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 $result->{isbn} =~ s/\|/OR/;
100 $query = "isbn:$result->{isbn}";
104 my $titleindex = 'ti,ext';
105 my $authorindex = 'au,ext';
108 $result->{title} =~ s /\\//g;
109 $result->{title} =~ s /\"//g;
110 $result->{title} =~ s /\(//g;
111 $result->{title} =~ s /\)//g;
113 $query = "$titleindex:\"$result->{title}\"";
114 if ( $result->{author} ) {
115 $result->{author} =~ s /\\//g;
116 $result->{author} =~ s /\"//g;
117 $result->{author} =~ s /\(//g;
118 $result->{author} =~ s /\)//g;
120 $query .= " $op $authorindex:\"$result->{author}\"";
124 my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
125 my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50);
127 if (!defined $error) {
128 foreach my $possible_duplicate_record (@{$searchresults}) {
129 my $marcrecord = new_record_from_zebra(
131 $possible_duplicate_record
134 my $result = TransformMarcToKoha({ record => $marcrecord });
136 # FIXME :: why 2 $biblionumber ?
138 push @results, $result->{'biblionumber'};
139 push @results, $result->{'title'};
148 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
150 This function provides a simple search API on the bibliographic catalog
156 * $query can be a simple keyword or a complete CCL query
157 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
158 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
159 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
160 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
165 Returns an array consisting of three elements
166 * $error is undefined unless an error is detected
167 * $results is a reference to an array of records.
168 * $total_hits is the number of hits that would have been returned with no limit
170 If an error is returned the two other return elements are undefined. If error itself is undefined
171 the other two elements are always defined
173 =item C<usage in the script:>
177 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
179 if (defined $error) {
180 $template->param(query_error => $error);
181 warn "error: ".$error;
182 output_html_with_http_headers $input, $cookie, $template->output;
186 my $hits = @{$marcresults};
189 for my $r ( @{$marcresults} ) {
190 my $marcrecord = MARC::File::USMARC::decode($r);
191 my $biblio = TransformMarcToKoha({ record => $marcrecord });
193 #build the iarray of hashs for the template.
195 title => $biblio->{'title'},
196 subtitle => $biblio->{'subtitle'},
197 biblionumber => $biblio->{'biblionumber'},
198 author => $biblio->{'author'},
199 publishercode => $biblio->{'publishercode'},
200 publicationyear => $biblio->{'publicationyear'},
205 $template->param(result=>\@results);
210 my ( $query, $offset, $max_results, $servers, %options ) = @_;
212 return ( 'No query entered', undef, undef ) unless $query;
213 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
214 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
221 # Initialize & Search Zebra
222 for ( my $i = 0 ; $i < @servers ; $i++ ) {
224 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
225 $query =~ s/:/=/g unless $options{skip_normalize};
226 $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]);
227 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
231 $zconns[$i]->errmsg() . " ("
232 . $zconns[$i]->errcode() . ") "
233 . $zconns[$i]->addinfo() . " "
234 . $zconns[$i]->diagset();
236 return ( $error, undef, undef ) if $zconns[$i]->errcode();
240 # caught a ZOOM::Exception
244 . $@->addinfo() . " "
246 warn $error." for query: $query";
247 return ( $error, undef, undef );
256 my $first_record = defined($offset) ? $offset + 1 : 1;
257 my $hits = $tmpresults[ $i - 1 ]->size();
258 $total_hits += $hits;
259 my $last_record = $hits;
260 if ( defined $max_results && $offset + $max_results < $hits ) {
261 $last_record = $offset + $max_results;
264 for my $j ( $first_record .. $last_record ) {
266 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
269 push @{$results}, $record if defined $record;
274 foreach my $zoom_query (@zoom_queries) {
275 $zoom_query->destroy();
278 return ( undef, $results, $total_hits );
283 ( undef, $results_hashref, \@facets_loop ) = getRecords (
285 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
286 $results_per_page, $offset, $branches, $itemtypes,
287 $query_type, $scan, $opac
290 The all singing, all dancing, multi-server, asynchronous, scanning,
291 searching, record nabbing, facet-building
293 See verbose embedded documentation.
299 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
300 $results_per_page, $offset, $branches, $itemtypes,
301 $query_type, $scan, $opac
304 my @servers = @$servers_ref;
305 my @sort_by = @$sort_by_ref;
306 $offset = 0 if $offset < 0;
308 # Initialize variables for the ZOOM connection and results object
311 my $results_hashref = ();
313 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
314 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search->as_list };
316 # Initialize variables for the faceted results objects
317 my $facets_counter = {};
318 my $facets_info = {};
319 my $facets = getFacets();
321 my @facets_loop; # stores the ref to array of hashes for template facets loop
323 ### LOOP THROUGH THE SERVERS
324 for ( my $i = 0 ; $i < @servers ; $i++ ) {
325 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
327 # perform the search, create the results objects
328 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
329 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
331 Koha::Logger->get->debug($simple_query) if $scan;
333 # Check if we've got a query_type defined, if so, use it
336 if ($query_type =~ /^ccl/) {
337 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
338 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
339 } elsif ($query_type =~ /^cql/) {
340 $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i]));
341 } elsif ($query_type =~ /^pqf/) {
342 $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i]));
344 warn "Unknown query_type '$query_type'. Results undetermined.";
347 $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
349 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
353 warn "WARNING: query problem with $query_to_use " . $@;
356 # Concatenate the sort_by limits and pass them to the results object
357 # Note: sort will override rank
359 foreach my $sort (@sort_by) {
360 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
361 $sort_by .= "1=1003 <i ";
