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 => $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({ record => $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({ record => $marcrecord });
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 ";
393 elsif ( $sort eq "biblionumber_az" || $sort eq "biblionumber_asc" ) {
394 $sort_by .= "1=12 <i ";
396 elsif ( $sort eq "biblionumber_za" || $sort eq "biblionumber_dsc" ) {
397 $sort_by .= "1=12 >i ";
400 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
403 if ( $sort_by && !$scan && $results[$i] ) {
404 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
405 warn "WARNING sort $sort_by failed";
408 } # finished looping through servers
410 # The big moment: asynchronously retrieve results from all servers
415 my ( $i, $size ) = @_;
418 # loop through the results
419 $results_hash->{'hits'} = $size;
421 if ( $offset + $results_per_page <= $size ) {
422 $times = $offset + $results_per_page;
428 for ( my $j = $offset ; $j < $times ; $j++ ) {
431 ## Check if it's an index scan
433 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
435 # here we create a minimal MARC record and hand it off to the
436 # template just like a normal result ... perhaps not ideal, but
438 my $tmprecord = MARC::Record->new();
439 $tmprecord->encoding('UTF-8');
443 # the minimal record in author/title (depending on MARC flavour)
444 if ( C4::Context->preference("marcflavour") eq
447 $tmptitle = MARC::Field->new(
452 $tmprecord->append_fields($tmptitle);
456 MARC::Field->new( '245', ' ', ' ', a => $term, );
458 MARC::Field->new( '100', ' ', ' ', a => $occ, );
459 $tmprecord->append_fields($tmptitle);
460 $tmprecord->append_fields($tmpauthor);
462 $results_hash->{'RECORDS'}[$j] =
463 $tmprecord->as_usmarc();
468 $record = $results[ $i - 1 ]->record($j)->raw();
469 # warn "RECORD $j:".$record;
470 $results_hash->{'RECORDS'}[$j] = $record;
474 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
476 # Fill the facets while we're looping, but only for the
477 # biblioserver and not for a scan
478 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
479 $facets_counter = GetFacets( $results[ $i - 1 ] );
480 $facets_info = _get_facets_info( $facets );
484 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
486 sort { $a cmp $b } keys %$facets_counter
489 my @this_facets_array;
492 $facets_counter->{$link_value}
493 ->{$b} <=> $facets_counter->{$link_value}
495 } keys %{ $facets_counter->{$link_value} }
498 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
499 my $facet_link_value = $one_facet;
500 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
502 # fix the length that will display in the label,
503 my $facet_label_value = $one_facet;
504 my $facet_max_length = C4::Context->preference(
505 'FacetLabelTruncationLength')
508 substr( $one_facet, 0, $facet_max_length )
510 if length($facet_label_value) >
513 # if it's a branch, label by the name, not the code,
514 if ( $link_value =~ /branch/ ) {
515 if ( defined $branches
516 && ref($branches) eq "HASH"
517 && defined $branches->{$one_facet}
518 && ref( $branches->{$one_facet} ) eq
522 $branches->{$one_facet}
526 $facet_label_value = "*";
530 # if it's a itemtype, label by the name, not the code,
531 if ( $link_value =~ /itype/ ) {
532 if ( defined $itemtypes
533 && ref($itemtypes) eq "HASH"
534 && defined $itemtypes->{$one_facet}
535 && ref( $itemtypes->{$one_facet} ) eq
539 $itemtypes->{$one_facet}
540 ->{translated_description};
544 # also, if it's a location code, use the name instead of the code
545 if ( $link_value =~ /location/ ) {
546 # TODO Retrieve all authorised values at once, instead of 1 query per entry
547 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
548 $facet_label_value = $av->count ? $av->next->opac_description : '';
551 # also, if it's a collection code, use the name instead of the code
552 if ( $link_value =~ /ccode/ ) {
553 # TODO Retrieve all authorised values at once, instead of 1 query per entry
554 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
555 $facet_label_value = $av->count ? $av->next->opac_description : '';
558 # but we're down with the whole label being in the link's title.
