preference("item-level_itypes")) {
2039 if ($itemtype && $itemtype->{notforloan}) {
2040 $can_place_holds = 0;
2043 $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object;
2045 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2046 $oldbiblio->{items_count} = $items_count;
2047 $oldbiblio->{available_items_loop} = \@available_items_loop;
2048 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2049 $oldbiblio->{other_items_loop} = \@other_items_loop;
2050 $oldbiblio->{availablecount} = $available_count;
2051 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2052 $oldbiblio->{onloancount} = $onloan_count;
2053 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2054 $oldbiblio->{othercount} = $other_count;
2055 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2056 $oldbiblio->{withdrawncount} = $withdrawn_count;
2057 $oldbiblio->{itemlostcount} = $itemlost_count;
2058 $oldbiblio->{damagedcount} = $itemdamaged_count;
2059 $oldbiblio->{intransitcount} = $item_in_transit_count;
2060 $oldbiblio->{onholdcount} = $item_onhold_count;
2061 $oldbiblio->{recalledcount} = $item_recalled_count;
2062 $oldbiblio->{orderedcount} = $ordered_count;
2063 $oldbiblio->{notforloancount} = $notforloan_count;
2065 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2066 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2067 my $subfields = substr $fieldspec, 3;
2068 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2069 my @alternateholdingsinfo = ();
2070 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2071 my $alternateholdingscount = 0;
2073 for my $field (@holdingsfields) {
2074 my %holding = ( holding => '' );
2075 my $havesubfield = 0;
2076 for my $subfield ($field->subfields()) {
2077 if ((index $subfields, $$subfield[0]) >= 0) {
2078 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2079 $holding{'holding'} .= $$subfield[1];
2083 if ($havesubfield) {
2084 push(@alternateholdingsinfo, \%holding);
2085 $alternateholdingscount++;
2089 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2090 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2093 push( @newresults, $oldbiblio );
2099 =head2 enabled_staff_search_views
2101 %hash = enabled_staff_search_views()
2103 This function returns a hash that contains three flags obtained from the system
2104 preferences, used to determine whether a particular staff search results view
2109 =item C<Output arg:>
2111 * $hash{can_view_MARC} is true only if the MARC view is enabled
2112 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2113 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2115 =item C<usage in the script:>
2119 $template->param ( C4::Search::enabled_staff_search_views );
2123 sub enabled_staff_search_views
2126 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2127 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2128 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2132 =head2 z3950_search_args
2134 $arrayref = z3950_search_args($matchpoints)
2136 This function returns an array reference that contains the search parameters to be
2137 passed to the Z39.50 search script (z3950_search.pl). The array elements
2138 are hash refs whose keys are name and value, and whose values are the
2139 name of a search parameter, the value of that search parameter and the URL encoded
2140 value of that parameter.
2142 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2144 The search parameter values are obtained from the bibliographic record whose
2145 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2147 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2148 a general purpose search argument. In this case, the returned array contains only
2149 entry: the key is 'title' and the value is derived from $matchpoints.
2151 If a search parameter value is undefined or empty, it is not included in the returned
2154 The returned array reference may be passed directly to the template parameters.
2158 =item C<Output arg:>
2160 * $array containing hash refs as described above
2162 =item C<usage in the script:>
2166 $data = Biblio::GetBiblioData($bibno);
2167 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2171 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2175 sub z3950_search_args {
2178 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2179 my $isbn = Business::ISBN->new( $isbn_string );
2181 if (defined $isbn && $isbn->is_valid)
2183 if ( ref($bibrec) ) {
2184 $bibrec->{isbn} = $isbn_string;
2185 $bibrec->{title} = undef;
2187 $bibrec = { isbn => $isbn_string };
2191 $bibrec = { title => $bibrec } if !ref $bibrec;
2194 for my $field (qw/ lccn isbn issn title author dewey subject /)
2196 push @$array, { name => $field, value => $bibrec->{$field} }
2197 if defined $bibrec->{$field};
2202 =head2 GetDistinctValues($field);
2204 C<$field> is a reference to the fields array
2208 sub GetDistinctValues {
2209 my ($fieldname,$string)=@_;
2210 # returns a reference to a hash of references to branches...
2211 if ($fieldname=~/\./){
2212 my ($table,$column)=split /\./, $fieldname;
2213 my $dbh = C4::Context->dbh;
2214 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 ");
2216 my $elements=$sth->fetchall_arrayref({});
2221 my @servers=qw<biblioserver authorityserver>;
2222 my (@zconns,@results);
2223 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2224 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2227 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2230 # The big moment: asynchronously retrieve results from all servers
2236 my ( $i, $size ) = @_;
2237 for ( my $j = 0 ; $j < $size ; $j++ ) {
2239 @hashscan{qw(value cnt)} =
2240 $results[ $i - 1 ]->display_term($j);
2241 push @elements, \%hashscan;
2249 =head2 _ZOOM_event_loop
2251 _ZOOM_event_loop(\@zconns, \@results, sub {
2252 my ( $i, $size ) = @_;
2256 Processes a ZOOM event loop and passes control to a closure for
2257 processing the results, and destroying the resultsets.
2261 sub _ZOOM_event_loop {
2262 my ($zconns, $results, $callback) = @_;
2263 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2264 my $ev = $zconns->[ $i - 1 ]->last_event();
2265 if ( $ev == ZOOM::Event::ZEND ) {
2266 next unless $results->[ $i - 1 ];
2267 my $size = $results->[ $i - 1 ]->size();
2269 $callback->($i, $size);
2274 foreach my $result (@$results) {
2279 =head2 new_record_from_zebra
2281 Given raw data from a searchengine result set, return a MARC::Record object
2283 This helper function is needed to take into account all the involved
2284 system preferences and configuration variables to properly create the
2285 MARC::Record object.
2287 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2288 data. If we are using DOM, then it has to be MARCXML.
2290 If we are using elasticsearch, it'll already be a MARC::Record and this
2291 function needs a new name.
2295 sub new_record_from_zebra {
2298 my $raw_data = shift;
2299 # Set the default indexing modes
2300 my $search_engine = C4::Context->preference("SearchEngine");
2301 if ($search_engine eq 'Elasticsearch') {
2302 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2304 my $index_mode = ( $server eq 'biblioserver' )
2305 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2306 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2308 my $marc_record = eval {
2309 if ( $index_mode eq 'dom' ) {
2310 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2312 MARC::Record->new_from_usmarc( $raw_data );
2319 return $marc_record;
2324 END { } # module clean-up code here (global destructor)
2331 Koha Development Team <http://koha-community.org/>