363 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
364 $sort_by .= "1=1003 >i ";
366 elsif ( $sort eq "popularity_asc" ) {
367 $sort_by .= "1=9003,4=109 <i ";
369 elsif ( $sort eq "popularity_dsc" ) {
370 $sort_by .= "1=9003,4=109 >i ";
372 elsif ( $sort eq "call_number_asc" ) {
373 $sort_by .= "1=8007 <i ";
375 elsif ( $sort eq "call_number_dsc" ) {
376 $sort_by .= "1=8007 >i ";
378 elsif ( $sort eq "pubdate_asc" ) {
379 $sort_by .= "1=31 <i ";
381 elsif ( $sort eq "pubdate_dsc" ) {
382 $sort_by .= "1=31 >i ";
384 elsif ( $sort eq "acqdate_asc" ) {
385 $sort_by .= "1=32 <i ";
387 elsif ( $sort eq "acqdate_dsc" ) {
388 $sort_by .= "1=32 >i ";
390 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
391 $sort_by .= "1=4 <i ";
393 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
394 $sort_by .= "1=4 >i ";
396 elsif ( $sort eq "biblionumber_az" || $sort eq "biblionumber_asc" ) {
397 $sort_by .= "1=12 <i ";
399 elsif ( $sort eq "biblionumber_za" || $sort eq "biblionumber_dsc" ) {
400 $sort_by .= "1=12 >i ";
403 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
406 if ( $sort_by && !$scan && $results[$i] ) {
407 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
408 warn "WARNING sort $sort_by failed";
411 } # finished looping through servers
413 # The big moment: asynchronously retrieve results from all servers
418 my ( $i, $size ) = @_;
421 # loop through the results
422 $results_hash->{'hits'} = $size;
424 if ( $offset + $results_per_page <= $size ) {
425 $times = $offset + $results_per_page;
431 for ( my $j = $offset ; $j < $times ; $j++ ) {
434 ## Check if it's an index scan
436 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
438 # here we create a minimal MARC record and hand it off to the
439 # template just like a normal result ... perhaps not ideal, but
441 my $tmprecord = MARC::Record->new();
442 $tmprecord->encoding('UTF-8');
446 # the minimal record in author/title (depending on MARC flavour)
447 if ( C4::Context->preference("marcflavour") eq
450 $tmptitle = MARC::Field->new(
455 $tmprecord->append_fields($tmptitle);
459 MARC::Field->new( '245', ' ', ' ', a => $term, );
461 MARC::Field->new( '100', ' ', ' ', a => $occ, );
462 $tmprecord->append_fields($tmptitle);
463 $tmprecord->append_fields($tmpauthor);
465 $results_hash->{'RECORDS'}[$j] =
466 $tmprecord->as_usmarc();
471 $record = $results[ $i - 1 ]->record($j)->raw();
472 # warn "RECORD $j:".$record;
473 $results_hash->{'RECORDS'}[$j] = $record;
477 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
479 # Fill the facets while we're looping, but only for the
480 # biblioserver and not for a scan
481 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
482 $facets_counter = GetFacets( $results[ $i - 1 ] );
483 $facets_info = _get_facets_info( $facets );
487 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
489 sort { $a cmp $b } keys %$facets_counter
492 my @this_facets_array;
495 $facets_counter->{$link_value}
496 ->{$b} <=> $facets_counter->{$link_value}
498 } keys %{ $facets_counter->{$link_value} }
501 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
502 my $facet_link_value = $one_facet;
503 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
505 # fix the length that will display in the label,
506 my $facet_label_value = $one_facet;
507 my $facet_max_length = C4::Context->preference(
508 'FacetLabelTruncationLength')
511 substr( $one_facet, 0, $facet_max_length )
513 if length($facet_label_value) >
516 # if it's a branch, label by the name, not the code,
517 if ( $link_value =~ /branch/ ) {
518 if ( defined $branches
519 && ref($branches) eq "HASH"
520 && defined $branches->{$one_facet}
521 && ref( $branches->{$one_facet} ) eq
525 $branches->{$one_facet}
529 $facet_label_value = "*";
533 # if it's a itemtype, label by the name, not the code,
534 if ( $link_value =~ /itype/ ) {
535 if ( defined $itemtypes
536 && ref($itemtypes) eq "HASH"
537 && defined $itemtypes->{$one_facet}
538 && ref( $itemtypes->{$one_facet} ) eq
542 $itemtypes->{$one_facet}
543 ->{translated_description};
547 # also, if it's a location code, use the name instead of the code
548 if ( $link_value =~ /location/ ) {
549 # TODO Retrieve all authorised values at once, instead of 1 query per entry
550 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
551 $facet_label_value = $av->count ? $av->next->opac_description : '';
554 # also, if it's a collection code, use the name instead of the code
555 if ( $link_value =~ /ccode/ ) {
556 # TODO Retrieve all authorised values at once, instead of 1 query per entry
557 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
558 $facet_label_value = $av->count ? $av->next->opac_description : '';
561 # but we're down with the whole label being in the link's title.
562 push @this_facets_array,
565 $facets_counter->{$link_value}
567 facet_label_value => $facet_label_value,
568 facet_title_value => $one_facet,
569 facet_link_value => $facet_link_value,
570 type_link_value => $link_value,
572 if ($facet_label_value);
577 type_link_value => $link_value,
578 type_id => $link_value . "_id",
580 . $facets_info->{$link_value}->{'label_value'} =>
582 facets => \@this_facets_array,
586 $facets_info->{$link_value}->{'label_value'} =~
589 and ( Koha::Libraries->search->count == 1 )
596 # This sorts the facets into alphabetical order
598 foreach my $f (@facets_loop) {
599 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
601 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
606 return ( undef, $results_hashref, \@facets_loop );
614 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
616 if ( $use_zebra_facets ) {
617 $facets = _get_facets_from_zebra( $rs );
619 $facets = _get_facets_from_records( $rs );
625 sub _get_facets_from_records {
629 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
630 my $facets_config = getFacets();
632 my $size = $rs->size();
633 my $jmax = $size > $facets_maxrecs
637 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
639 my $marc_record = new_record_from_zebra (
641 $rs->record( $j )->raw()
644 if ( ! defined $marc_record ) {
645 warn "ERROR DECODING RECORD - $@: " .
646 $rs->record( $j )->raw();
650 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
656 =head2 _get_facets_data_from_record
658 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
660 Internal function that extracts facets information from a MARC::Record object
661 and populates $facets_counter for using in getRecords.
