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;
34 use Koha::RecordProcessor;
40 our (@ISA, @EXPORT_OK);
51 enabled_staff_search_views
60 C4::Search - Functions for searching the Koha catalog.
64 See opac/opac-search.pl or catalogue/search.pl for example of usage
68 This module provides searching functions for Koha's bibliographic databases
74 # make all your functions, whether exported or not;
78 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
80 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
86 my $dbh = C4::Context->dbh;
87 my $result = TransformMarcToKoha( $record, '' );
91 # search duplicate on ISBN, easy and fast..
93 if ( $result->{isbn} ) {
94 $result->{isbn} =~ s/\(.*$//;
95 $result->{isbn} =~ s/\s+$//;
96 $query = "isbn:$result->{isbn}";
100 my $titleindex = 'ti,ext';
101 my $authorindex = 'au,ext';
104 $result->{title} =~ s /\\//g;
105 $result->{title} =~ s /\"//g;
106 $result->{title} =~ s /\(//g;
107 $result->{title} =~ s /\)//g;
109 $query = "$titleindex:\"$result->{title}\"";
110 if ( $result->{author} ) {
111 $result->{author} =~ s /\\//g;
112 $result->{author} =~ s /\"//g;
113 $result->{author} =~ s /\(//g;
114 $result->{author} =~ s /\)//g;
116 $query .= " $op $authorindex:\"$result->{author}\"";
120 my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
121 my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50);
123 if (!defined $error) {
124 foreach my $possible_duplicate_record (@{$searchresults}) {
125 my $marcrecord = new_record_from_zebra(
127 $possible_duplicate_record
130 my $result = TransformMarcToKoha( $marcrecord, '' );
132 # FIXME :: why 2 $biblionumber ?
134 push @results, $result->{'biblionumber'};
135 push @results, $result->{'title'};
144 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
146 This function provides a simple search API on the bibliographic catalog
152 * $query can be a simple keyword or a complete CCL query
153 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
154 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
155 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
156 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
161 Returns an array consisting of three elements
162 * $error is undefined unless an error is detected
163 * $results is a reference to an array of records.
164 * $total_hits is the number of hits that would have been returned with no limit
166 If an error is returned the two other return elements are undefined. If error itself is undefined
167 the other two elements are always defined
169 =item C<usage in the script:>
173 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
175 if (defined $error) {
176 $template->param(query_error => $error);
177 warn "error: ".$error;
178 output_html_with_http_headers $input, $cookie, $template->output;
182 my $hits = @{$marcresults};
185 for my $r ( @{$marcresults} ) {
186 my $marcrecord = MARC::File::USMARC::decode($r);
187 my $biblio = TransformMarcToKoha($marcrecord,q{});
189 #build the iarray of hashs for the template.
191 title => $biblio->{'title'},
192 subtitle => $biblio->{'subtitle'},
193 biblionumber => $biblio->{'biblionumber'},
194 author => $biblio->{'author'},
195 publishercode => $biblio->{'publishercode'},
196 publicationyear => $biblio->{'publicationyear'},
201 $template->param(result=>\@results);
206 my ( $query, $offset, $max_results, $servers, %options ) = @_;
208 return ( 'No query entered', undef, undef ) unless $query;
209 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
210 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
217 # Initialize & Search Zebra
218 for ( my $i = 0 ; $i < @servers ; $i++ ) {
220 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
221 $query =~ s/:/=/g unless $options{skip_normalize};
222 $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]);
223 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
227 $zconns[$i]->errmsg() . " ("
228 . $zconns[$i]->errcode() . ") "
229 . $zconns[$i]->addinfo() . " "
230 . $zconns[$i]->diagset();
232 return ( $error, undef, undef ) if $zconns[$i]->errcode();
236 # caught a ZOOM::Exception
240 . $@->addinfo() . " "
242 warn $error." for query: $query";
243 return ( $error, undef, undef );
252 my $first_record = defined($offset) ? $offset + 1 : 1;
253 my $hits = $tmpresults[ $i - 1 ]->size();
254 $total_hits += $hits;
255 my $last_record = $hits;
256 if ( defined $max_results && $offset + $max_results < $hits ) {
257 $last_record = $offset + $max_results;
260 for my $j ( $first_record .. $last_record ) {
262 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
265 push @{$results}, $record if defined $record;
270 foreach my $zoom_query (@zoom_queries) {
271 $zoom_query->destroy();
274 return ( undef, $results, $total_hits );
