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 );
26 use C4::XSLT qw( XSLTParse4Display );
27 use C4::Reserves qw( GetReserveStatus );
28 use C4::Charset qw( SetUTF8Flag );
29 use Koha::AuthorisedValues;
35 use Koha::RecordProcessor;
41 our (@ISA, @EXPORT_OK);
52 enabled_staff_search_views
61 C4::Search - Functions for searching the Koha catalog.
65 See opac/opac-search.pl or catalogue/search.pl for example of usage
69 This module provides searching functions for Koha's bibliographic databases
75 # make all your functions, whether exported or not;
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $record, '' );
92 # search duplicate on ISBN, easy and fast..
94 if ( $result->{isbn} ) {
95 $result->{isbn} =~ s/\(.*$//;
96 $result->{isbn} =~ s/\s+$//;
97 $query = "isbn:$result->{isbn}";
101 my $titleindex = 'ti,ext';
102 my $authorindex = 'au,ext';
105 $result->{title} =~ s /\\//g;
106 $result->{title} =~ s /\"//g;
107 $result->{title} =~ s /\(//g;
108 $result->{title} =~ s /\)//g;
110 $query = "$titleindex:\"$result->{title}\"";
111 if ( $result->{author} ) {
112 $result->{author} =~ s /\\//g;
113 $result->{author} =~ s /\"//g;
114 $result->{author} =~ s /\(//g;
115 $result->{author} =~ s /\)//g;
117 $query .= " $op $authorindex:\"$result->{author}\"";
121 my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
122 my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50);
124 if (!defined $error) {
125 foreach my $possible_duplicate_record (@{$searchresults}) {
126 my $marcrecord = new_record_from_zebra(
128 $possible_duplicate_record
131 my $result = TransformMarcToKoha( $marcrecord, '' );
133 # FIXME :: why 2 $biblionumber ?
135 push @results, $result->{'biblionumber'};
136 push @results, $result->{'title'};
145 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
147 This function provides a simple search API on the bibliographic catalog
153 * $query can be a simple keyword or a complete CCL query
154 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
155 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
156 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
157 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
162 Returns an array consisting of three elements
163 * $error is undefined unless an error is detected
164 * $results is a reference to an array of records.
165 * $total_hits is the number of hits that would have been returned with no limit
167 If an error is returned the two other return elements are undefined. If error itself is undefined
168 the other two elements are always defined
170 =item C<usage in the script:>
174 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
176 if (defined $error) {
177 $template->param(query_error => $error);
178 warn "error: ".$error;
179 output_html_with_http_headers $input, $cookie, $template->output;
183 my $hits = @{$marcresults};
186 for my $r ( @{$marcresults} ) {
187 my $marcrecord = MARC::File::USMARC::decode($r);
188 my $biblio = TransformMarcToKoha($marcrecord,q{});
190 #build the iarray of hashs for the template.
192 title => $biblio->{'title'},
193 subtitle => $biblio->{'subtitle'},
194 biblionumber => $biblio->{'biblionumber'},
195 author => $biblio->{'author'},
196 publishercode => $biblio->{'publishercode'},
197 publicationyear => $biblio->{'publicationyear'},
202 $template->param(result=>\@results);
207 my ( $query, $offset, $max_results, $servers, %options ) = @_;
209 return ( 'No query entered', undef, undef ) unless $query;
210 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
211 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
218 # Initialize & Search Zebra
219 for ( my $i = 0 ; $i < @servers ; $i++ ) {
221 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
222 $query =~ s/:/=/g unless $options{skip_normalize};
223 $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]);
224 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
228 $zconns[$i]->errmsg() . " ("
229 . $zconns[$i]->errcode() . ") "
230 . $zconns[$i]->addinfo() . " "
231 . $zconns[$i]->diagset();
233 return ( $error, undef, undef ) if $zconns[$i]->errcode();
237 # caught a ZOOM::Exception
241 . $@->addinfo() . " "
243 warn $error." for query: $query";
244 return ( $error, undef, undef );
253 my $first_record = defined($offset) ? $offset + 1 : 1;
254 my $hits = $tmpresults[ $i - 1 ]->size();
255 $total_hits += $hits;
256 my $last_record = $hits;
257 if ( defined $max_results && $offset + $max_results < $hits ) {
258 $last_record = $offset + $max_results;
261 for my $j ( $first_record .. $last_record ) {
263 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
266 push @{$results}, $record if defined $record;
271 foreach my $zoom_query (@zoom_queries) {
272 $zoom_query->destroy();
275 return ( undef, $results, $total_hits );