663 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
668 sub _get_facets_data_from_record {
670 my ( $marc_record, $facets, $facets_counter ) = @_;
672 for my $facet (@$facets) {
676 foreach my $tag ( @{ $facet->{ tags } } ) {
678 # tag number is the first three digits
679 my $tag_num = substr( $tag, 0, 3 );
680 # subfields are the remainder
681 my $subfield_letters = substr( $tag, 3 );
683 my @fields = $marc_record->field( $tag_num );
684 foreach my $field (@fields) {
685 # If $field->indicator(1) eq 'z', it means it is a 'see from'
686 # field introduced because of IncludeSeeFromInSearches, so skip it
687 next if $field->indicator(1) eq 'z';
689 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
690 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
692 unless ( grep { $_ eq $data } @used_datas ) {
693 push @used_datas, $data;
694 $facets_counter->{ $facet->{ idx } }->{ $data }++;
701 =head2 _get_facets_from_zebra
703 my $facets = _get_facets_from_zebra( $result_set )
705 Retrieves facets for a specified result set. It loops through the facets defined
706 in C4::Koha::getFacets and returns a hash with the following structure:
716 sub _get_facets_from_zebra {
720 # save current elementSetName
721 my $elementSetName = $rs->option( 'elementSetName' );
723 my $facets_loop = getFacets();
724 my $facets_data = {};
725 # loop through defined facets and fill the facets hashref
726 foreach my $facet ( @$facets_loop ) {
728 my $idx = $facet->{ idx };
729 my $sep = $facet->{ sep };
730 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
731 if ( $facet_values ) {
732 # we've actually got a result
733 $facets_data->{ $idx } = $facet_values;
736 # set elementSetName to its previous value to avoid side effects
737 $rs->option( elementSetName => $elementSetName );
742 =head2 _get_facet_from_result_set
745 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
747 Internal function that extracts facet information for a specific index ($facet_idx) and
748 returns a hash containing facet values and count:
751 $facet_value => $count ,
755 Warning: this function has the side effect of changing the elementSetName for the result
756 set. It is a helper function for the main loop, which takes care of backing it up for
761 sub _get_facet_from_result_set {
763 my $facet_idx = shift;
767 my $internal_sep = '<*>';
768 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
770 return if ( ! defined $facet_idx || ! defined $rs );
771 # zebra's facet element, untokenized index
772 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
773 # configure zebra results for retrieving the desired facet
774 $rs->option( elementSetName => $facet_element );
775 # get the facet record from result set
776 my $facet = $rs->record( 0 )->raw;
777 # if the facet has no restuls...
778 return if !defined $facet;
779 # TODO: benchmark DOM vs. SAX performance
780 my $facet_dom = XML::LibXML->load_xml(
783 my @terms = $facet_dom->getElementsByTagName('term');
787 foreach my $term ( @terms ) {
788 my $facet_value = $term->textContent;
789 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
790 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
791 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
797 =head2 _get_facets_info
799 my $facets_info = C4::Search::_get_facets_info( $facets )
801 Internal function that extracts facets information and properly builds
802 the data structure needed to render facet labels.
806 sub _get_facets_info {
810 my $facets_info = {};
812 for my $facet ( @$facets ) {
813 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
820 sub _detect_truncation {
821 my ( $operand, $index ) = @_;
822 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
825 my @wordlist = split( /\s/, $operand );
826 foreach my $word (@wordlist) {
827 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
828 push @rightlefttruncated, $word;
830 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
831 push @lefttruncated, $word;
833 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
834 push @righttruncated, $word;
836 elsif ( index( $word, "*" ) < 0 ) {
837 push @nontruncated, $word;
840 push @regexpr, $word;
844 \@nontruncated, \@righttruncated, \@lefttruncated,
845 \@rightlefttruncated, \@regexpr
850 sub _build_stemmed_operand {
851 my ($operand,$lang) = @_;
852 require Lingua::Stem::Snowball ;
853 my $stemmed_operand=q{};
855 # Stemmer needs language
856 return $operand unless $lang;
858 # If operand contains a digit, it is almost certainly an identifier, and should
859 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
860 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
861 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
862 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
863 return $operand if $operand =~ /\d/;
865 # FIXME: the locale should be set based on the user's language and/or search choice
867 # Make sure we only use the first two letters from the language code
868 $lang = lc(substr($lang, 0, 2));
869 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
870 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
871 if ($lang eq 'nb' || $lang eq 'nn') {
874 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
875 encoding => "UTF-8" );
877 my @words = split( / /, $operand );
878 my @stems = $stemmer->stem(\@words);
879 for my $stem (@stems) {
880 $stemmed_operand .= "$stem";
881 $stemmed_operand .= "?"
882 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
883 $stemmed_operand .= " ";
886 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
887 return $stemmed_operand;
891 sub _build_weighted_query {
893 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
894 # pretty well but could work much better if we had a smarter query parser
895 my ( $operand, $stemmed_operand, $index ) = @_;
896 my $stemming = C4::Context->preference("QueryStemming") || 0;
897 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
898 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
899 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
901 my $weighted_query = "(rk=("; # Specifies that we're applying rank
903 # Keyword, or, no index specified
904 if ( ( $index eq 'kw' ) || ( !$index ) ) {
906 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
907 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
908 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
909 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
910 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
911 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
912 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
913 if $fuzzy_enabled; # add fuzzy, word list
914 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
915 if ( $stemming and $stemmed_operand )
916 ; # add stemming, right truncation
917 $weighted_query .= " or wrdl,r9=\"$operand\"";
919 # embedded sorting: 0 a-z; 1 z-a
920 # $weighted_query .= ") or (sort1,aut=1";
923 # Barcode searches should skip this process
924 elsif ( $index eq 'bc' ) {
925 $weighted_query .= "bc=\"$operand\"";
928 # Authority-number searches should skip this process
929 elsif ( $index eq 'an' ) {
930 $weighted_query .= "an=\"$operand\"";
933 # If the index is numeric, don't autoquote it.