279 ( undef, $results_hashref, \@facets_loop ) = getRecords (
281 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
282 $results_per_page, $offset, $branches, $itemtypes,
283 $query_type, $scan, $opac
286 The all singing, all dancing, multi-server, asynchronous, scanning,
287 searching, record nabbing, facet-building
289 See verbose embedded documentation.
295 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
296 $results_per_page, $offset, $branches, $itemtypes,
297 $query_type, $scan, $opac
300 my @servers = @$servers_ref;
301 my @sort_by = @$sort_by_ref;
302 $offset = 0 if $offset < 0;
304 # Initialize variables for the ZOOM connection and results object
307 my $results_hashref = ();
309 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
310 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
312 # Initialize variables for the faceted results objects
313 my $facets_counter = {};
314 my $facets_info = {};
315 my $facets = getFacets();
317 my @facets_loop; # stores the ref to array of hashes for template facets loop
319 ### LOOP THROUGH THE SERVERS
320 for ( my $i = 0 ; $i < @servers ; $i++ ) {
321 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
323 # perform the search, create the results objects
324 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
325 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
327 Koha::Logger->get->debug($simple_query) if $scan;
329 # Check if we've got a query_type defined, if so, use it
332 if ($query_type =~ /^ccl/) {
333 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
334 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
335 } elsif ($query_type =~ /^cql/) {
336 $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i]));
337 } elsif ($query_type =~ /^pqf/) {
338 $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i]));
340 warn "Unknown query_type '$query_type'. Results undetermined.";
343 $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
345 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
349 warn "WARNING: query problem with $query_to_use " . $@;
352 # Concatenate the sort_by limits and pass them to the results object
353 # Note: sort will override rank
355 foreach my $sort (@sort_by) {
356 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
357 $sort_by .= "1=1003 <i ";
359 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
360 $sort_by .= "1=1003 >i ";
362 elsif ( $sort eq "popularity_asc" ) {
363 $sort_by .= "1=9003 <i ";
365 elsif ( $sort eq "popularity_dsc" ) {
366 $sort_by .= "1=9003 >i ";
368 elsif ( $sort eq "call_number_asc" ) {
369 $sort_by .= "1=8007 <i ";
371 elsif ( $sort eq "call_number_dsc" ) {
372 $sort_by .= "1=8007 >i ";
374 elsif ( $sort eq "pubdate_asc" ) {
375 $sort_by .= "1=31 <i ";
377 elsif ( $sort eq "pubdate_dsc" ) {
378 $sort_by .= "1=31 >i ";
380 elsif ( $sort eq "acqdate_asc" ) {
381 $sort_by .= "1=32 <i ";
383 elsif ( $sort eq "acqdate_dsc" ) {
384 $sort_by .= "1=32 >i ";
386 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
387 $sort_by .= "1=4 <i ";
389 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
390 $sort_by .= "1=4 >i ";
393 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
396 if ( $sort_by && !$scan && $results[$i] ) {
397 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
398 warn "WARNING sort $sort_by failed";
401 } # finished looping through servers
403 # The big moment: asynchronously retrieve results from all servers
408 my ( $i, $size ) = @_;
411 # loop through the results
412 $results_hash->{'hits'} = $size;
414 if ( $offset + $results_per_page <= $size ) {
415 $times = $offset + $results_per_page;
421 for ( my $j = $offset ; $j < $times ; $j++ ) {
424 ## Check if it's an index scan
426 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
428 # here we create a minimal MARC record and hand it off to the
429 # template just like a normal result ... perhaps not ideal, but
431 my $tmprecord = MARC::Record->new();
432 $tmprecord->encoding('UTF-8');
436 # the minimal record in author/title (depending on MARC flavour)
437 if ( C4::Context->preference("marcflavour") eq
440 $tmptitle = MARC::Field->new(
445 $tmprecord->append_fields($tmptitle);
449 MARC::Field->new( '245', ' ', ' ', a => $term, );
451 MARC::Field->new( '100', ' ', ' ', a => $occ, );
452 $tmprecord->append_fields($tmptitle);
453 $tmprecord->append_fields($tmpauthor);
455 $results_hash->{'RECORDS'}[$j] =
456 $tmprecord->as_usmarc();
461 $record = $results[ $i - 1 ]->record($j)->raw();
462 # warn "RECORD $j:".$record;
463 $results_hash->{'RECORDS'}[$j] = $record;
467 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
469 # Fill the facets while we're looping, but only for the
470 # biblioserver and not for a scan
471 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
472 $facets_counter = GetFacets( $results[ $i - 1 ] );
473 $facets_info = _get_facets_info( $facets );
477 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
479 sort { $a cmp $b } keys %$facets_counter
482 my @this_facets_array;
485 $facets_counter->{$link_value}
486 ->{$b} <=> $facets_counter->{$link_value}
488 } keys %{ $facets_counter->{$link_value} }
491 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
492 my $facet_link_value = $one_facet;
493 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
495 # fix the length that will display in the label,
496 my $facet_label_value = $one_facet;
497 my $facet_max_length = C4::Context->preference(
498 'FacetLabelTruncationLength')
501 substr( $one_facet, 0, $facet_max_length )
503 if length($facet_label_value) >
506 # if it's a branch, label by the name, not the code,
507 if ( $link_value =~ /branch/ ) {
508 if ( defined $branches
509 && ref($branches) eq "HASH"
510 && defined $branches->{$one_facet}
511 && ref( $branches->{$one_facet} ) eq
515 $branches->{$one_facet}
519 $facet_label_value = "*";
523 # if it's a itemtype, label by the name, not the code,
524 if ( $link_value =~ /itype/ ) {
525 if ( defined $itemtypes
526 && ref($itemtypes) eq "HASH"
527 && defined $itemtypes->{$one_facet}
528 && ref( $itemtypes->{$one_facet} ) eq
532 $itemtypes->{$one_facet}
533 ->{translated_description};
537 # also, if it's a location code, use the name instead of the code
538 if ( $link_value =~ /location/ ) {
539 # TODO Retrieve all authorised values at once, instead of 1 query per entry
540 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
541 $facet_label_value = $av->count ? $av->next->opac_description : '';
544 # also, if it's a collection code, use the name instead of the code
545 if ( $link_value =~ /ccode/ ) {
546 # TODO Retrieve all authorised values at once, instead of 1 query per entry
547 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
548 $facet_label_value = $av->count ? $av->next->opac_description : '';
551 # but we're down with the whole label being in the link's title.
552 push @this_facets_array,
555 $facets_counter->{$link_value}
557 facet_label_value => $facet_label_value,
558 facet_title_value => $one_facet,
559 facet_link_value => $facet_link_value,
560 type_link_value => $link_value,
562 if ($facet_label_value);
567 type_link_value => $link_value,
568 type_id => $link_value . "_id",
570 . $facets_info->{$link_value}->{'label_value'} =>
572 facets => \@this_facets_array,
576 $facets_info->{$link_value}->{'label_value'} =~
579 and ( Koha::Libraries->search->count == 1 )
586 # This sorts the facets into alphabetical order
588 foreach my $f (@facets_loop) {
589 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
591 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
596 return ( undef, $results_hashref, \@facets_loop );
604 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
606 if ( $use_zebra_facets ) {
607 $facets = _get_facets_from_zebra( $rs );
609 $facets = _get_facets_from_records( $rs );
615 sub _get_facets_from_records {
619 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
620 my $facets_config = getFacets();
622 my $size = $rs->size();
623 my $jmax = $size > $facets_maxrecs
627 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
629 my $marc_record = new_record_from_zebra (
631 $rs->record( $j )->raw()
634 if ( ! defined $marc_record ) {
635 warn "ERROR DECODING RECORD - $@: " .