280 ( undef, $results_hashref, \@facets_loop ) = getRecords (
282 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
283 $results_per_page, $offset, $branches, $itemtypes,
284 $query_type, $scan, $opac
287 The all singing, all dancing, multi-server, asynchronous, scanning,
288 searching, record nabbing, facet-building
290 See verbose embedded documentation.
296 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
297 $results_per_page, $offset, $branches, $itemtypes,
298 $query_type, $scan, $opac
301 my @servers = @$servers_ref;
302 my @sort_by = @$sort_by_ref;
303 $offset = 0 if $offset < 0;
305 # Initialize variables for the ZOOM connection and results object
308 my $results_hashref = ();
310 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
311 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search->as_list };
313 # Initialize variables for the faceted results objects
314 my $facets_counter = {};
315 my $facets_info = {};
316 my $facets = getFacets();
318 my @facets_loop; # stores the ref to array of hashes for template facets loop
320 ### LOOP THROUGH THE SERVERS
321 for ( my $i = 0 ; $i < @servers ; $i++ ) {
322 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
324 # perform the search, create the results objects
325 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
326 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
328 Koha::Logger->get->debug($simple_query) if $scan;
330 # Check if we've got a query_type defined, if so, use it
333 if ($query_type =~ /^ccl/) {
334 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
335 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
336 } elsif ($query_type =~ /^cql/) {
337 $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i]));
338 } elsif ($query_type =~ /^pqf/) {
339 $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i]));
341 warn "Unknown query_type '$query_type'. Results undetermined.";
344 $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
346 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
350 warn "WARNING: query problem with $query_to_use " . $@;
353 # Concatenate the sort_by limits and pass them to the results object
354 # Note: sort will override rank
356 foreach my $sort (@sort_by) {
357 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
358 $sort_by .= "1=1003 <i ";
360 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
361 $sort_by .= "1=1003 >i ";
363 elsif ( $sort eq "popularity_asc" ) {
364 $sort_by .= "1=9003 <i ";
366 elsif ( $sort eq "popularity_dsc" ) {
367 $sort_by .= "1=9003 >i ";
369 elsif ( $sort eq "call_number_asc" ) {
370 $sort_by .= "1=8007 <i ";
372 elsif ( $sort eq "call_number_dsc" ) {
373 $sort_by .= "1=8007 >i ";
375 elsif ( $sort eq "pubdate_asc" ) {
376 $sort_by .= "1=31 <i ";
378 elsif ( $sort eq "pubdate_dsc" ) {
379 $sort_by .= "1=31 >i ";
381 elsif ( $sort eq "acqdate_asc" ) {
382 $sort_by .= "1=32 <i ";
384 elsif ( $sort eq "acqdate_dsc" ) {
385 $sort_by .= "1=32 >i ";
387 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
388 $sort_by .= "1=4 <i ";
390 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
391 $sort_by .= "1=4 >i ";
394 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
397 if ( $sort_by && !$scan && $results[$i] ) {
398 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
399 warn "WARNING sort $sort_by failed";
402 } # finished looping through servers
404 # The big moment: asynchronously retrieve results from all servers
409 my ( $i, $size ) = @_;
412 # loop through the results
413 $results_hash->{'hits'} = $size;
415 if ( $offset + $results_per_page <= $size ) {
416 $times = $offset + $results_per_page;
422 for ( my $j = $offset ; $j < $times ; $j++ ) {
425 ## Check if it's an index scan
427 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
429 # here we create a minimal MARC record and hand it off to the
430 # template just like a normal result ... perhaps not ideal, but
432 my $tmprecord = MARC::Record->new();
433 $tmprecord->encoding('UTF-8');
437 # the minimal record in author/title (depending on MARC flavour)
438 if ( C4::Context->preference("marcflavour") eq
441 $tmptitle = MARC::Field->new(
446 $tmprecord->append_fields($tmptitle);
450 MARC::Field->new( '245', ' ', ' ', a => $term, );
452 MARC::Field->new( '100', ' ', ' ', a => $occ, );
453 $tmprecord->append_fields($tmptitle);
454 $tmprecord->append_fields($tmpauthor);
456 $results_hash->{'RECORDS'}[$j] =
457 $tmprecord->as_usmarc();
462 $record = $results[ $i - 1 ]->record($j)->raw();
463 # warn "RECORD $j:".$record;
464 $results_hash->{'RECORDS'}[$j] = $record;
468 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
470 # Fill the facets while we're looping, but only for the
471 # biblioserver and not for a scan
472 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
473 $facets_counter = GetFacets( $results[ $i - 1 ] );
474 $facets_info = _get_facets_info( $facets );
478 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
480 sort { $a cmp $b } keys %$facets_counter
483 my @this_facets_array;
486 $facets_counter->{$link_value}
487 ->{$b} <=> $facets_counter->{$link_value}
489 } keys %{ $facets_counter->{$link_value} }
492 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
493 my $facet_link_value = $one_facet;
494 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
496 # fix the length that will display in the label,
497 my $facet_label_value = $one_facet;
498 my $facet_max_length = C4::Context->preference(
499 'FacetLabelTruncationLength')
502 substr( $one_facet, 0, $facet_max_length )
504 if length($facet_label_value) >
507 # if it's a branch, label by the name, not the code,
508 if ( $link_value =~ /branch/ ) {
509 if ( defined $branches
510 && ref($branches) eq "HASH"
511 && defined $branches->{$one_facet}
512 && ref( $branches->{$one_facet} ) eq
516 $branches->{$one_facet}
520 $facet_label_value = "*";
524 # if it's a itemtype, label by the name, not the code,
525 if ( $link_value =~ /itype/ ) {
526 if ( defined $itemtypes
527 && ref($itemtypes) eq "HASH"
528 && defined $itemtypes->{$one_facet}
529 && ref( $itemtypes->{$one_facet} ) eq
533 $itemtypes->{$one_facet}
534 ->{translated_description};
538 # also, if it's a location code, use the name instead of the code
539 if ( $link_value =~ /location/ ) {
540 # TODO Retrieve all authorised values at once, instead of 1 query per entry
541 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
542 $facet_label_value = $av->count ? $av->next->opac_description : '';
545 # also, if it's a collection code, use the name instead of the code
546 if ( $link_value =~ /ccode/ ) {
547 # TODO Retrieve all authorised values at once, instead of 1 query per entry
548 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
549 $facet_label_value = $av->count ? $av->next->opac_description : '';
552 # but we're down with the whole label being in the link's title.
553 push @this_facets_array,
556 $facets_counter->{$link_value}
558 facet_label_value => $facet_label_value,
559 facet_title_value => $one_facet,
560 facet_link_value => $facet_link_value,
561 type_link_value => $link_value,
563 if ($facet_label_value);
568 type_link_value => $link_value,
569 type_id => $link_value . "_id",
571 . $facets_info->{$link_value}->{'label_value'} =>
573 facets => \@this_facets_array,
577 $facets_info->{$link_value}->{'label_value'} =~
580 and ( Koha::Libraries->search->count == 1 )
587 # This sorts the facets into alphabetical order
589 foreach my $f (@facets_loop) {
590 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
592 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
597 return ( undef, $results_hashref, \@facets_loop );
605 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
607 if ( $use_zebra_facets ) {
608 $facets = _get_facets_from_zebra( $rs );
610 $facets = _get_facets_from_records( $rs );
616 sub _get_facets_from_records {
620 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
621 my $facets_config = getFacets();
623 my $size = $rs->size();
624 my $jmax = $size > $facets_maxrecs
628 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
630 my $marc_record = new_record_from_zebra (
632 $rs->record( $j )->raw()
635 if ( ! defined $marc_record ) {
636 warn "ERROR DECODING RECORD - $@: " .