559 push @this_facets_array,
562 $facets_counter->{$link_value}
564 facet_label_value => $facet_label_value,
565 facet_title_value => $one_facet,
566 facet_link_value => $facet_link_value,
567 type_link_value => $link_value,
569 if ($facet_label_value);
574 type_link_value => $link_value,
575 type_id => $link_value . "_id",
577 . $facets_info->{$link_value}->{'label_value'} =>
579 facets => \@this_facets_array,
583 $facets_info->{$link_value}->{'label_value'} =~
586 and ( Koha::Libraries->search->count == 1 )
593 # This sorts the facets into alphabetical order
595 foreach my $f (@facets_loop) {
596 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
598 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
603 return ( undef, $results_hashref, \@facets_loop );
611 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
613 if ( $use_zebra_facets ) {
614 $facets = _get_facets_from_zebra( $rs );
616 $facets = _get_facets_from_records( $rs );
622 sub _get_facets_from_records {
626 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
627 my $facets_config = getFacets();
629 my $size = $rs->size();
630 my $jmax = $size > $facets_maxrecs
634 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
636 my $marc_record = new_record_from_zebra (
638 $rs->record( $j )->raw()
641 if ( ! defined $marc_record ) {
642 warn "ERROR DECODING RECORD - $@: " .
643 $rs->record( $j )->raw();
647 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
653 =head2 _get_facets_data_from_record
655 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
657 Internal function that extracts facets information from a MARC::Record object
658 and populates $facets_counter for using in getRecords.
660 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
665 sub _get_facets_data_from_record {
667 my ( $marc_record, $facets, $facets_counter ) = @_;
669 for my $facet (@$facets) {
673 foreach my $tag ( @{ $facet->{ tags } } ) {
675 # tag number is the first three digits
676 my $tag_num = substr( $tag, 0, 3 );
677 # subfields are the remainder
678 my $subfield_letters = substr( $tag, 3 );
680 my @fields = $marc_record->field( $tag_num );
681 foreach my $field (@fields) {
682 # If $field->indicator(1) eq 'z', it means it is a 'see from'
683 # field introduced because of IncludeSeeFromInSearches, so skip it
684 next if $field->indicator(1) eq 'z';
686 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
687 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
689 unless ( grep { $_ eq $data } @used_datas ) {
690 push @used_datas, $data;
691 $facets_counter->{ $facet->{ idx } }->{ $data }++;
698 =head2 _get_facets_from_zebra
700 my $facets = _get_facets_from_zebra( $result_set )
702 Retrieves facets for a specified result set. It loops through the facets defined
703 in C4::Koha::getFacets and returns a hash with the following structure:
713 sub _get_facets_from_zebra {
717 # save current elementSetName
718 my $elementSetName = $rs->option( 'elementSetName' );
720 my $facets_loop = getFacets();
721 my $facets_data = {};
722 # loop through defined facets and fill the facets hashref
723 foreach my $facet ( @$facets_loop ) {
725 my $idx = $facet->{ idx };
726 my $sep = $facet->{ sep };
727 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
728 if ( $facet_values ) {
729 # we've actually got a result
730 $facets_data->{ $idx } = $facet_values;
733 # set elementSetName to its previous value to avoid side effects
734 $rs->option( elementSetName => $elementSetName );
739 =head2 _get_facet_from_result_set
742 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
744 Internal function that extracts facet information for a specific index ($facet_idx) and
745 returns a hash containing facet values and count:
748 $facet_value => $count ,
752 Warning: this function has the side effect of changing the elementSetName for the result
753 set. It is a helper function for the main loop, which takes care of backing it up for
758 sub _get_facet_from_result_set {
760 my $facet_idx = shift;
764 my $internal_sep = '<*>';
765 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
767 return if ( ! defined $facet_idx || ! defined $rs );
768 # zebra's facet element, untokenized index
769 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
770 # configure zebra results for retrieving the desired facet
771 $rs->option( elementSetName => $facet_element );
772 # get the facet record from result set
773 my $facet = $rs->record( 0 )->raw;
774 # if the facet has no restuls...