934 elsif ( $index =~ /,st-numeric$/ ) {
935 $weighted_query .= " $index=$operand";
938 # If the index already has more than one qualifier, wrap the operand
939 # in quotes and pass it back (assumption is that the user knows what they
940 # are doing and won't appreciate us mucking up their query
941 elsif ( $index =~ ',' ) {
942 $weighted_query .= " $index=\"$operand\"";
945 #TODO: build better cases based on specific search indexes
947 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
948 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
949 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
950 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
951 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
952 if $fuzzy_enabled; # add fuzzy, word list
953 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
954 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
957 $weighted_query .= "))"; # close rank specification
958 return $weighted_query;
963 Return an array with available indexes.
987 'Author-personal-bibliography',
997 'Chronological-subdivision',
1007 'Conference-name-heading',
1008 'Conference-name-see',
1009 'Conference-name-seealso',
1013 'Control-number-identifier',
1017 'Corporate-name-heading',
1018 'Corporate-name-see',
1019 'Corporate-name-seealso',
1020 'Country-publication',
1023 'date-entered-on-file',
1024 'Date-of-acquisition',
1025 'Date-of-publication',
1026 'Date-time-last-modified',
1027 'Dewey-classification',
1028 'Dissertation-information',
1037 'Geographic-subdivision',
1040 'Heading-use-main-or-added-entry',
1041 'Heading-use-series-added-entry ',
1042 'Heading-use-subject-added-entry',
1046 'Identifier-standard',
1047 'Illustration-code',
1049 'Index-term-uncontrolled',
1050 'Interest-age-level',
1051 'Interest-grade-level',
1061 'language-original',
1071 'Local-classification',
1074 'Match-heading-see-from',
1079 'Multipart-resource-level',
1084 'Name-geographic-heading',
1085 'Name-geographic-see',
1086 'Name-geographic-seealso',
1092 'Other-control-number',
1095 'Personal-name-heading',
1096 'Personal-name-see',
1097 'Personal-name-seealso',
1099 'Place-publication',
1106 'Reading-grade-level',
1107 'Record-control-number',
1118 'Subject-heading-thesaurus',
1119 'Subject-name-personal',
1120 'Subject-subdivision',
1129 'Term-genre-form-heading',
1130 'Term-genre-form-see',
1131 'Term-genre-form-seealso',
1137 'Title-uniform-heading',
1138 'Title-uniform-see',
1139 'Title-uniform-seealso',
1149 'classification-source',
1151 'coded-location-qualifier',
1162 'Local-classification',
1165 'materials-specified',
1170 'Number-local-acquisition',
1175 'replacementpricedate',
1193 $simple_query, $query_cgi,
1194 $query_desc, $limit,
1195 $limit_cgi, $limit_desc,
1196 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1198 Build queries and limits in CCL, CGI, Human,
1199 handle truncation, stemming, field weighting, fuzziness, etc.
1201 See verbose embedded documentation.
1207 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1212 my @operators = $operators ? @$operators : ();
1213 my @indexes = $indexes ? @$indexes : ();
1214 my @operands = $operands ? @$operands : ();
1215 my @limits = $limits ? @$limits : ();
1216 my @sort_by = $sort_by ? @$sort_by : ();
1218 my $stemming = C4::Context->preference("QueryStemming") || 0;
1219 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1220 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1221 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1223 my $query = $operands[0] // "";
1224 my $simple_query = $operands[0];
1226 # initialize the variables we're passing back
1235 my $cclindexes = getIndexes();
1236 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1237 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1239 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1241 $query = "ccl=$query" if $cclq;
1245 my %group_OR_limits;
1246 my $availability_limit;
1247 foreach my $this_limit (@limits) {
1248 next unless $this_limit;
1249 if ( $this_limit =~ /available/ ) {
1251 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1253 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1254 $availability_limit .=
1255 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1256 $limit_cgi .= "&limit=available";
1260 # group_OR_limits, prefixed by mc-
1261 # OR every member of the group
1262 elsif ( $this_limit =~ /mc/ ) {
1263 my ($k,$v) = split(/:/, $this_limit,2);
1264 if ( $k !~ /mc-i(tem)?type/ ) {
1265 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1266 $this_limit =~ tr/"//d;
1267 $this_limit = $k.':"'.$v.'"';
1270 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1271 $limit_desc .= " or " if $group_OR_limits{$k};
1272 $group_OR_limits{$k} .= "$this_limit";
1273 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1274 $limit_desc .= " $this_limit";
1276 elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) {
1277 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1278 $limit .= " and " if $limit || $query;
1279 my $branchfield = C4::Context->preference('SearchLimitLibrary');
1281 if( $this_limit =~ '^multibranchlimit:' ){
1282 my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ );
1283 my $search_group = Koha::Library::Groups->find( $group_id );
1284 @branchcodes = map { $_->branchcode } $search_group->all_libraries;
1285 @branchcodes = sort { $a cmp $b } @branchcodes;
1287 @branchcodes = ( $this_limit =~ /^branch:(.*)$/ );
1291 if ( $branchfield eq "homebranch" ) {
1292 $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes;
1294 elsif ( $branchfield eq "holdingbranch" ) {
1295 $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes;
1298 $this_limit = sprintf "(%s or %s)",
1299 join( " or ", map { 'homebranch: ' . $_ } @branchcodes ),
1300 join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes );
1303 $limit .= "$this_limit";
1304 $limit_desc .= " $this_limit";
1305 } elsif ( $this_limit =~ '^search_filter:' ) {
1306 # Here we will get the query as a string, append to the limits, and pass through buildQuery
1307 # again to clean the terms and handle nested filters
1308 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1309 my ($filter_id) = ( $this_limit =~ /^search_filter:(.*)$/ );
1310 my $search_filter = Koha::SearchFilters->find( $filter_id );
1311 next unless $search_filter;
1312 my ($expanded_lim, $query_lim) = $search_filter->expand_filter;
1313 push @$expanded_lim, $query_lim;
1314 my ( $error, undef, undef, undef, undef, $fixed_limit, undef, undef, undef ) = buildQuery ( undef, undef, undef, $expanded_lim, undef, undef, $lang);
1315 $limit .= " and " if $limit || $query;
1316 $limit .= "$fixed_limit";
1317 $limit_desc .= " $limit";
1320 # Regular old limits
1322 $limit .= " and " if $limit || $query;
1323 $limit .= "$this_limit";
1324 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1325 $limit_desc .= " $this_limit";
1328 foreach my $k (keys (%group_OR_limits)) {
1329 $limit .= " and " if ( $query || $limit );
1330 $limit .= "($group_OR_limits{$k})";
1332 if ($availability_limit) {
1333 $limit .= " and " if ( $query || $limit );
1334 $limit .= "($availability_limit)";
1337 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1339 if ( $query =~ /^ccl=/ ) {
1341 # This is needed otherwise ccl= and &limit won't work together, and
1342 # this happens when selecting a subject on the opac-detail page
1343 my $original_q = $q; # without available part
1344 $q .= $limit if $limit;
1345 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1347 if ( $query =~ /^cql=/ ) {
1348 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1350 if ( $query =~ /^pqf=/ ) {
1352 $query_cgi = "q=pqf=".uri_escape_utf8($');
1353 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1356 # pass nested queries directly
1357 # FIXME: need better handling of some of these variables in this case
1358 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1359 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1360 # if ( $query =~ /(\(|\))/ ) {
1362 # undef, $query, $simple_query, $query_cgi,
1363 # $query, $limit, $limit_cgi, $limit_desc,
1368 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1369 # query operands and indexes and add stemming, truncation, field weighting, etc.