637 $rs->record( $j )->raw();
641 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
647 =head2 _get_facets_data_from_record
649 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
651 Internal function that extracts facets information from a MARC::Record object
652 and populates $facets_counter for using in getRecords.
654 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
659 sub _get_facets_data_from_record {
661 my ( $marc_record, $facets, $facets_counter ) = @_;
663 for my $facet (@$facets) {
667 foreach my $tag ( @{ $facet->{ tags } } ) {
669 # tag number is the first three digits
670 my $tag_num = substr( $tag, 0, 3 );
671 # subfields are the remainder
672 my $subfield_letters = substr( $tag, 3 );
674 my @fields = $marc_record->field( $tag_num );
675 foreach my $field (@fields) {
676 # If $field->indicator(1) eq 'z', it means it is a 'see from'
677 # field introduced because of IncludeSeeFromInSearches, so skip it
678 next if $field->indicator(1) eq 'z';
680 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
681 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
683 unless ( grep { $_ eq $data } @used_datas ) {
684 push @used_datas, $data;
685 $facets_counter->{ $facet->{ idx } }->{ $data }++;
692 =head2 _get_facets_from_zebra
694 my $facets = _get_facets_from_zebra( $result_set )
696 Retrieves facets for a specified result set. It loops through the facets defined
697 in C4::Koha::getFacets and returns a hash with the following structure:
707 sub _get_facets_from_zebra {
711 # save current elementSetName
712 my $elementSetName = $rs->option( 'elementSetName' );
714 my $facets_loop = getFacets();
715 my $facets_data = {};
716 # loop through defined facets and fill the facets hashref
717 foreach my $facet ( @$facets_loop ) {
719 my $idx = $facet->{ idx };
720 my $sep = $facet->{ sep };
721 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
722 if ( $facet_values ) {
723 # we've actually got a result
724 $facets_data->{ $idx } = $facet_values;
727 # set elementSetName to its previous value to avoid side effects
728 $rs->option( elementSetName => $elementSetName );
733 =head2 _get_facet_from_result_set
736 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
738 Internal function that extracts facet information for a specific index ($facet_idx) and
739 returns a hash containing facet values and count:
742 $facet_value => $count ,
746 Warning: this function has the side effect of changing the elementSetName for the result
747 set. It is a helper function for the main loop, which takes care of backing it up for
752 sub _get_facet_from_result_set {
754 my $facet_idx = shift;
758 my $internal_sep = '<*>';
759 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
761 return if ( ! defined $facet_idx || ! defined $rs );
762 # zebra's facet element, untokenized index
763 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
764 # configure zebra results for retrieving the desired facet
765 $rs->option( elementSetName => $facet_element );
766 # get the facet record from result set
767 my $facet = $rs->record( 0 )->raw;
768 # if the facet has no restuls...
769 return if !defined $facet;
770 # TODO: benchmark DOM vs. SAX performance
771 my $facet_dom = XML::LibXML->load_xml(
774 my @terms = $facet_dom->getElementsByTagName('term');
778 foreach my $term ( @terms ) {
779 my $facet_value = $term->textContent;
780 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
781 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
782 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
788 =head2 _get_facets_info
790 my $facets_info = C4::Search::_get_facets_info( $facets )
792 Internal function that extracts facets information and properly builds
793 the data structure needed to render facet labels.
797 sub _get_facets_info {
801 my $facets_info = {};
803 for my $facet ( @$facets ) {
804 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
811 sub _detect_truncation {
812 my ( $operand, $index ) = @_;
813 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
816 my @wordlist = split( /\s/, $operand );
817 foreach my $word (@wordlist) {
818 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
819 push @rightlefttruncated, $word;
821 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
822 push @lefttruncated, $word;
824 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
825 push @righttruncated, $word;
827 elsif ( index( $word, "*" ) < 0 ) {
828 push @nontruncated, $word;
831 push @regexpr, $word;
835 \@nontruncated, \@righttruncated, \@lefttruncated,
836 \@rightlefttruncated, \@regexpr
841 sub _build_stemmed_operand {
842 my ($operand,$lang) = @_;
843 require Lingua::Stem::Snowball ;
844 my $stemmed_operand=q{};
846 # Stemmer needs language
847 return $operand unless $lang;
849 # If operand contains a digit, it is almost certainly an identifier, and should
850 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
851 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
852 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
853 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
854 return $operand if $operand =~ /\d/;
856 # FIXME: the locale should be set based on the user's language and/or search choice
858 # Make sure we only use the first two letters from the language code
859 $lang = lc(substr($lang, 0, 2));
860 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
861 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
862 if ($lang eq 'nb' || $lang eq 'nn') {
865 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
866 encoding => "UTF-8" );
868 my @words = split( / /, $operand );
869 my @stems = $stemmer->stem(\@words);
870 for my $stem (@stems) {
871 $stemmed_operand .= "$stem";
872 $stemmed_operand .= "?"
873 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
874 $stemmed_operand .= " ";
877 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
878 return $stemmed_operand;
882 sub _build_weighted_query {
884 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
885 # pretty well but could work much better if we had a smarter query parser
886 my ( $operand, $stemmed_operand, $index ) = @_;
887 my $stemming = C4::Context->preference("QueryStemming") || 0;
888 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
889 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
890 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
892 my $weighted_query = "(rk=("; # Specifies that we're applying rank
894 # Keyword, or, no index specified
895 if ( ( $index eq 'kw' ) || ( !$index ) ) {
897 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
898 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
899 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
900 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
901 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
902 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
903 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
904 if $fuzzy_enabled; # add fuzzy, word list
905 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
906 if ( $stemming and $stemmed_operand )
907 ; # add stemming, right truncation
908 $weighted_query .= " or wrdl,r9=\"$operand\"";
910 # embedded sorting: 0 a-z; 1 z-a
911 # $weighted_query .= ") or (sort1,aut=1";
914 # Barcode searches should skip this process
915 elsif ( $index eq 'bc' ) {
916 $weighted_query .= "bc=\"$operand\"";
919 # Authority-number searches should skip this process
920 elsif ( $index eq 'an' ) {
921 $weighted_query .= "an=\"$operand\"";
924 # If the index is numeric, don't autoquote it.