636 $rs->record( $j )->raw();
640 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
646 =head2 _get_facets_data_from_record
648 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
650 Internal function that extracts facets information from a MARC::Record object
651 and populates $facets_counter for using in getRecords.
653 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
658 sub _get_facets_data_from_record {
660 my ( $marc_record, $facets, $facets_counter ) = @_;
662 for my $facet (@$facets) {
666 foreach my $tag ( @{ $facet->{ tags } } ) {
668 # tag number is the first three digits
669 my $tag_num = substr( $tag, 0, 3 );
670 # subfields are the remainder
671 my $subfield_letters = substr( $tag, 3 );
673 my @fields = $marc_record->field( $tag_num );
674 foreach my $field (@fields) {
675 # If $field->indicator(1) eq 'z', it means it is a 'see from'
676 # field introduced because of IncludeSeeFromInSearches, so skip it
677 next if $field->indicator(1) eq 'z';
679 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
680 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
682 unless ( grep { $_ eq $data } @used_datas ) {
683 push @used_datas, $data;
684 $facets_counter->{ $facet->{ idx } }->{ $data }++;
691 =head2 _get_facets_from_zebra
693 my $facets = _get_facets_from_zebra( $result_set )
695 Retrieves facets for a specified result set. It loops through the facets defined
696 in C4::Koha::getFacets and returns a hash with the following structure:
706 sub _get_facets_from_zebra {
710 # save current elementSetName
711 my $elementSetName = $rs->option( 'elementSetName' );
713 my $facets_loop = getFacets();
714 my $facets_data = {};
715 # loop through defined facets and fill the facets hashref
716 foreach my $facet ( @$facets_loop ) {
718 my $idx = $facet->{ idx };
719 my $sep = $facet->{ sep };
720 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
721 if ( $facet_values ) {
722 # we've actually got a result
723 $facets_data->{ $idx } = $facet_values;
726 # set elementSetName to its previous value to avoid side effects
727 $rs->option( elementSetName => $elementSetName );
732 =head2 _get_facet_from_result_set
735 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
737 Internal function that extracts facet information for a specific index ($facet_idx) and
738 returns a hash containing facet values and count:
741 $facet_value => $count ,
745 Warning: this function has the side effect of changing the elementSetName for the result
746 set. It is a helper function for the main loop, which takes care of backing it up for
751 sub _get_facet_from_result_set {
753 my $facet_idx = shift;
757 my $internal_sep = '<*>';
758 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
760 return if ( ! defined $facet_idx || ! defined $rs );
761 # zebra's facet element, untokenized index
762 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
763 # configure zebra results for retrieving the desired facet
764 $rs->option( elementSetName => $facet_element );
765 # get the facet record from result set
766 my $facet = $rs->record( 0 )->raw;
767 # if the facet has no restuls...
768 return if !defined $facet;
769 # TODO: benchmark DOM vs. SAX performance
770 my $facet_dom = XML::LibXML->load_xml(
773 my @terms = $facet_dom->getElementsByTagName('term');
777 foreach my $term ( @terms ) {
778 my $facet_value = $term->textContent;
779 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
780 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
781 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
787 =head2 _get_facets_info
789 my $facets_info = C4::Search::_get_facets_info( $facets )
791 Internal function that extracts facets information and properly builds
792 the data structure needed to render facet labels.
796 sub _get_facets_info {
800 my $facets_info = {};
802 for my $facet ( @$facets ) {
803 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
810 sub _detect_truncation {
811 my ( $operand, $index ) = @_;
812 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
815 my @wordlist = split( /\s/, $operand );
816 foreach my $word (@wordlist) {
817 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
818 push @rightlefttruncated, $word;
820 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
821 push @lefttruncated, $word;
823 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
824 push @righttruncated, $word;
826 elsif ( index( $word, "*" ) < 0 ) {
827 push @nontruncated, $word;
830 push @regexpr, $word;
834 \@nontruncated, \@righttruncated, \@lefttruncated,
835 \@rightlefttruncated, \@regexpr
840 sub _build_stemmed_operand {
841 my ($operand,$lang) = @_;
842 require Lingua::Stem::Snowball ;
843 my $stemmed_operand=q{};
845 # Stemmer needs language
846 return $operand unless $lang;
848 # If operand contains a digit, it is almost certainly an identifier, and should
849 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
850 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
851 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
852 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
853 return $operand if $operand =~ /\d/;
855 # FIXME: the locale should be set based on the user's language and/or search choice
857 # Make sure we only use the first two letters from the language code
858 $lang = lc(substr($lang, 0, 2));
859 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
860 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
861 if ($lang eq 'nb' || $lang eq 'nn') {
864 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
865 encoding => "UTF-8" );
867 my @words = split( / /, $operand );
868 my @stems = $stemmer->stem(\@words);
869 for my $stem (@stems) {
870 $stemmed_operand .= "$stem";
871 $stemmed_operand .= "?"