775 return if !defined $facet;
776 # TODO: benchmark DOM vs. SAX performance
777 my $facet_dom = XML::LibXML->load_xml(
780 my @terms = $facet_dom->getElementsByTagName('term');
784 foreach my $term ( @terms ) {
785 my $facet_value = $term->textContent;
786 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
787 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
788 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
794 =head2 _get_facets_info
796 my $facets_info = C4::Search::_get_facets_info( $facets )
798 Internal function that extracts facets information and properly builds
799 the data structure needed to render facet labels.
803 sub _get_facets_info {
807 my $facets_info = {};
809 for my $facet ( @$facets ) {
810 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
817 sub _detect_truncation {
818 my ( $operand, $index ) = @_;
819 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
822 my @wordlist = split( /\s/, $operand );
823 foreach my $word (@wordlist) {
824 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
825 push @rightlefttruncated, $word;
827 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
828 push @lefttruncated, $word;
830 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
831 push @righttruncated, $word;
833 elsif ( index( $word, "*" ) < 0 ) {
834 push @nontruncated, $word;
837 push @regexpr, $word;
841 \@nontruncated, \@righttruncated, \@lefttruncated,
842 \@rightlefttruncated, \@regexpr
847 sub _build_stemmed_operand {
848 my ($operand,$lang) = @_;
849 require Lingua::Stem::Snowball ;
850 my $stemmed_operand=q{};
852 # Stemmer needs language
853 return $operand unless $lang;
855 # If operand contains a digit, it is almost certainly an identifier, and should
856 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
857 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
858 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
859 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
860 return $operand if $operand =~ /\d/;
862 # FIXME: the locale should be set based on the user's language and/or search choice
864 # Make sure we only use the first two letters from the language code
865 $lang = lc(substr($lang, 0, 2));
866 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
867 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
868 if ($lang eq 'nb' || $lang eq 'nn') {
871 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
872 encoding => "UTF-8" );
874 my @words = split( / /, $operand );
875 my @stems = $stemmer->stem(\@words);
876 for my $stem (@stems) {
877 $stemmed_operand .= "$stem";
878 $stemmed_operand .= "?"
879 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
880 $stemmed_operand .= " ";
883 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
884 return $stemmed_operand;
888 sub _build_weighted_query {
890 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
891 # pretty well but could work much better if we had a smarter query parser
892 my ( $operand, $stemmed_operand, $index ) = @_;
893 my $stemming = C4::Context->preference("QueryStemming") || 0;
894 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
895 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
896 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
898 my $weighted_query = "(rk=("; # Specifies that we're applying rank
900 # Keyword, or, no index specified
901 if ( ( $index eq 'kw' ) || ( !$index ) ) {
903 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
904 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
905 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
906 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
907 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
908 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
909 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
910 if $fuzzy_enabled; # add fuzzy, word list
911 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
912 if ( $stemming and $stemmed_operand )
913 ; # add stemming, right truncation
914 $weighted_query .= " or wrdl,r9=\"$operand\"";
916 # embedded sorting: 0 a-z; 1 z-a
917 # $weighted_query .= ") or (sort1,aut=1";
920 # Barcode searches should skip this process
921 elsif ( $index eq 'bc' ) {
922 $weighted_query .= "bc=\"$operand\"";
925 # Authority-number searches should skip this process
926 elsif ( $index eq 'an' ) {
927 $weighted_query .= "an=\"$operand\"";
930 # If the index is numeric, don't autoquote it.