925 elsif ( $index =~ /,st-numeric$/ ) {
926 $weighted_query .= " $index=$operand";
929 # If the index already has more than one qualifier, wrap the operand
930 # in quotes and pass it back (assumption is that the user knows what they
931 # are doing and won't appreciate us mucking up their query
932 elsif ( $index =~ ',' ) {
933 $weighted_query .= " $index=\"$operand\"";
936 #TODO: build better cases based on specific search indexes
938 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
939 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
940 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
941 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
942 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
943 if $fuzzy_enabled; # add fuzzy, word list
944 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
945 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
948 $weighted_query .= "))"; # close rank specification
949 return $weighted_query;
954 Return an array with available indexes.
978 'Author-personal-bibliography',
988 'Chronological-subdivision',
998 'Conference-name-heading',
999 'Conference-name-see',
1000 'Conference-name-seealso',
1003 'Control-number-identifier',
1007 'Corporate-name-heading',
1008 'Corporate-name-see',
1009 'Corporate-name-seealso',
1010 'Country-publication',
1013 'date-entered-on-file',
1014 'Date-of-acquisition',
1015 'Date-of-publication',
1016 'Date-time-last-modified',
1017 'Dewey-classification',
1018 'Dissertation-information',
1027 'Geographic-subdivision',
1030 'Heading-use-main-or-added-entry',
1031 'Heading-use-series-added-entry ',
1032 'Heading-use-subject-added-entry',
1036 'Identifier-standard',
1037 'Illustration-code',
1039 'Index-term-uncontrolled',
1040 'Interest-age-level',
1041 'Interest-grade-level',
1051 'language-original',
1061 'Local-classification',
1064 'Match-heading-see-from',
1069 'Multipart-resource-level',
1074 'Name-geographic-heading',
1075 'Name-geographic-see',
1076 'Name-geographic-seealso',
1082 'Other-control-number',
1085 'Personal-name-heading',
1086 'Personal-name-see',
1087 'Personal-name-seealso',
1089 'Place-publication',
1096 'Reading-grade-level',
1097 'Record-control-number',
1108 'Subject-heading-thesaurus',
1109 'Subject-name-personal',
1110 'Subject-subdivision',
1119 'Term-genre-form-heading',
1120 'Term-genre-form-see',
1121 'Term-genre-form-seealso',
1127 'Title-uniform-heading',
1128 'Title-uniform-see',
1129 'Title-uniform-seealso',
1139 'classification-source',
1141 'coded-location-qualifier',
1152 'Local-classification',
1155 'materials-specified',
1160 'Number-local-acquisition',
1165 'replacementpricedate',
1183 $simple_query, $query_cgi,
1184 $query_desc, $limit,
1185 $limit_cgi, $limit_desc,
1186 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1188 Build queries and limits in CCL, CGI, Human,
1189 handle truncation, stemming, field weighting, fuzziness, etc.
1191 See verbose embedded documentation.
1197 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1202 my @operators = $operators ? @$operators : ();
1203 my @indexes = $indexes ? @$indexes : ();
1204 my @operands = $operands ? @$operands : ();
1205 my @limits = $limits ? @$limits : ();
1206 my @sort_by = $sort_by ? @$sort_by : ();
1208 my $stemming = C4::Context->preference("QueryStemming") || 0;
1209 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1210 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1211 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1213 my $query = $operands[0] // "";
1214 my $simple_query = $operands[0];
1216 # initialize the variables we're passing back
1225 my $cclindexes = getIndexes();
1226 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1227 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1229 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1231 $query = "ccl=$query" if $cclq;
1234 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1236 if ( $query =~ /^ccl=/ ) {
1238 # This is needed otherwise ccl= and &limit won't work together, and
1239 # this happens when selecting a subject on the opac-detail page
1240 @limits = grep {!/^$/} @limits;
1241 my $original_q = $q; # without available part
1242 unless ( grep { $_ eq 'available' } @limits ) {
1243 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1247 if ( grep { $_ eq 'available' } @limits ) {
1248 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1249 @limits = grep {!/^available$/} @limits;
1251 $q .= ' and '.join(' and ', @limits) if @limits;
1253 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1255 if ( $query =~ /^cql=/ ) {
1256 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1258 if ( $query =~ /^pqf=/ ) {
1260 $query_cgi = "q=pqf=".uri_escape_utf8($');
1261 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1264 # pass nested queries directly
1265 # FIXME: need better handling of some of these variables in this case
1266 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1267 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1268 # if ( $query =~ /(\(|\))/ ) {
1270 # undef, $query, $simple_query, $query_cgi,
1271 # $query, $limit, $limit_cgi, $limit_desc,
1276 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1277 # query operands and indexes and add stemming, truncation, field weighting, etc.
1278 # Once we do so, we'll end up with a value in $query, just like if we had an
1279 # incoming $query from the user
1282 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1283 my $previous_operand
1284 ; # a flag used to keep track if there was a previous query
1285 # if there was, we can apply the current operator
1287 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1289 # COMBINE OPERANDS, INDEXES AND OPERATORS
1290 if ( ($operands[$i] // '') ne '' ) {
1291 $operands[$i]=~s/^\s+//;
1293 # A flag to determine whether or not to add the index to the query
1296 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1297 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1301 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1303 my $operand = $operands[$i];
1304 my $index = $indexes[$i] || 'kw';
1306 # Add index-specific attributes
1308 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1309 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1310 #which is processed higher up in this sub. Other than that, year searches are typically
1311 #handled as limits which are not processed her either.