872 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
873 $stemmed_operand .= " ";
876 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
877 return $stemmed_operand;
881 sub _build_weighted_query {
883 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
884 # pretty well but could work much better if we had a smarter query parser
885 my ( $operand, $stemmed_operand, $index ) = @_;
886 my $stemming = C4::Context->preference("QueryStemming") || 0;
887 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
888 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
889 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
891 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
893 # Keyword, or, no index specified
894 if ( ( $index eq 'kw' ) || ( !$index ) ) {
896 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
897 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
898 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
899 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
900 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
901 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
902 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
903 if $fuzzy_enabled; # add fuzzy, word list
904 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
905 if ( $stemming and $stemmed_operand )
906 ; # add stemming, right truncation
907 $weighted_query .= " or wrdl,r9=\"$operand\"";
909 # embedded sorting: 0 a-z; 1 z-a
910 # $weighted_query .= ") or (sort1,aut=1";
913 # Barcode searches should skip this process
914 elsif ( $index eq 'bc' ) {
915 $weighted_query .= "bc=\"$operand\"";
918 # Authority-number searches should skip this process
919 elsif ( $index eq 'an' ) {
920 $weighted_query .= "an=\"$operand\"";
923 # If the index is numeric, don't autoquote it.
924 elsif ( $index =~ /,st-numeric$/ ) {
925 $weighted_query .= " $index=$operand";
928 # If the index already has more than one qualifier, wrap the operand
929 # in quotes and pass it back (assumption is that the user knows what they
930 # are doing and won't appreciate us mucking up their query
931 elsif ( $index =~ ',' ) {
932 $weighted_query .= " $index=\"$operand\"";
935 #TODO: build better cases based on specific search indexes
937 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
938 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
939 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
940 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
941 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
942 if $fuzzy_enabled; # add fuzzy, word list
943 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
944 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
947 $weighted_query .= "))"; # close rank specification
948 return $weighted_query;
953 Return an array with available indexes.
977 'Author-personal-bibliography',
987 'Chronological-subdivision',
997 'Conference-name-heading',
998 'Conference-name-see',
999 'Conference-name-seealso',
1002 'Control-number-identifier',
1006 'Corporate-name-heading',
1007 'Corporate-name-see',
1008 'Corporate-name-seealso',
1009 'Country-publication',
1012 'date-entered-on-file',
1013 'Date-of-acquisition',
1014 'Date-of-publication',
1015 'Date-time-last-modified',
1016 'Dewey-classification',
1017 'Dissertation-information',
1026 'Geographic-subdivision',
1029 'Heading-use-main-or-added-entry',
1030 'Heading-use-series-added-entry ',
1031 'Heading-use-subject-added-entry',
1035 'Identifier-standard',
1036 'Illustration-code',
1038 'Index-term-uncontrolled',
1039 'Interest-age-level',
1040 'Interest-grade-level',
1050 'language-original',
1060 'Local-classification',
1063 'Match-heading-see-from',
1068 'Multipart-resource-level',
1073 'Name-geographic-heading',
1074 'Name-geographic-see',
1075 'Name-geographic-seealso',
1081 'Other-control-number',
1084 'Personal-name-heading',
1085 'Personal-name-see',
1086 'Personal-name-seealso',
1088 'Place-publication',
1095 'Reading-grade-level',
1096 'Record-control-number',
1107 'Subject-heading-thesaurus',
1108 'Subject-name-personal',
1109 'Subject-subdivision',
1118 'Term-genre-form-heading',
1119 'Term-genre-form-see',
1120 'Term-genre-form-seealso',
1126 'Title-uniform-heading',
1127 'Title-uniform-see',
1128 'Title-uniform-seealso',
1138 'classification-source',
1140 'coded-location-qualifier',
1151 'Local-classification',
1154 'materials-specified',
1159 'Number-local-acquisition',
1164 'replacementpricedate',
1182 $simple_query, $query_cgi,
1183 $query_desc, $limit,
1184 $limit_cgi, $limit_desc,
1185 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1187 Build queries and limits in CCL, CGI, Human,
1188 handle truncation, stemming, field weighting, fuzziness, etc.
1190 See verbose embedded documentation.
1196 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1201 my @operators = $operators ? @$operators : ();
1202 my @indexes = $indexes ? @$indexes : ();
1203 my @operands = $operands ? @$operands : ();
1204 my @limits = $limits ? @$limits : ();
1205 my @sort_by = $sort_by ? @$sort_by : ();
1207 my $stemming = C4::Context->preference("QueryStemming") || 0;
1208 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1209 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1210 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1212 my $query = $operands[0] // "";
1213 my $simple_query = $operands[0];
1215 # initialize the variables we're passing back
1224 my $cclindexes = getIndexes();
1225 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1226 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1228 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1230 $query = "ccl=$query" if $cclq;
1233 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1235 if ( $query =~ /^ccl=/ ) {
1237 # This is needed otherwise ccl= and &limit won't work together, and
1238 # this happens when selecting a subject on the opac-detail page
1239 @limits = grep {!/^$/} @limits;
1240 my $original_q = $q; # without available part
1241 unless ( grep { $_ eq 'available' } @limits ) {
1242 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1246 if ( grep { $_ eq 'available' } @limits ) {
1247 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1248 @limits = grep {!/^available$/} @limits;
1250 $q .= ' and '.join(' and ', @limits) if @limits;
1252 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1254 if ( $query =~ /^cql=/ ) {
1255 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1257 if ( $query =~ /^pqf=/ ) {
1259 $query_cgi = "q=pqf=".uri_escape_utf8($');
1260 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1263 # pass nested queries directly
1264 # FIXME: need better handling of some of these variables in this case
1265 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1266 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1267 # if ( $query =~ /(\(|\))/ ) {
1269 # undef, $query, $simple_query, $query_cgi,
1270 # $query, $limit, $limit_cgi, $limit_desc,
1275 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1276 # query operands and indexes and add stemming, truncation, field weighting, etc.
1277 # Once we do so, we'll end up with a value in $query, just like if we had an
1278 # incoming $query from the user
1281 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1282 my $previous_operand
1283 ; # a flag used to keep track if there was a previous query
1284 # if there was, we can apply the current operator
1286 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1288 # COMBINE OPERANDS, INDEXES AND OPERATORS
1289 if ( ($operands[$i] // '') ne '' ) {
1290 $operands[$i]=~s/^\s+//;
1292 # A flag to determine whether or not to add the index to the query
1295 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1296 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1300 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1302 my $operand = $operands[$i];
1303 my $index = $indexes[$i] || 'kw';
1305 # Add index-specific attributes
1307 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1308 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1309 #which is processed higher up in this sub. Other than that, year searches are typically
1310 #handled as limits which are not processed her either.