931 elsif ( $index =~ /,st-numeric$/ ) {
932 $weighted_query .= " $index=$operand";
935 # If the index already has more than one qualifier, wrap the operand
936 # in quotes and pass it back (assumption is that the user knows what they
937 # are doing and won't appreciate us mucking up their query
938 elsif ( $index =~ ',' ) {
939 $weighted_query .= " $index=\"$operand\"";
942 #TODO: build better cases based on specific search indexes
944 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
945 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
946 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
947 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
948 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
949 if $fuzzy_enabled; # add fuzzy, word list
950 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
951 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
954 $weighted_query .= "))"; # close rank specification
955 return $weighted_query;
960 Return an array with available indexes.
984 'Author-personal-bibliography',
994 'Chronological-subdivision',
1004 'Conference-name-heading',
1005 'Conference-name-see',
1006 'Conference-name-seealso',
1009 'Control-number-identifier',
1013 'Corporate-name-heading',
1014 'Corporate-name-see',
1015 'Corporate-name-seealso',
1016 'Country-publication',
1019 'date-entered-on-file',
1020 'Date-of-acquisition',
1021 'Date-of-publication',
1022 'Date-time-last-modified',
1023 'Dewey-classification',
1024 'Dissertation-information',
1033 'Geographic-subdivision',
1036 'Heading-use-main-or-added-entry',
1037 'Heading-use-series-added-entry ',
1038 'Heading-use-subject-added-entry',
1042 'Identifier-standard',
1043 'Illustration-code',
1045 'Index-term-uncontrolled',
1046 'Interest-age-level',
1047 'Interest-grade-level',
1057 'language-original',
1067 'Local-classification',
1070 'Match-heading-see-from',
1075 'Multipart-resource-level',
1080 'Name-geographic-heading',
1081 'Name-geographic-see',
1082 'Name-geographic-seealso',
1088 'Other-control-number',
1091 'Personal-name-heading',
1092 'Personal-name-see',
1093 'Personal-name-seealso',
1095 'Place-publication',
1102 'Reading-grade-level',
1103 'Record-control-number',
1114 'Subject-heading-thesaurus',
1115 'Subject-name-personal',
1116 'Subject-subdivision',
1125 'Term-genre-form-heading',
1126 'Term-genre-form-see',
1127 'Term-genre-form-seealso',
1133 'Title-uniform-heading',
1134 'Title-uniform-see',
1135 'Title-uniform-seealso',
1145 'classification-source',
1147 'coded-location-qualifier',
1158 'Local-classification',
1161 'materials-specified',
1166 'Number-local-acquisition',
1171 'replacementpricedate',
1189 $simple_query, $query_cgi,
1190 $query_desc, $limit,
1191 $limit_cgi, $limit_desc,
1192 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1194 Build queries and limits in CCL, CGI, Human,
1195 handle truncation, stemming, field weighting, fuzziness, etc.