1370 # Once we do so, we'll end up with a value in $query, just like if we had an
1371 # incoming $query from the user
1374 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1375 my $previous_operand
1376 ; # a flag used to keep track if there was a previous query
1377 # if there was, we can apply the current operator
1379 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1381 # COMBINE OPERANDS, INDEXES AND OPERATORS
1382 if ( ($operands[$i] // '') ne '' ) {
1383 $operands[$i]=~s/^\s+//;
1385 # A flag to determine whether or not to add the index to the query
1388 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1389 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1393 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1395 my $operand = $operands[$i];
1396 my $index = $indexes[$i] || 'kw';
1398 # Add index-specific attributes
1400 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1401 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1402 #which is processed higher up in this sub. Other than that, year searches are typically
1403 #handled as limits which are not processed her either.
1405 # Search ranges: Date of Publication, st-numeric
1406 if ( $index =~ /(yr|st-numeric)/ ) {
1407 #weight_fields/relevance search causes errors with date ranges
1408 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1409 #In the case of YYYY-YYYY, it will return no results
1410 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1413 # Date of Acquisition
1414 elsif ( $index =~ /acqdate/ ) {
1415 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1416 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1417 #top of the results just because they have lots of item records matching that date.
1418 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1420 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1422 # ISBN,ISSN,Standard Number, don't need special treatment
1423 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1425 $stemming, $auto_truncation,
1426 $weight_fields, $fuzzy_enabled
1429 if ( $index eq 'nb' ) {
1430 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1431 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1432 $operands[$i] = $operand = '(' . join( ' OR ', map { 'nb=' . $_ } @isbns ) . ')';
1433 $indexes[$i] = $index = 'kw';
1436 if ( $index eq 'ns' ) {
1437 if ( C4::Context->preference("SearchWithISSNVariations") ) {
1438 my @issns = C4::Koha::GetVariationsOfISSN( $operand );
1439 $operands[$i] = $operand = '(' . join( ' OR ', map { 'ns=' . $_ } @issns ) . ')';
1440 $indexes[$i] = $index = 'kw';
1445 # Set default structure attribute (word list)
1446 my $struct_attr = q{};
1447 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1448 $struct_attr = ",wrdl";
1451 # Some helpful index variants
1452 my $index_plus = $index . $struct_attr . ':';
1453 my $index_plus_comma = $index . $struct_attr . ',';
1455 if ($auto_truncation){
1456 unless ( $index =~ /,(st-|phr|ext)/ ) {
1457 #FIXME only valid with LTR scripts
1458 $operand=join(" ",map{
1459 (index($_,"*")>0?"$_":"$_*")
1460 }split (/\s+/,$operand));
1465 my $truncated_operand = q{};
1466 my( $nontruncated, $righttruncated, $lefttruncated,
1467 $rightlefttruncated, $regexpr
1468 ) = _detect_truncation( $operand, $index );
1470 Koha::Logger->get->debug(
1471 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1475 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1476 scalar(@$rightlefttruncated) > 0 )
1479 # Don't field weight or add the index to the query, we do it here
1481 undef $weight_fields;
1482 my $previous_truncation_operand;
1483 if (scalar @$nontruncated) {
1484 $truncated_operand .= "$index_plus @$nontruncated ";
1485 $previous_truncation_operand = 1;
1487 if (scalar @$righttruncated) {
1488 $truncated_operand .= "and " if $previous_truncation_operand;
1489 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1490 $previous_truncation_operand = 1;
1492 if (scalar @$lefttruncated) {
1493 $truncated_operand .= "and " if $previous_truncation_operand;
1494 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1495 $previous_truncation_operand = 1;
1497 if (scalar @$rightlefttruncated) {
1498 $truncated_operand .= "and " if $previous_truncation_operand;
1499 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1500 $previous_truncation_operand = 1;
1503 $operand = $truncated_operand if $truncated_operand;
1504 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1507 my $stemmed_operand = q{};
1508 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1511 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1513 # Handle Field Weighting
1514 my $weighted_operand = q{};
1515 if ($weight_fields) {
1516 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1517 $operand = $weighted_operand;
1521 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1523 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1525 #N.B. Truncation is mutually exclusive with Weighted Queries,
1526 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1527 #the need for this relevance wrapper.
1528 $operand = "(rk=($operand))" unless $weight_fields;
1530 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1532 query_cgi => $query_cgi,
1533 query_desc => $query_desc,
1534 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1535 parsed_operand => $operand,
1536 original_operand => $operands[$i] // '',
1538 index_plus => $index_plus,
1539 indexes_set => $indexes_set,
1540 previous_operand => $previous_operand,
1546 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1548 # Normalize the query and limit strings
1549 # This is flawed , means we can't search anything with : in it
1550 # if user wants to do ccl or cql, start the query with that
1551 # $query =~ s/:/=/g;
1552 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1553 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1554 $query =~ s/(?<=(wrdl)):/=/g;
1555 $query =~ s/(?<=(trn|phr)):/=/g;
1556 $query =~ s/(?<=(st-numeric)):/=/g;
1557 $query =~ s/(?<=(st-year)):/=/g;
1558 $query =~ s/(?<=(st-date-normalized)):/=/g;
1560 # Removing warnings for later substitutions
1562 $query_desc //= q{};
1565 $limit_desc //= q{};
1567 $simple_query //= q{};
1569 for ( $query, $query_desc, $limit, $limit_desc ) {
1570 s/ +/ /g; # remove extra spaces
1571 s/^ //g; # remove any beginning spaces
1572 s/ $//g; # remove any ending spaces
1573 s/==/=/g; # remove double == from query
1575 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1577 for ($query_cgi,$simple_query) {
1580 # append the limit to the query
1581 $query .= " " . $limit;
1583 Koha::Logger->get->debug(
1584 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1585 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1588 undef, $query, $simple_query, $query_cgi,
1589 $query_desc, $limit, $limit_cgi, $limit_desc,
1594 =head2 _build_initial_query
1596 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1598 Build a section of the initial query containing indexes, operators, and operands.