1312 # Search ranges: Date of Publication, st-numeric
1313 if ( $index =~ /(yr|st-numeric)/ ) {
1314 #weight_fields/relevance search causes errors with date ranges
1315 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1316 #In the case of YYYY-YYYY, it will return no results
1317 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1320 # Date of Acquisition
1321 elsif ( $index =~ /acqdate/ ) {
1322 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1323 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1324 #top of the results just because they have lots of item records matching that date.
1325 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1327 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1329 # ISBN,ISSN,Standard Number, don't need special treatment
1330 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1332 $stemming, $auto_truncation,
1333 $weight_fields, $fuzzy_enabled
1336 if ( $index eq 'nb' ) {
1337 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1338 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1339 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1340 $indexes[$i] = $index = 'kw';
1345 # Set default structure attribute (word list)
1346 my $struct_attr = q{};
1347 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1348 $struct_attr = ",wrdl";
1351 # Some helpful index variants
1352 my $index_plus = $index . $struct_attr . ':';
1353 my $index_plus_comma = $index . $struct_attr . ',';
1355 if ($auto_truncation){
1356 unless ( $index =~ /,(st-|phr|ext)/ ) {
1357 #FIXME only valid with LTR scripts
1358 $operand=join(" ",map{
1359 (index($_,"*")>0?"$_":"$_*")
1360 }split (/\s+/,$operand));
1365 my $truncated_operand = q{};
1366 my( $nontruncated, $righttruncated, $lefttruncated,
1367 $rightlefttruncated, $regexpr
1368 ) = _detect_truncation( $operand, $index );
1370 Koha::Logger->get->debug(
1371 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1375 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1376 scalar(@$rightlefttruncated) > 0 )
1379 # Don't field weight or add the index to the query, we do it here
1381 undef $weight_fields;
1382 my $previous_truncation_operand;
1383 if (scalar @$nontruncated) {
1384 $truncated_operand .= "$index_plus @$nontruncated ";
1385 $previous_truncation_operand = 1;
1387 if (scalar @$righttruncated) {
1388 $truncated_operand .= "and " if $previous_truncation_operand;
1389 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1390 $previous_truncation_operand = 1;
1392 if (scalar @$lefttruncated) {
1393 $truncated_operand .= "and " if $previous_truncation_operand;
1394 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1395 $previous_truncation_operand = 1;
1397 if (scalar @$rightlefttruncated) {
1398 $truncated_operand .= "and " if $previous_truncation_operand;
1399 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1400 $previous_truncation_operand = 1;
1403 $operand = $truncated_operand if $truncated_operand;
1404 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1407 my $stemmed_operand = q{};
1408 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1411 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1413 # Handle Field Weighting
1414 my $weighted_operand = q{};
1415 if ($weight_fields) {
1416 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1417 $operand = $weighted_operand;
1421 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1423 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1425 #N.B. Truncation is mutually exclusive with Weighted Queries,
1426 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1427 #the need for this relevance wrapper.
1428 $operand = "(rk=($operand))" unless $weight_fields;
1430 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1432 query_cgi => $query_cgi,
1433 query_desc => $query_desc,
1434 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1435 parsed_operand => $operand,
1436 original_operand => $operands[$i] // '',
1438 index_plus => $index_plus,
1439 indexes_set => $indexes_set,
1440 previous_operand => $previous_operand,
1446 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1449 my %group_OR_limits;
1450 my $availability_limit;
1451 foreach my $this_limit (@limits) {
1452 next unless $this_limit;
1453 if ( $this_limit =~ /available/ ) {
1455 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1457 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1458 $availability_limit .=
1459 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1460 $limit_cgi .= "&limit=available";
1464 # group_OR_limits, prefixed by mc-
1465 # OR every member of the group
1466 elsif ( $this_limit =~ /mc/ ) {
1467 my ($k,$v) = split(/:/, $this_limit,2);
1468 if ( $k !~ /mc-i(tem)?type/ ) {
1469 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1470 $this_limit =~ tr/"//d;
1471 $this_limit = $k.':"'.$v.'"';
1474 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1475 $limit_desc .= " or " if $group_OR_limits{$k};
1476 $group_OR_limits{$k} .= "$this_limit";
1477 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1478 $limit_desc .= " $this_limit";
1480 elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) {
1481 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1482 $limit .= " and " if $limit || $query;
1483 my $branchfield = C4::Context->preference('SearchLimitLibrary');
1485 if( $this_limit =~ '^multibranchlimit:' ){
1486 my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ );
1487 my $search_group = Koha::Library::Groups->find( $group_id );
1488 @branchcodes = map { $_->branchcode } $search_group->all_libraries;
1489 @branchcodes = sort { $a cmp $b } @branchcodes;
1491 @branchcodes = ( $this_limit =~ /^branch:(.*)$/ );
1495 if ( $branchfield eq "homebranch" ) {
1496 $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes;
1498 elsif ( $branchfield eq "holdingbranch" ) {
1499 $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes;
1502 $this_limit = sprintf "(%s or %s)",
1503 join( " or ", map { 'homebranch: ' . $_ } @branchcodes ),
1504 join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes );
1507 $limit .= "$this_limit";
1508 $limit_desc .= " $this_limit";
1511 # Regular old limits
1513 $limit .= " and " if $limit || $query;
1514 $limit .= "$this_limit";
1515 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1516 $limit_desc .= " $this_limit";
1519 foreach my $k (keys (%group_OR_limits)) {
1520 $limit .= " and " if ( $query || $limit );
1521 $limit .= "($group_OR_limits{$k})";
1523 if ($availability_limit) {
1524 $limit .= " and " if ( $query || $limit );
1525 $limit .= "($availability_limit)";
1528 # Normalize the query and limit strings
1529 # This is flawed , means we can't search anything with : in it
1530 # if user wants to do ccl or cql, start the query with that
1531 # $query =~ s/:/=/g;
1532 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1533 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1534 $query =~ s/(?<=(wrdl)):/=/g;
1535 $query =~ s/(?<=(trn|phr)):/=/g;
1536 $query =~ s/(?<=(st-numeric)):/=/g;
1537 $query =~ s/(?<=(st-year)):/=/g;
1538 $query =~ s/(?<=(st-date-normalized)):/=/g;
1540 # Removing warnings for later substitutions
1542 $query_desc //= q{};
1545 $limit_desc //= q{};
1547 $simple_query //= q{};
1549 for ( $query, $query_desc, $limit, $limit_desc ) {
1550 s/ +/ /g; # remove extra spaces
1551 s/^ //g; # remove any beginning spaces
1552 s/ $//g; # remove any ending spaces
1553 s/==/=/g; # remove double == from query
1555 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1557 for ($query_cgi,$simple_query) {
1560 # append the limit to the query
1561 $query .= " " . $limit;
1563 Koha::Logger->get->debug(
1564 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1565 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1568 undef, $query, $simple_query, $query_cgi,
1569 $query_desc, $limit, $limit_cgi, $limit_desc,
1574 =head2 _build_initial_query
1576 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1578 Build a section of the initial query containing indexes, operators, and operands.