1197 See verbose embedded documentation.
1203 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1208 my @operators = $operators ? @$operators : ();
1209 my @indexes = $indexes ? @$indexes : ();
1210 my @operands = $operands ? @$operands : ();
1211 my @limits = $limits ? @$limits : ();
1212 my @sort_by = $sort_by ? @$sort_by : ();
1214 my $stemming = C4::Context->preference("QueryStemming") || 0;
1215 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1216 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1217 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1219 my $query = $operands[0] // "";
1220 my $simple_query = $operands[0];
1222 # initialize the variables we're passing back
1231 my $cclindexes = getIndexes();
1232 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1233 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1235 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1237 $query = "ccl=$query" if $cclq;
1241 my %group_OR_limits;
1242 my $availability_limit;
1243 foreach my $this_limit (@limits) {
1244 next unless $this_limit;
1245 if ( $this_limit =~ /available/ ) {
1247 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1249 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1250 $availability_limit .=
1251 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1252 $limit_cgi .= "&limit=available";
1256 # group_OR_limits, prefixed by mc-
1257 # OR every member of the group
1258 elsif ( $this_limit =~ /mc/ ) {
1259 my ($k,$v) = split(/:/, $this_limit,2);
1260 if ( $k !~ /mc-i(tem)?type/ ) {
1261 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1262 $this_limit =~ tr/"//d;
1263 $this_limit = $k.':"'.$v.'"';
1266 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1267 $limit_desc .= " or " if $group_OR_limits{$k};
1268 $group_OR_limits{$k} .= "$this_limit";
1269 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1270 $limit_desc .= " $this_limit";
1272 elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) {
1273 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1274 $limit .= " and " if $limit || $query;
1275 my $branchfield = C4::Context->preference('SearchLimitLibrary');
1277 if( $this_limit =~ '^multibranchlimit:' ){
1278 my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ );
1279 my $search_group = Koha::Library::Groups->find( $group_id );
1280 @branchcodes = map { $_->branchcode } $search_group->all_libraries;
1281 @branchcodes = sort { $a cmp $b } @branchcodes;
1283 @branchcodes = ( $this_limit =~ /^branch:(.*)$/ );
1287 if ( $branchfield eq "homebranch" ) {
1288 $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes;
1290 elsif ( $branchfield eq "holdingbranch" ) {
1291 $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes;
1294 $this_limit = sprintf "(%s or %s)",
1295 join( " or ", map { 'homebranch: ' . $_ } @branchcodes ),
1296 join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes );
1299 $limit .= "$this_limit";
1300 $limit_desc .= " $this_limit";
1301 } elsif ( $this_limit =~ '^search_filter:' ) {
1302 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1303 my ($filter_id) = ( $this_limit =~ /^search_filter:(.*)$/ );
1304 my $search_filter = Koha::SearchFilters->find( $filter_id );
1305 next unless $search_filter;
1306 my $expanded = $search_filter->expand_filter;
1307 my ( $error, undef, undef, undef, undef, $fixed_limit, undef, undef, undef ) = buildQuery ( undef, undef, undef, $expanded, undef, undef, $lang);
1308 $limit .= " and " if $limit || $query;
1309 $limit .= "$fixed_limit";
1310 $limit_desc .= " $limit";
1313 # Regular old limits
1315 $limit .= " and " if $limit || $query;
1316 $limit .= "$this_limit";
1317 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1318 $limit_desc .= " $this_limit";
1321 foreach my $k (keys (%group_OR_limits)) {
1322 $limit .= " and " if ( $query || $limit );
1323 $limit .= "($group_OR_limits{$k})";
1325 if ($availability_limit) {
1326 $limit .= " and " if ( $query || $limit );
1327 $limit .= "($availability_limit)";
1330 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1332 if ( $query =~ /^ccl=/ ) {
1334 # This is needed otherwise ccl= and &limit won't work together, and
1335 # this happens when selecting a subject on the opac-detail page
1336 my $original_q = $q; # without available part
1337 $q .= $limit if $limit;
1338 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1340 if ( $query =~ /^cql=/ ) {
1341 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1343 if ( $query =~ /^pqf=/ ) {
1345 $query_cgi = "q=pqf=".uri_escape_utf8($');
1346 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1349 # pass nested queries directly
1350 # FIXME: need better handling of some of these variables in this case
1351 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1352 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1353 # if ( $query =~ /(\(|\))/ ) {
1355 # undef, $query, $simple_query, $query_cgi,
1356 # $query, $limit, $limit_cgi, $limit_desc,
1361 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1362 # query operands and indexes and add stemming, truncation, field weighting, etc.
1363 # Once we do so, we'll end up with a value in $query, just like if we had an
1364 # incoming $query from the user
1367 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1368 my $previous_operand
1369 ; # a flag used to keep track if there was a previous query
1370 # if there was, we can apply the current operator
1372 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1374 # COMBINE OPERANDS, INDEXES AND OPERATORS
1375 if ( ($operands[$i] // '') ne '' ) {
1376 $operands[$i]=~s/^\s+//;
1378 # A flag to determine whether or not to add the index to the query
1381 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1382 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1386 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1388 my $operand = $operands[$i];
1389 my $index = $indexes[$i] || 'kw';
1391 # Add index-specific attributes
1393 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1394 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1395 #which is processed higher up in this sub. Other than that, year searches are typically
1396 #handled as limits which are not processed her either.