1602 sub _build_initial_query {
1606 if ($params->{previous_operand}){
1607 #If there is a previous operand, add a supplied operator or the default 'and'
1608 $operator = ($params->{operator}) ? ($params->{operator}) : 'AND';
1611 #NOTE: indexes_set is typically set when doing truncation or field weighting
1612 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1614 #e.g. "kw,wrdl:test"
1615 #e.g. " and kw,wrdl:test"
1616 $params->{query} .= " " . $operator . " " . $operand;
1618 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1619 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1620 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if ( $params->{original_operand} ne '' );
1622 #e.g. " and kw,wrdl: test"
1623 $params->{query_desc} .= " " . $operator . " " . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1625 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1627 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1630 =head2 searchResults
1632 my @search_results = searchResults($search_context, $searchdesc, $hits,
1633 $results_per_page, $offset, $scan,
1636 Format results in a form suitable for passing to the template
1640 # IMO this subroutine is pretty messy still -- it's responsible for
1641 # building the HTML output for the template
1643 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1644 my $dbh = C4::Context->dbh;
1649 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1650 my ($is_opac, $hidelostitems);
1651 if ($search_context->{'interface'} eq 'opac') {
1652 $hidelostitems = C4::Context->preference('hidelostitems');
1656 my $record_processor = Koha::RecordProcessor->new({
1657 filters => 'ViewPolicy'
1660 #Build branchnames hash
1661 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' })->as_list;
1663 # FIXME - We build an authorised values hash here, using the default framework
1664 # though it is possible to have different authvals for different fws.
1666 my $shelflocations =
1667 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1669 # get notforloan authorised value list (see $shelflocations FIXME)
1670 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1671 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1674 my $itemtypes = Koha::ItemTypes->search_with_localization;
1675 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1677 #search item field code
1678 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1680 ## find column names of items related to MARC
1681 my %subfieldstosearch;
1682 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1683 for my $column ( @columns ) {
1684 my ( $tagfield, $tagsubfield ) =
1685 &GetMarcFromKohaField( "items." . $column );
1686 if ( defined $tagsubfield ) {
1687 $subfieldstosearch{$column} = $tagsubfield;
1691 # handle which records to actually retrieve
1692 my $times; # Times is which record to process up to
1693 if ( $hits && $offset + $results_per_page <= $hits ) {
1694 $times = $offset + $results_per_page;
1697 $times = $hits; # If less hits than results_per_page+offset we go to the end
1700 my $marcflavour = C4::Context->preference("marcflavour");
1701 # We get the biblionumber position in MARC
1702 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1704 # set stuff for XSLT processing here once, not later again for every record we retrieved
1706 my $userenv = C4::Context->userenv;
1708 = ( defined $userenv and $userenv->{number} )
1709 ? Koha::Patrons->find( $userenv->{number} )
1711 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1713 # loop through all of the records we've retrieved
1714 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1718 # For Scan searches we built USMARC data
1719 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1721 # Normal search, render from Zebra's output
1722 $marcrecord = new_record_from_zebra(
1727 if ( ! defined $marcrecord ) {
1728 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1736 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1737 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1739 SetUTF8Flag($marcrecord);
1740 my $oldbiblio = TransformMarcToKoha({ record => $marcrecord, limit_table => 'no_items' });
1741 $oldbiblio->{result_number} = $i + 1;
1743 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1744 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1745 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1746 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1747 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1749 # edition information, if any
1750 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1752 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1753 # add imageurl to itemtype if there is one
1754 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1755 # Build summary if there is one (the summary is defined in the itemtypes table)
1756 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1758 # Pull out the items fields
1759 my @fields = $marcrecord->field($itemtag);
1760 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1761 my $marcflavor = C4::Context->preference("marcflavour");
1763 # adding linked items that belong to host records
1764 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1765 my $analyticsfield = '773';
1766 if ($marcflavor eq 'MARC21') {
1767 $analyticsfield = '773';
1768 } elsif ($marcflavor eq 'UNIMARC') {
1769 $analyticsfield = '461';
1771 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1772 my $hostbiblionumber = $hostfield->subfield("0");
1773 my $linkeditemnumber = $hostfield->subfield("9");
1774 if( $hostbiblionumber ) {
1775 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1776 if ($linkeditemmarc) {
1777 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1778 if ($linkeditemfield) {
1779 push( @fields, $linkeditemfield );
1786 # Setting item statuses for display
1787 my @available_items_loop;
1788 my @onloan_items_loop;
1789 my @other_items_loop;
1791 my $available_items;
1795 my $ordered_count = 0;
1796 my $available_count = 0;
1797 my $onloan_count = 0;
1798 my $longoverdue_count = 0;
1799 my $other_count = 0;
1800 my $withdrawn_count = 0;
1801 my $itemlost_count = 0;
1802 my $hideatopac_count = 0;
1803 my $itembinding_count = 0;
1804 my $itemdamaged_count = 0;
1805 my $item_in_transit_count = 0;
1806 my $item_onhold_count = 0;
1807 my $notforloan_count = 0;
1808 my $item_recalled_count = 0;
1809 my $items_count = scalar(@fields);
1810 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1811 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1812 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1814 # loop through every item
1815 foreach my $field (@fields) {
1818 # populate the items hash
1819 foreach my $code ( keys %subfieldstosearch ) {
1820 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1823 unless ( $item->{itemnumber} ) {
1824 warn "MARC item without itemnumber retrieved for biblio ($oldbiblio->{biblionumber})";
1828 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1832 # hidden based on OpacHiddenItems syspref or because lost
1833 my $hi = Koha::Items->search( { itemnumber => $item->{itemnumber} } )
1834 ->filter_by_visible_in_opac({ patron => $search_context->{patron} });
1835 unless ( $hi->count ) {
1836 push @hiddenitems, $item->{itemnumber};
1837 $hideatopac_count++;
1842 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1843 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1845 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1846 if ($item->{$hbranch}) {
1847 $item->{'branchname'} = $branches{$item->{$hbranch}};
1849 elsif ($item->{$otherbranch}) { # Last resort
1850 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1854 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1855 . ( $item->{location} ? $item->{location} : q{} )
1856 . ( $item->{itype} ? $item->{itype} : q{} )
1857 . ( $item->{ccode} ? $item->{ccode} : q{} )
1858 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1859 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1860 if ( $item->{onloan}
1862 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1865 my $key = $prefix . $item->{onloan} . $item->{barcode};
1866 $onloan_items->{$key}->{due_date} = $item->{onloan};
1867 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1868 $onloan_items->{$key}->{branchname} = $item->{branchname};
1869 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1870 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1871 $onloan_items->{$key}->{description} = $item->{description};
1872 $onloan_items->{$key}->{imageurl} =
1873 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1874 $onloan_items->{$key}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE');
1876 # if something's checked out and lost, mark it as 'long overdue'
1877 if ( $item->{itemlost} ) {
1878 $onloan_items->{$key}->{longoverdue}++;
1879 $longoverdue_count++;
1883 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1886 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1887 $item->{notforloan} = 1 if !$item->{notforloan} &&
1888 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1891 if ( $item->{notforloan} < 0 ) {
1893 } elsif ( $item->{notforloan} > 0 ) {
1894 $notforloan_count++;
1897 # is item in transit?