1313 # Search ranges: Date of Publication, st-numeric
1314 if ( $index =~ /(yr|st-numeric)/ ) {
1315 #weight_fields/relevance search causes errors with date ranges
1316 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1317 #In the case of YYYY-YYYY, it will return no results
1318 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1321 # Date of Acquisition
1322 elsif ( $index =~ /acqdate/ ) {
1323 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1324 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1325 #top of the results just because they have lots of item records matching that date.
1326 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1328 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1330 # ISBN,ISSN,Standard Number, don't need special treatment
1331 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1333 $stemming, $auto_truncation,
1334 $weight_fields, $fuzzy_enabled
1337 if ( $index eq 'nb' ) {
1338 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1339 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1340 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1341 $indexes[$i] = $index = 'kw';
1346 # Set default structure attribute (word list)
1347 my $struct_attr = q{};
1348 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1349 $struct_attr = ",wrdl";
1352 # Some helpful index variants
1353 my $index_plus = $index . $struct_attr . ':';
1354 my $index_plus_comma = $index . $struct_attr . ',';
1356 if ($auto_truncation){
1357 unless ( $index =~ /,(st-|phr|ext)/ ) {
1358 #FIXME only valid with LTR scripts
1359 $operand=join(" ",map{
1360 (index($_,"*")>0?"$_":"$_*")
1361 }split (/\s+/,$operand));
1366 my $truncated_operand = q{};
1367 my( $nontruncated, $righttruncated, $lefttruncated,
1368 $rightlefttruncated, $regexpr
1369 ) = _detect_truncation( $operand, $index );
1371 Koha::Logger->get->debug(
1372 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1376 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1377 scalar(@$rightlefttruncated) > 0 )
1380 # Don't field weight or add the index to the query, we do it here
1382 undef $weight_fields;
1383 my $previous_truncation_operand;
1384 if (scalar @$nontruncated) {
1385 $truncated_operand .= "$index_plus @$nontruncated ";
1386 $previous_truncation_operand = 1;
1388 if (scalar @$righttruncated) {
1389 $truncated_operand .= "and " if $previous_truncation_operand;
1390 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1391 $previous_truncation_operand = 1;
1393 if (scalar @$lefttruncated) {
1394 $truncated_operand .= "and " if $previous_truncation_operand;
1395 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1396 $previous_truncation_operand = 1;
1398 if (scalar @$rightlefttruncated) {
1399 $truncated_operand .= "and " if $previous_truncation_operand;
1400 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1401 $previous_truncation_operand = 1;
1404 $operand = $truncated_operand if $truncated_operand;
1405 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1408 my $stemmed_operand = q{};
1409 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1412 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1414 # Handle Field Weighting
1415 my $weighted_operand = q{};
1416 if ($weight_fields) {
1417 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1418 $operand = $weighted_operand;
1422 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1424 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1426 #N.B. Truncation is mutually exclusive with Weighted Queries,
1427 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1428 #the need for this relevance wrapper.
1429 $operand = "(rk=($operand))" unless $weight_fields;
1431 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1433 query_cgi => $query_cgi,
1434 query_desc => $query_desc,
1435 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1436 parsed_operand => $operand,
1437 original_operand => $operands[$i] // '',
1439 index_plus => $index_plus,
1440 indexes_set => $indexes_set,
1441 previous_operand => $previous_operand,
1447 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1450 my %group_OR_limits;
1451 my $availability_limit;
1452 foreach my $this_limit (@limits) {
1453 next unless $this_limit;
1454 if ( $this_limit =~ /available/ ) {
1456 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1458 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1459 $availability_limit .=
1460 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1461 $limit_cgi .= "&limit=available";
1465 # group_OR_limits, prefixed by mc-
1466 # OR every member of the group
1467 elsif ( $this_limit =~ /mc/ ) {
1468 my ($k,$v) = split(/:/, $this_limit,2);
1469 if ( $k !~ /mc-i(tem)?type/ ) {
1470 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1471 $this_limit =~ tr/"//d;
1472 $this_limit = $k.':"'.$v.'"';
1475 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1476 $limit_desc .= " or " if $group_OR_limits{$k};
1477 $group_OR_limits{$k} .= "$this_limit";
1478 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1479 $limit_desc .= " $this_limit";
1481 elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) {
1482 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1483 $limit .= " and " if $limit || $query;
1484 my $branchfield = C4::Context->preference('SearchLimitLibrary');
1486 if( $this_limit =~ '^multibranchlimit:' ){
1487 my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ );
1488 my $search_group = Koha::Library::Groups->find( $group_id );
1489 @branchcodes = map { $_->branchcode } $search_group->all_libraries;
1490 @branchcodes = sort { $a cmp $b } @branchcodes;
1492 @branchcodes = ( $this_limit =~ /^branch:(.*)$/ );
1496 if ( $branchfield eq "homebranch" ) {
1497 $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes;
1499 elsif ( $branchfield eq "holdingbranch" ) {
1500 $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes;
1503 $this_limit = sprintf "(%s or %s)",
1504 join( " or ", map { 'homebranch: ' . $_ } @branchcodes ),
1505 join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes );
1508 $limit .= "$this_limit";
1509 $limit_desc .= " $this_limit";
1512 # Regular old limits
1514 $limit .= " and " if $limit || $query;
1515 $limit .= "$this_limit";
1516 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1517 $limit_desc .= " $this_limit";
1520 foreach my $k (keys (%group_OR_limits)) {
1521 $limit .= " and " if ( $query || $limit );
1522 $limit .= "($group_OR_limits{$k})";
1524 if ($availability_limit) {
1525 $limit .= " and " if ( $query || $limit );
1526 $limit .= "($availability_limit)";
1529 # Normalize the query and limit strings
1530 # This is flawed , means we can't search anything with : in it
1531 # if user wants to do ccl or cql, start the query with that
1532 # $query =~ s/:/=/g;
1533 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1534 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1535 $query =~ s/(?<=(wrdl)):/=/g;
1536 $query =~ s/(?<=(trn|phr)):/=/g;
1537 $query =~ s/(?<=(st-numeric)):/=/g;
1538 $query =~ s/(?<=(st-year)):/=/g;
1539 $query =~ s/(?<=(st-date-normalized)):/=/g;
1541 # Removing warnings for later substitutions
1543 $query_desc //= q{};
1546 $limit_desc //= q{};
1548 $simple_query //= q{};
1550 for ( $query, $query_desc, $limit, $limit_desc ) {
1551 s/ +/ /g; # remove extra spaces
1552 s/^ //g; # remove any beginning spaces
1553 s/ $//g; # remove any ending spaces
1554 s/==/=/g; # remove double == from query
1556 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1558 for ($query_cgi,$simple_query) {
1561 # append the limit to the query
1562 $query .= " " . $limit;
1564 Koha::Logger->get->debug(
1565 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1566 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1569 undef, $query, $simple_query, $query_cgi,
1570 $query_desc, $limit, $limit_cgi, $limit_desc,
1575 =head2 _build_initial_query
1577 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1579 Build a section of the initial query containing indexes, operators, and operands.