1582 sub _build_initial_query {
1586 if ($params->{previous_operand}){
1587 #If there is a previous operand, add a supplied operator or the default 'and'
1588 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1591 #NOTE: indexes_set is typically set when doing truncation or field weighting
1592 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1594 #e.g. "kw,wrdl:test"
1595 #e.g. " and kw,wrdl:test"
1596 $params->{query} .= $operator . $operand;
1598 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1599 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1600 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1602 #e.g. " and kw,wrdl: test"
1603 $params->{query_desc} .= $operator . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1605 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1607 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1610 =head2 searchResults
1612 my @search_results = searchResults($search_context, $searchdesc, $hits,
1613 $results_per_page, $offset, $scan,
1616 Format results in a form suitable for passing to the template
1620 # IMO this subroutine is pretty messy still -- it's responsible for
1621 # building the HTML output for the template
1623 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1624 my $dbh = C4::Context->dbh;
1629 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1630 my ($is_opac, $hidelostitems);
1631 if ($search_context->{'interface'} eq 'opac') {
1632 $hidelostitems = C4::Context->preference('hidelostitems');
1636 my $record_processor = Koha::RecordProcessor->new({
1637 filters => 'ViewPolicy'
1640 #Build branchnames hash
1641 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1643 # FIXME - We build an authorised values hash here, using the default framework
1644 # though it is possible to have different authvals for different fws.
1646 my $shelflocations =
1647 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1649 # get notforloan authorised value list (see $shelflocations FIXME)
1650 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1651 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1654 my $itemtypes = Koha::ItemTypes->search_with_localization;
1655 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1657 #search item field code
1658 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1660 ## find column names of items related to MARC
1661 my %subfieldstosearch;
1662 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1663 for my $column ( @columns ) {
1664 my ( $tagfield, $tagsubfield ) =
1665 &GetMarcFromKohaField( "items." . $column );
1666 if ( defined $tagsubfield ) {
1667 $subfieldstosearch{$column} = $tagsubfield;
1671 # handle which records to actually retrieve
1672 my $times; # Times is which record to process up to
1673 if ( $hits && $offset + $results_per_page <= $hits ) {
1674 $times = $offset + $results_per_page;
1677 $times = $hits; # If less hits than results_per_page+offset we go to the end
1680 my $marcflavour = C4::Context->preference("marcflavour");
1681 # We get the biblionumber position in MARC
1682 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1684 # set stuff for XSLT processing here once, not later again for every record we retrieved
1686 my $userenv = C4::Context->userenv;
1688 = ( defined $userenv and $userenv->{number} )
1689 ? Koha::Patrons->find( $userenv->{number} )
1691 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1693 # loop through all of the records we've retrieved
1694 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1698 # For Scan searches we built USMARC data
1699 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1701 # Normal search, render from Zebra's output
1702 $marcrecord = new_record_from_zebra(
1707 if ( ! defined $marcrecord ) {
1708 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1716 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1717 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1719 SetUTF8Flag($marcrecord);
1720 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw, 'no_items' );
1721 $oldbiblio->{result_number} = $i + 1;
1723 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1724 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1725 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1726 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1727 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1729 # edition information, if any
1730 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1732 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1733 # add imageurl to itemtype if there is one
1734 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1735 # Build summary if there is one (the summary is defined in the itemtypes table)
1736 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1738 # Pull out the items fields
1739 my @fields = $marcrecord->field($itemtag);
1740 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1741 my $marcflavor = C4::Context->preference("marcflavour");
1743 # adding linked items that belong to host records
1744 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1745 my $analyticsfield = '773';
1746 if ($marcflavor eq 'MARC21') {
1747 $analyticsfield = '773';
1748 } elsif ($marcflavor eq 'UNIMARC') {
1749 $analyticsfield = '461';
1751 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1752 my $hostbiblionumber = $hostfield->subfield("0");
1753 my $linkeditemnumber = $hostfield->subfield("9");
1754 if( $hostbiblionumber ) {
1755 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1756 if ($linkeditemmarc) {
1757 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1758 if ($linkeditemfield) {
1759 push( @fields, $linkeditemfield );
1766 # Setting item statuses for display
1767 my @available_items_loop;
1768 my @onloan_items_loop;
1769 my @other_items_loop;
1771 my $available_items;
1775 my $ordered_count = 0;
1776 my $available_count = 0;
1777 my $onloan_count = 0;
1778 my $longoverdue_count = 0;
1779 my $other_count = 0;