1898 my $transfertwhen = '';
1899 my ($transfertfrom, $transfertto);
1901 # is item on the reserve shelf?
1902 my $reservestatus = '';
1904 # is item a waiting recall?
1905 my $recallstatus = '';
1907 unless ($item->{withdrawn}
1908 || $item->{itemlost}
1910 || $item->{notforloan}
1911 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1912 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1914 # A couple heuristics to limit how many times
1915 # we query the database for item transfer information, sacrificing
1916 # accuracy in some cases for speed;
1918 # 1. don't query if item has one of the other statuses
1919 # 2. don't check transit status if the bib has
1920 # more than 20 items
1922 # FIXME: to avoid having the query the database like this, and to make
1923 # the in transit status count as unavailable for search limiting,
1924 # should map transit status to record indexed in Zebra.
1926 my $item_object = Koha::Items->find($item->{itemnumber});
1927 my $transfer = defined($item_object) ? $item_object->get_transfer : undef;
1928 ( $transfertwhen, $transfertfrom, $transfertto ) =
1931 $transfer->datesent, $transfer->frombranch,
1935 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1936 if ( C4::Context->preference('UseRecalls') ) {
1937 if ( Koha::Recalls->search({ item_id => $item->{itemnumber}, status => 'waiting' })->count ) {
1938 $recallstatus = 'Waiting';
1943 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1944 if ( $item->{withdrawn}
1945 || $item->{itemlost}
1947 || $item->{notforloan}
1948 || $reservestatus eq 'Waiting'
1949 || $recallstatus eq 'Waiting'
1950 || ($transfertwhen && $transfertwhen ne ''))
1952 $withdrawn_count++ if $item->{withdrawn};
1953 $itemlost_count++ if $item->{itemlost};
1954 $itemdamaged_count++ if $item->{damaged};
1955 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1956 $item_onhold_count++ if $reservestatus eq 'Waiting';
1957 $item_recalled_count++ if $recallstatus eq 'Waiting';
1958 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1962 my $key = $prefix . $item->{status};
1963 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1964 $other_items->{$key}->{$_} = $item->{$_};
1966 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1967 $other_items->{$key}->{recalled} = ($recallstatus) ? 1 : 0;
1968 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1969 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1970 $other_items->{$key}->{count}++ if $item->{$hbranch};
1971 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1972 $other_items->{$key}->{description} = $item->{description};
1973 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1974 $other_items->{$key}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE');
1979 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1980 foreach (qw(branchname itemcallnumber description)) {
1981 $available_items->{$prefix}->{$_} = $item->{$_};
1983 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1984 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1985 $available_items->{$prefix}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE');
1988 } # notforloan, item level and biblioitem level
1990 # if all items are hidden, do not show the record
1991 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1995 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1996 for my $key ( sort keys %$onloan_items ) {
1997 (++$onloanitemscount > $maxitems) and last;
1998 push @onloan_items_loop, $onloan_items->{$key};
2000 for my $key ( sort keys %$other_items ) {
2001 (++$otheritemscount > $maxitems) and last;
2002 push @other_items_loop, $other_items->{$key};
2004 for my $key ( sort keys %$available_items ) {
2005 (++$availableitemscount > $maxitems) and last;
2006 push @available_items_loop, $available_items->{$key}
2009 # XSLT processing of some stuff
2010 # we fetched the sysprefs already before the loop through all retrieved record!