1583 sub _build_initial_query {
1587 if ($params->{previous_operand}){
1588 #If there is a previous operand, add a supplied operator or the default 'and'
1589 $operator = ($params->{operator}) ? ($params->{operator}) : 'AND';
1592 #NOTE: indexes_set is typically set when doing truncation or field weighting
1593 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1595 #e.g. "kw,wrdl:test"
1596 #e.g. " and kw,wrdl:test"
1597 $params->{query} .= " " . $operator . " " . $operand;
1599 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1600 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1601 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if ( $params->{original_operand} ne '' );
1603 #e.g. " and kw,wrdl: test"
1604 $params->{query_desc} .= " " . $operator . " " . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1606 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1608 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1611 =head2 searchResults
1613 my @search_results = searchResults($search_context, $searchdesc, $hits,
1614 $results_per_page, $offset, $scan,
1617 Format results in a form suitable for passing to the template
1621 # IMO this subroutine is pretty messy still -- it's responsible for
1622 # building the HTML output for the template
1624 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1625 my $dbh = C4::Context->dbh;
1630 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1631 my ($is_opac, $hidelostitems);
1632 if ($search_context->{'interface'} eq 'opac') {
1633 $hidelostitems = C4::Context->preference('hidelostitems');
1637 my $record_processor = Koha::RecordProcessor->new({
1638 filters => 'ViewPolicy'
1641 #Build branchnames hash
1642 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' })->as_list;
1644 # FIXME - We build an authorised values hash here, using the default framework
1645 # though it is possible to have different authvals for different fws.
1647 my $shelflocations =
1648 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1650 # get notforloan authorised value list (see $shelflocations FIXME)
1651 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1652 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1655 my $itemtypes = Koha::ItemTypes->search_with_localization;
1656 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1658 #search item field code
1659 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1661 ## find column names of items related to MARC
1662 my %subfieldstosearch;
1663 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1664 for my $column ( @columns ) {
1665 my ( $tagfield, $tagsubfield ) =
1666 &GetMarcFromKohaField( "items." . $column );
1667 if ( defined $tagsubfield ) {
1668 $subfieldstosearch{$column} = $tagsubfield;
1672 # handle which records to actually retrieve
1673 my $times; # Times is which record to process up to
1674 if ( $hits && $offset + $results_per_page <= $hits ) {
1675 $times = $offset + $results_per_page;
1678 $times = $hits; # If less hits than results_per_page+offset we go to the end
1681 my $marcflavour = C4::Context->preference("marcflavour");
1682 # We get the biblionumber position in MARC
1683 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1685 # set stuff for XSLT processing here once, not later again for every record we retrieved
1687 my $userenv = C4::Context->userenv;
1689 = ( defined $userenv and $userenv->{number} )
1690 ? Koha::Patrons->find( $userenv->{number} )
1692 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1694 # loop through all of the records we've retrieved
1695 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1699 # For Scan searches we built USMARC data
1700 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1702 # Normal search, render from Zebra's output
1703 $marcrecord = new_record_from_zebra(
1708 if ( ! defined $marcrecord ) {
1709 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1717 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1718 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1720 SetUTF8Flag($marcrecord);
1721 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw, 'no_items' );
1722 $oldbiblio->{result_number} = $i + 1;
1724 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1725 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1726 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1727 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1728 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1730 # edition information, if any
1731 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1733 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1734 # add imageurl to itemtype if there is one
1735 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1736 # Build summary if there is one (the summary is defined in the itemtypes table)
1737 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1739 # Pull out the items fields
1740 my @fields = $marcrecord->field($itemtag);
1741 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1742 my $marcflavor = C4::Context->preference("marcflavour");
1744 # adding linked items that belong to host records
1745 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1746 my $analyticsfield = '773';
1747 if ($marcflavor eq 'MARC21') {
1748 $analyticsfield = '773';
1749 } elsif ($marcflavor eq 'UNIMARC') {
1750 $analyticsfield = '461';
1752 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1753 my $hostbiblionumber = $hostfield->subfield("0");
1754 my $linkeditemnumber = $hostfield->subfield("9");
1755 if( $hostbiblionumber ) {
1756 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1757 if ($linkeditemmarc) {
1758 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1759 if ($linkeditemfield) {
1760 push( @fields, $linkeditemfield );
1767 # Setting item statuses for display
1768 my @available_items_loop;
1769 my @onloan_items_loop;
1770 my @other_items_loop;
1772 my $available_items;
1776 my $ordered_count = 0;
1777 my $available_count = 0;
1778 my $onloan_count = 0;
1779 my $longoverdue_count = 0;
1780 my $other_count = 0;
1781 my $withdrawn_count = 0;
1782 my $itemlost_count = 0;
1783 my $hideatopac_count = 0;
1784 my $itembinding_count = 0;
1785 my $itemdamaged_count = 0;
1786 my $item_in_transit_count = 0;