1780 my $withdrawn_count = 0;
1781 my $itemlost_count = 0;
1782 my $hideatopac_count = 0;
1783 my $itembinding_count = 0;
1784 my $itemdamaged_count = 0;
1785 my $item_in_transit_count = 0;
1786 my $item_onhold_count = 0;
1787 my $notforloan_count = 0;
1788 my $items_count = scalar(@fields);
1789 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1790 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1791 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1793 # loop through every item
1794 foreach my $field (@fields) {
1797 # populate the items hash
1798 foreach my $code ( keys %subfieldstosearch ) {
1799 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1801 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1805 # hidden because lost
1806 if ($hidelostitems && $item->{itemlost}) {
1807 $hideatopac_count++;
1810 # hidden based on OpacHiddenItems syspref
1811 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
1813 push @hiddenitems, @hi;
1814 $hideatopac_count++;
1819 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1820 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1822 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1823 if ($item->{$hbranch}) {
1824 $item->{'branchname'} = $branches{$item->{$hbranch}};
1826 elsif ($item->{$otherbranch}) { # Last resort
1827 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1831 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1832 . ( $item->{location} ? $item->{location} : q{} )
1833 . ( $item->{itype} ? $item->{itype} : q{} )
1834 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1835 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1836 if ( $item->{onloan}
1838 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1841 my $key = $prefix . $item->{onloan} . $item->{barcode};
1842 $onloan_items->{$key}->{due_date} = $item->{onloan};
1843 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1844 $onloan_items->{$key}->{branchname} = $item->{branchname};
1845 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1846 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1847 $onloan_items->{$key}->{description} = $item->{description};
1848 $onloan_items->{$key}->{imageurl} =
1849 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1851 # if something's checked out and lost, mark it as 'long overdue'
1852 if ( $item->{itemlost} ) {
1853 $onloan_items->{$key}->{longoverdue}++;
1854 $longoverdue_count++;
1858 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1861 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1862 $item->{notforloan} = 1 if !$item->{notforloan} &&
1863 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1866 if ( $item->{notforloan} < 0 ) {
1868 } elsif ( $item->{notforloan} > 0 ) {
1869 $notforloan_count++;
1872 # is item in transit?
1873 my $transfertwhen = '';
1874 my ($transfertfrom, $transfertto);
1876 # is item on the reserve shelf?
1877 my $reservestatus = '';
1879 unless ($item->{withdrawn}
1880 || $item->{itemlost}
1882 || $item->{notforloan}
1883 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1884 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1886 # A couple heuristics to limit how many times
1887 # we query the database for item transfer information, sacrificing
1888 # accuracy in some cases for speed;
1890 # 1. don't query if item has one of the other statuses
1891 # 2. don't check transit status if the bib has
1892 # more than 20 items
1894 # FIXME: to avoid having the query the database like this, and to make
1895 # the in transit status count as unavailable for search limiting,
1896 # should map transit status to record indexed in Zebra.
1898 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1899 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1902 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1903 if ( $item->{withdrawn}
1904 || $item->{itemlost}
1906 || $item->{notforloan}
1907 || $reservestatus eq 'Waiting'
1908 || ($transfertwhen && $transfertwhen ne ''))
1910 $withdrawn_count++ if $item->{withdrawn};
1911 $itemlost_count++ if $item->{itemlost};
1912 $itemdamaged_count++ if $item->{damaged};
1913 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1914 $item_onhold_count++ if $reservestatus eq 'Waiting';
1915 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1919 my $key = $prefix . $item->{status};
1920 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1921 $other_items->{$key}->{$_} = $item->{$_};
1923 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1924 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1925 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1926 $other_items->{$key}->{count}++ if $item->{$hbranch};
1927 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1928 $other_items->{$key}->{description} = $item->{description};
1929 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1934 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1935 foreach (qw(branchname itemcallnumber description)) {
1936 $available_items->{$prefix}->{$_} = $item->{$_};
1938 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1939 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1942 } # notforloan, item level and biblioitem level
1944 # if all items are hidden, do not show the record
1945 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1949 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1950 for my $key ( sort keys %$onloan_items ) {
1951 (++$onloanitemscount > $maxitems) and last;
1952 push @onloan_items_loop, $onloan_items->{$key};
1954 for my $key ( sort keys %$other_items ) {
1955 (++$otheritemscount > $maxitems) and last;
1956 push @other_items_loop, $other_items->{$key};
1958 for my $key ( sort keys %$available_items ) {
1959 (++$availableitemscount > $maxitems) and last;
1960 push @available_items_loop, $available_items->{$key}
1963 # XSLT processing of some stuff
1964 # we fetched the sysprefs already before the loop through all retrieved record!