2012 $record_processor->options({
2013 frameworkcode => $fw,
2014 interface => $search_context->{'interface'}
2017 $record_processor->process($marcrecord);
2019 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
2021 biblionumber => $oldbiblio->{biblionumber},
2022 record => $marcrecord,
2025 ? 'OPACXSLTResultsDisplay'
2026 : 'XSLTResultsDisplay'
2029 hidden_items => \@hiddenitems,
2030 xslt_variables => $xslt_variables,
2035 my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2036 $oldbiblio->{biblio_object} = $biblio_object;
2037 $oldbiblio->{coins} = eval { $biblio_object->get_coins }
2039 && C4::Context->preference('COinSinOPACResults')
2042 my $can_place_holds = 1;
2043 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2044 if (!C4::Context->preference("item-level_itypes")) {
2045 if ($itemtype && $itemtype->{notforloan}) {
2046 $can_place_holds = 0;
2049 $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object;
2051 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2052 $oldbiblio->{items_count} = $items_count;
2053 $oldbiblio->{available_items_loop} = \@available_items_loop;
2054 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2055 $oldbiblio->{other_items_loop} = \@other_items_loop;
2056 $oldbiblio->{availablecount} = $available_count;
2057 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2058 $oldbiblio->{onloancount} = $onloan_count;
2059 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2060 $oldbiblio->{othercount} = $other_count;
2061 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2062 $oldbiblio->{withdrawncount} = $withdrawn_count;
2063 $oldbiblio->{itemlostcount} = $itemlost_count;
2064 $oldbiblio->{damagedcount} = $itemdamaged_count;
2065 $oldbiblio->{intransitcount} = $item_in_transit_count;
2066 $oldbiblio->{onholdcount} = $item_onhold_count;
2067 $oldbiblio->{recalledcount} = $item_recalled_count;
2068 $oldbiblio->{orderedcount} = $ordered_count;
2069 $oldbiblio->{notforloancount} = $notforloan_count;
2071 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2072 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2073 my $subfields = substr $fieldspec, 3;
2074 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2075 my @alternateholdingsinfo = ();
2076 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2078 for my $field (@holdingsfields) {
2079 my %holding = ( holding => '' );
2080 my $havesubfield = 0;
2081 for my $subfield ($field->subfields()) {
2082 if ((index $subfields, $$subfield[0]) >= 0) {
2083 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2084 $holding{'holding'} .= $$subfield[1];
2088 if ($havesubfield) {
2089 push(@alternateholdingsinfo, \%holding);
2093 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2096 push( @newresults, $oldbiblio );
2102 =head2 enabled_staff_search_views
2104 %hash = enabled_staff_search_views()
2106 This function returns a hash that contains three flags obtained from the system
2107 preferences, used to determine whether a particular staff search results view
2112 =item C<Output arg:>
2114 * $hash{can_view_MARC} is true only if the MARC view is enabled
2115 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2116 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2118 =item C<usage in the script:>
2122 $template->param ( C4::Search::enabled_staff_search_views );
2126 sub enabled_staff_search_views
2129 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2130 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2131 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2135 =head2 z3950_search_args
2137 $arrayref = z3950_search_args($matchpoints)
2139 This function returns an array reference that contains the search parameters to be
2140 passed to the Z39.50 search script (z3950_search.pl). The array elements
2141 are hash refs whose keys are name and value, and whose values are the
2142 name of a search parameter, the value of that search parameter and the URL encoded
2143 value of that parameter.
2145 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2147 The search parameter values are obtained from the bibliographic record whose
2148 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2150 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2151 a general purpose search argument. In this case, the returned array contains only
2152 entry: the key is 'title' and the value is derived from $matchpoints.
2154 If a search parameter value is undefined or empty, it is not included in the returned
2157 The returned array reference may be passed directly to the template parameters.
2161 =item C<Output arg:>
2163 * $array containing hash refs as described above
2165 =item C<usage in the script:>
2169 $data = Biblio::GetBiblioData($bibno);
2170 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2174 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2178 sub z3950_search_args {
2181 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2182 my $isbn = Business::ISBN->new( $isbn_string );
2184 if (defined $isbn && $isbn->is_valid)
2186 if ( ref($bibrec) ) {
2187 $bibrec->{isbn} = $isbn_string;
2188 $bibrec->{title} = undef;
2190 $bibrec = { isbn => $isbn_string };
2194 $bibrec = { title => $bibrec } if !ref $bibrec;
2197 for my $field (qw/ lccn isbn issn title author dewey subject /)
2199 push @$array, { name => $field, value => $bibrec->{$field} }
2200 if defined $bibrec->{$field};
2205 =head2 GetDistinctValues($field);
2207 C<$field> is a reference to the fields array
2211 sub GetDistinctValues {
2212 my ($fieldname,$string)=@_;
2213 # returns a reference to a hash of references to branches...
2214 if ($fieldname=~/\./){
2215 my ($table,$column)=split /\./, $fieldname;
2216 my $dbh = C4::Context->dbh;
2217 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 ");
2219 my $elements=$sth->fetchall_arrayref({});
2224 my @servers=qw<biblioserver authorityserver>;
2225 my (@zconns,@results);
2226 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2227 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2230 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2233 # The big moment: asynchronously retrieve results from all servers
2239 my ( $i, $size ) = @_;
2240 for ( my $j = 0 ; $j < $size ; $j++ ) {
2242 @hashscan{qw(value cnt)} =
2243 $results[ $i - 1 ]->display_term($j);
2244 push @elements, \%hashscan;
2252 =head2 _ZOOM_event_loop
2254 _ZOOM_event_loop(\@zconns, \@results, sub {
2255 my ( $i, $size ) = @_;
2259 Processes a ZOOM event loop and passes control to a closure for
2260 processing the results, and destroying the resultsets.
2264 sub _ZOOM_event_loop {
2265 my ($zconns, $results, $callback) = @_;
2266 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2267 my $ev = $zconns->[ $i - 1 ]->last_event();
2268 if ( $ev == ZOOM::Event::ZEND ) {
2269 next unless $results->[ $i - 1 ];
2270 my $size = $results->[ $i - 1 ]->size();
2272 $callback->($i, $size);
2277 foreach my $result (@$results) {
2282 =head2 new_record_from_zebra
2284 Given raw data from a searchengine result set, return a MARC::Record object
2286 This helper function is needed to take into account all the involved
2287 system preferences and configuration variables to properly create the
2288 MARC::Record object.
2290 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2291 data. If we are using DOM, then it has to be MARCXML.
2293 If we are using elasticsearch, it'll already be a MARC::Record and this
2294 function needs a new name.
2298 sub new_record_from_zebra {
2301 my $raw_data = shift;
2302 # Set the default indexing modes
2303 my $search_engine = C4::Context->preference("SearchEngine");
2304 if ($search_engine eq 'Elasticsearch') {
2305 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2307 my $index_mode = ( $server eq 'biblioserver' )
2308 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2309 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2311 my $marc_record = eval {
2312 if ( $index_mode eq 'dom' ) {
2313 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2315 MARC::Record->new_from_usmarc( $raw_data );
2322 return $marc_record;
2327 END { } # module clean-up code here (global destructor)
2334 Koha Development Team <http://koha-community.org/>