1787 my $item_onhold_count = 0;
1788 my $notforloan_count = 0;
1789 my $item_recalled_count = 0;
1790 my $items_count = scalar(@fields);
1791 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1792 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1793 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1795 # loop through every item
1796 foreach my $field (@fields) {
1799 # populate the items hash
1800 foreach my $code ( keys %subfieldstosearch ) {
1801 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1803 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1807 # hidden because lost
1808 if ($hidelostitems && $item->{itemlost}) {
1809 push @hiddenitems, $item->{itemnumber};
1810 $hideatopac_count++;
1813 # hidden based on OpacHiddenItems syspref
1814 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
1816 push @hiddenitems, @hi;
1817 $hideatopac_count++;
1822 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1823 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1825 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1826 if ($item->{$hbranch}) {
1827 $item->{'branchname'} = $branches{$item->{$hbranch}};
1829 elsif ($item->{$otherbranch}) { # Last resort
1830 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1834 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1835 . ( $item->{location} ? $item->{location} : q{} )
1836 . ( $item->{itype} ? $item->{itype} : q{} )
1837 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1838 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1839 if ( $item->{onloan}
1841 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1844 my $key = $prefix . $item->{onloan} . $item->{barcode};
1845 $onloan_items->{$key}->{due_date} = $item->{onloan};
1846 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1847 $onloan_items->{$key}->{branchname} = $item->{branchname};
1848 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1849 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1850 $onloan_items->{$key}->{description} = $item->{description};
1851 $onloan_items->{$key}->{imageurl} =
1852 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1854 # if something's checked out and lost, mark it as 'long overdue'
1855 if ( $item->{itemlost} ) {
1856 $onloan_items->{$key}->{longoverdue}++;
1857 $longoverdue_count++;
1861 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1864 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1865 $item->{notforloan} = 1 if !$item->{notforloan} &&
1866 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1869 if ( $item->{notforloan} < 0 ) {
1871 } elsif ( $item->{notforloan} > 0 ) {
1872 $notforloan_count++;
1875 # is item in transit?
1876 my $transfertwhen = '';
1877 my ($transfertfrom, $transfertto);
1879 # is item on the reserve shelf?
1880 my $reservestatus = '';
1882 # is item a waiting recall?
1883 my $recallstatus = '';
1885 unless ($item->{withdrawn}
1886 || $item->{itemlost}
1888 || $item->{notforloan}
1889 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1890 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1892 # A couple heuristics to limit how many times
1893 # we query the database for item transfer information, sacrificing
1894 # accuracy in some cases for speed;
1896 # 1. don't query if item has one of the other statuses
1897 # 2. don't check transit status if the bib has
1898 # more than 20 items
1900 # FIXME: to avoid having the query the database like this, and to make
1901 # the in transit status count as unavailable for search limiting,
1902 # should map transit status to record indexed in Zebra.
1904 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1905 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1906 if ( C4::Context->preference('UseRecalls') ) {
1907 if ( Koha::Recalls->search({ item_id => $item->{itemnumber}, status => 'waiting' })->count ) {
1908 $recallstatus = 'Waiting';
1913 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1914 if ( $item->{withdrawn}
1915 || $item->{itemlost}
1917 || $item->{notforloan}
1918 || $reservestatus eq 'Waiting'
1919 || $recallstatus eq 'Waiting'
1920 || ($transfertwhen && $transfertwhen ne ''))
1922 $withdrawn_count++ if $item->{withdrawn};
1923 $itemlost_count++ if $item->{itemlost};
1924 $itemdamaged_count++ if $item->{damaged};
1925 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1926 $item_onhold_count++ if $reservestatus eq 'Waiting';
1927 $item_recalled_count++ if $recallstatus eq 'Waiting';
1928 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1932 my $key = $prefix . $item->{status};
1933 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1934 $other_items->{$key}->{$_} = $item->{$_};
1936 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1937 $other_items->{$key}->{recalled} = ($recallstatus) ? 1 : 0;
1938 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1939 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1940 $other_items->{$key}->{count}++ if $item->{$hbranch};
1941 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1942 $other_items->{$key}->{description} = $item->{description};
1943 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1948 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1949 foreach (qw(branchname itemcallnumber description)) {
1950 $available_items->{$prefix}->{$_} = $item->{$_};
1952 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1953 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1956 } # notforloan, item level and biblioitem level
1958 # if all items are hidden, do not show the record
1959 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1963 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1964 for my $key ( sort keys %$onloan_items ) {
1965 (++$onloanitemscount > $maxitems) and last;
1966 push @onloan_items_loop, $onloan_items->{$key};
1968 for my $key ( sort keys %$other_items ) {
1969 (++$otheritemscount > $maxitems) and last;
1970 push @other_items_loop, $other_items->{$key};
1972 for my $key ( sort keys %$available_items ) {
1973 (++$availableitemscount > $maxitems) and last;
1974 push @available_items_loop, $available_items->{$key}
1977 # XSLT processing of some stuff
1978 # we fetched the sysprefs already before the loop through all retrieved record!