1966 $record_processor->options({
1967 frameworkcode => $fw,
1968 interface => $search_context->{'interface'}
1971 $record_processor->process($marcrecord);
1973 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1975 biblionumber => $oldbiblio->{biblionumber},
1976 record => $marcrecord,
1979 ? 'OPACXSLTResultsDisplay'
1980 : 'XSLTResultsDisplay'
1983 hidden_items => \@hiddenitems,
1984 xslt_variables => $xslt_variables
1989 my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} );
1990 $oldbiblio->{biblio_object} = $biblio_object;
1992 my $can_place_holds = 1;
1993 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1994 if (!C4::Context->preference("item-level_itypes")) {
1995 if ($itemtype && $itemtype->{notforloan}) {
1996 $can_place_holds = 0;
1999 $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object;
2001 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2002 $oldbiblio->{items_count} = $items_count;
2003 $oldbiblio->{available_items_loop} = \@available_items_loop;
2004 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2005 $oldbiblio->{other_items_loop} = \@other_items_loop;
2006 $oldbiblio->{availablecount} = $available_count;
2007 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2008 $oldbiblio->{onloancount} = $onloan_count;
2009 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2010 $oldbiblio->{othercount} = $other_count;
2011 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2012 $oldbiblio->{withdrawncount} = $withdrawn_count;
2013 $oldbiblio->{itemlostcount} = $itemlost_count;
2014 $oldbiblio->{damagedcount} = $itemdamaged_count;
2015 $oldbiblio->{intransitcount} = $item_in_transit_count;
2016 $oldbiblio->{onholdcount} = $item_onhold_count;
2017 $oldbiblio->{orderedcount} = $ordered_count;
2018 $oldbiblio->{notforloancount} = $notforloan_count;
2020 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2021 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2022 my $subfields = substr $fieldspec, 3;
2023 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2024 my @alternateholdingsinfo = ();
2025 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2026 my $alternateholdingscount = 0;
2028 for my $field (@holdingsfields) {
2029 my %holding = ( holding => '' );
2030 my $havesubfield = 0;
2031 for my $subfield ($field->subfields()) {
2032 if ((index $subfields, $$subfield[0]) >= 0) {
2033 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2034 $holding{'holding'} .= $$subfield[1];
2038 if ($havesubfield) {
2039 push(@alternateholdingsinfo, \%holding);
2040 $alternateholdingscount++;
2044 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2045 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2048 push( @newresults, $oldbiblio );
2054 =head2 enabled_staff_search_views
2056 %hash = enabled_staff_search_views()
2058 This function returns a hash that contains three flags obtained from the system
2059 preferences, used to determine whether a particular staff search results view
2064 =item C<Output arg:>
2066 * $hash{can_view_MARC} is true only if the MARC view is enabled
2067 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2068 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2070 =item C<usage in the script:>
2074 $template->param ( C4::Search::enabled_staff_search_views );
2078 sub enabled_staff_search_views
2081 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2082 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2083 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2087 =head2 z3950_search_args
2089 $arrayref = z3950_search_args($matchpoints)
2091 This function returns an array reference that contains the search parameters to be
2092 passed to the Z39.50 search script (z3950_search.pl). The array elements
2093 are hash refs whose keys are name and value, and whose values are the
2094 name of a search parameter, the value of that search parameter and the URL encoded
2095 value of that parameter.
2097 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2099 The search parameter values are obtained from the bibliographic record whose
2100 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2102 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2103 a general purpose search argument. In this case, the returned array contains only
2104 entry: the key is 'title' and the value is derived from $matchpoints.
2106 If a search parameter value is undefined or empty, it is not included in the returned
2109 The returned array reference may be passed directly to the template parameters.
2113 =item C<Output arg:>
2115 * $array containing hash refs as described above
2117 =item C<usage in the script:>
2121 $data = Biblio::GetBiblioData($bibno);
2122 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2126 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2130 sub z3950_search_args {
2133 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2134 my $isbn = Business::ISBN->new( $isbn_string );
2136 if (defined $isbn && $isbn->is_valid)
2138 if ( ref($bibrec) ) {
2139 $bibrec->{isbn} = $isbn_string;
2140 $bibrec->{title} = undef;
2142 $bibrec = { isbn => $isbn_string };
2146 $bibrec = { title => $bibrec } if !ref $bibrec;
2149 for my $field (qw/ lccn isbn issn title author dewey subject /)
2151 push @$array, { name => $field, value => $bibrec->{$field} }
2152 if defined $bibrec->{$field};
2157 =head2 GetDistinctValues($field);
2159 C<$field> is a reference to the fields array
2163 sub GetDistinctValues {
2164 my ($fieldname,$string)=@_;
2165 # returns a reference to a hash of references to branches...
2166 if ($fieldname=~/\./){
2167 my ($table,$column)=split /\./, $fieldname;
2168 my $dbh = C4::Context->dbh;
2169 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 ");
2171 my $elements=$sth->fetchall_arrayref({});
2176 my @servers=qw<biblioserver authorityserver>;
2177 my (@zconns,@results);
2178 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2179 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2182 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2185 # The big moment: asynchronously retrieve results from all servers
2191 my ( $i, $size ) = @_;
2192 for ( my $j = 0 ; $j < $size ; $j++ ) {
2194 @hashscan{qw(value cnt)} =
2195 $results[ $i - 1 ]->display_term($j);
2196 push @elements, \%hashscan;
2204 =head2 _ZOOM_event_loop
2206 _ZOOM_event_loop(\@zconns, \@results, sub {
2207 my ( $i, $size ) = @_;
2211 Processes a ZOOM event loop and passes control to a closure for
2212 processing the results, and destroying the resultsets.
2216 sub _ZOOM_event_loop {
2217 my ($zconns, $results, $callback) = @_;
2218 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2219 my $ev = $zconns->[ $i - 1 ]->last_event();
2220 if ( $ev == ZOOM::Event::ZEND ) {
2221 next unless $results->[ $i - 1 ];
2222 my $size = $results->[ $i - 1 ]->size();
2224 $callback->($i, $size);
2229 foreach my $result (@$results) {
2234 =head2 new_record_from_zebra
2236 Given raw data from a searchengine result set, return a MARC::Record object
2238 This helper function is needed to take into account all the involved
2239 system preferences and configuration variables to properly create the
2240 MARC::Record object.
2242 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2243 data. If we are using DOM, then it has to be MARCXML.
2245 If we are using elasticsearch, it'll already be a MARC::Record and this
2246 function needs a new name.
2250 sub new_record_from_zebra {
2253 my $raw_data = shift;
2254 # Set the default indexing modes
2255 my $search_engine = C4::Context->preference("SearchEngine");
2256 if ($search_engine eq 'Elasticsearch') {
2257 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2259 my $index_mode = ( $server eq 'biblioserver' )
2260 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2261 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2263 my $marc_record = eval {
2264 if ( $index_mode eq 'dom' ) {
2265 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2267 MARC::Record->new_from_usmarc( $raw_data );
2274 return $marc_record;
2279 END { } # module clean-up code here (global destructor)
2286 Koha Development Team <http://koha-community.org/>