1980 $record_processor->options({
1981 frameworkcode => $fw,
1982 interface => $search_context->{'interface'}
1985 $record_processor->process($marcrecord);
1987 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1989 biblionumber => $oldbiblio->{biblionumber},
1990 record => $marcrecord,
1993 ? 'OPACXSLTResultsDisplay'
1994 : 'XSLTResultsDisplay'
1997 hidden_items => \@hiddenitems,
1998 xslt_variables => $xslt_variables
2003 my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2004 $oldbiblio->{biblio_object} = $biblio_object;
2006 my $can_place_holds = 1;
2007 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2008 if (!C4::Context->preference("item-level_itypes")) {
2009 if ($itemtype && $itemtype->{notforloan}) {
2010 $can_place_holds = 0;
2013 $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object;
2015 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2016 $oldbiblio->{items_count} = $items_count;
2017 $oldbiblio->{available_items_loop} = \@available_items_loop;
2018 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2019 $oldbiblio->{other_items_loop} = \@other_items_loop;
2020 $oldbiblio->{availablecount} = $available_count;
2021 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2022 $oldbiblio->{onloancount} = $onloan_count;
2023 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2024 $oldbiblio->{othercount} = $other_count;
2025 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2026 $oldbiblio->{withdrawncount} = $withdrawn_count;
2027 $oldbiblio->{itemlostcount} = $itemlost_count;
2028 $oldbiblio->{damagedcount} = $itemdamaged_count;
2029 $oldbiblio->{intransitcount} = $item_in_transit_count;
2030 $oldbiblio->{onholdcount} = $item_onhold_count;
2031 $oldbiblio->{recalledcount} = $item_recalled_count;
2032 $oldbiblio->{orderedcount} = $ordered_count;
2033 $oldbiblio->{notforloancount} = $notforloan_count;
2035 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2036 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2037 my $subfields = substr $fieldspec, 3;
2038 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2039 my @alternateholdingsinfo = ();
2040 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2041 my $alternateholdingscount = 0;
2043 for my $field (@holdingsfields) {
2044 my %holding = ( holding => '' );
2045 my $havesubfield = 0;
2046 for my $subfield ($field->subfields()) {
2047 if ((index $subfields, $$subfield[0]) >= 0) {
2048 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2049 $holding{'holding'} .= $$subfield[1];
2053 if ($havesubfield) {
2054 push(@alternateholdingsinfo, \%holding);
2055 $alternateholdingscount++;
2059 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2060 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2063 push( @newresults, $oldbiblio );
2069 =head2 enabled_staff_search_views
2071 %hash = enabled_staff_search_views()
2073 This function returns a hash that contains three flags obtained from the system
2074 preferences, used to determine whether a particular staff search results view
2079 =item C<Output arg:>
2081 * $hash{can_view_MARC} is true only if the MARC view is enabled
2082 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2083 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2085 =item C<usage in the script:>
2089 $template->param ( C4::Search::enabled_staff_search_views );
2093 sub enabled_staff_search_views
2096 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2097 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2098 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2102 =head2 z3950_search_args
2104 $arrayref = z3950_search_args($matchpoints)
2106 This function returns an array reference that contains the search parameters to be
2107 passed to the Z39.50 search script (z3950_search.pl). The array elements
2108 are hash refs whose keys are name and value, and whose values are the
2109 name of a search parameter, the value of that search parameter and the URL encoded
2110 value of that parameter.
2112 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2114 The search parameter values are obtained from the bibliographic record whose
2115 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2117 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2118 a general purpose search argument. In this case, the returned array contains only
2119 entry: the key is 'title' and the value is derived from $matchpoints.
2121 If a search parameter value is undefined or empty, it is not included in the returned
2124 The returned array reference may be passed directly to the template parameters.
2128 =item C<Output arg:>
2130 * $array containing hash refs as described above
2132 =item C<usage in the script:>
2136 $data = Biblio::GetBiblioData($bibno);
2137 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2141 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2145 sub z3950_search_args {
2148 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2149 my $isbn = Business::ISBN->new( $isbn_string );
2151 if (defined $isbn && $isbn->is_valid)
2153 if ( ref($bibrec) ) {
2154 $bibrec->{isbn} = $isbn_string;
2155 $bibrec->{title} = undef;
2157 $bibrec = { isbn => $isbn_string };
2161 $bibrec = { title => $bibrec } if !ref $bibrec;
2164 for my $field (qw/ lccn isbn issn title author dewey subject /)
2166 push @$array, { name => $field, value => $bibrec->{$field} }
2167 if defined $bibrec->{$field};
2172 =head2 GetDistinctValues($field);
2174 C<$field> is a reference to the fields array
2178 sub GetDistinctValues {
2179 my ($fieldname,$string)=@_;
2180 # returns a reference to a hash of references to branches...
2181 if ($fieldname=~/\./){
2182 my ($table,$column)=split /\./, $fieldname;
2183 my $dbh = C4::Context->dbh;
2184 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 ");
2186 my $elements=$sth->fetchall_arrayref({});
2191 my @servers=qw<biblioserver authorityserver>;
2192 my (@zconns,@results);
2193 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2194 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2197 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2200 # The big moment: asynchronously retrieve results from all servers
2206 my ( $i, $size ) = @_;
2207 for ( my $j = 0 ; $j < $size ; $j++ ) {
2209 @hashscan{qw(value cnt)} =
2210 $results[ $i - 1 ]->display_term($j);
2211 push @elements, \%hashscan;
2219 =head2 _ZOOM_event_loop
2221 _ZOOM_event_loop(\@zconns, \@results, sub {
2222 my ( $i, $size ) = @_;
2226 Processes a ZOOM event loop and passes control to a closure for
2227 processing the results, and destroying the resultsets.
2231 sub _ZOOM_event_loop {
2232 my ($zconns, $results, $callback) = @_;
2233 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2234 my $ev = $zconns->[ $i - 1 ]->last_event();
2235 if ( $ev == ZOOM::Event::ZEND ) {
2236 next unless $results->[ $i - 1 ];
2237 my $size = $results->[ $i - 1 ]->size();
2239 $callback->($i, $size);
2244 foreach my $result (@$results) {
2249 =head2 new_record_from_zebra
2251 Given raw data from a searchengine result set, return a MARC::Record object
2253 This helper function is needed to take into account all the involved
2254 system preferences and configuration variables to properly create the
2255 MARC::Record object.
2257 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2258 data. If we are using DOM, then it has to be MARCXML.
2260 If we are using elasticsearch, it'll already be a MARC::Record and this
2261 function needs a new name.
2265 sub new_record_from_zebra {
2268 my $raw_data = shift;
2269 # Set the default indexing modes
2270 my $search_engine = C4::Context->preference("SearchEngine");
2271 if ($search_engine eq 'Elasticsearch') {
2272 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2274 my $index_mode = ( $server eq 'biblioserver' )
2275 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2276 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2278 my $marc_record = eval {
2279 if ( $index_mode eq 'dom' ) {
2280 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2282 MARC::Record->new_from_usmarc( $raw_data );
2289 return $marc_record;
2294 END { } # module clean-up code here (global destructor)
2301 Koha Development Team <http://koha-community.org/>