1398 # Search ranges: Date of Publication, st-numeric
1399 if ( $index =~ /(yr|st-numeric)/ ) {
1400 #weight_fields/relevance search causes errors with date ranges
1401 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1402 #In the case of YYYY-YYYY, it will return no results
1403 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1406 # Date of Acquisition
1407 elsif ( $index =~ /acqdate/ ) {
1408 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1409 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1410 #top of the results just because they have lots of item records matching that date.
1411 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1413 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1415 # ISBN,ISSN,Standard Number, don't need special treatment
1416 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1418 $stemming, $auto_truncation,
1419 $weight_fields, $fuzzy_enabled
1422 if ( $index eq 'nb' ) {
1423 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1424 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1425 $operands[$i] = $operand = '(' . join( ' OR ', map { 'nb=' . $_ } @isbns ) . ')';
1426 $indexes[$i] = $index = 'kw';
1429 if ( $index eq 'ns' ) {
1430 if ( C4::Context->preference("SearchWithISSNVariations") ) {
1431 my @issns = C4::Koha::GetVariationsOfISSN( $operand );
1432 $operands[$i] = $operand = '(' . join( ' OR ', map { 'ns=' . $_ } @issns ) . ')';
1433 $indexes[$i] = $index = 'kw';
1438 # Set default structure attribute (word list)
1439 my $struct_attr = q{};
1440 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1441 $struct_attr = ",wrdl";
1444 # Some helpful index variants
1445 my $index_plus = $index . $struct_attr . ':';
1446 my $index_plus_comma = $index . $struct_attr . ',';
1448 if ($auto_truncation){
1449 unless ( $index =~ /,(st-|phr|ext)/ ) {
1450 #FIXME only valid with LTR scripts
1451 $operand=join(" ",map{
1452 (index($_,"*")>0?"$_":"$_*")
1453 }split (/\s+/,$operand));
1458 my $truncated_operand = q{};
1459 my( $nontruncated, $righttruncated, $lefttruncated,
1460 $rightlefttruncated, $regexpr
1461 ) = _detect_truncation( $operand, $index );
1463 Koha::Logger->get->debug(
1464 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1468 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1469 scalar(@$rightlefttruncated) > 0 )
1472 # Don't field weight or add the index to the query, we do it here
1474 undef $weight_fields;
1475 my $previous_truncation_operand;
1476 if (scalar @$nontruncated) {
1477 $truncated_operand .= "$index_plus @$nontruncated ";
1478 $previous_truncation_operand = 1;
1480 if (scalar @$righttruncated) {
1481 $truncated_operand .= "and " if $previous_truncation_operand;
1482 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1483 $previous_truncation_operand = 1;
1485 if (scalar @$lefttruncated) {
1486 $truncated_operand .= "and " if $previous_truncation_operand;
1487 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1488 $previous_truncation_operand = 1;
1490 if (scalar @$rightlefttruncated) {
1491 $truncated_operand .= "and " if $previous_truncation_operand;
1492 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1493 $previous_truncation_operand = 1;
1496 $operand = $truncated_operand if $truncated_operand;
1497 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1500 my $stemmed_operand = q{};
1501 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1504 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1506 # Handle Field Weighting
1507 my $weighted_operand = q{};
1508 if ($weight_fields) {
1509 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1510 $operand = $weighted_operand;
1514 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1516 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1518 #N.B. Truncation is mutually exclusive with Weighted Queries,
1519 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1520 #the need for this relevance wrapper.
1521 $operand = "(rk=($operand))" unless $weight_fields;
1523 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1525 query_cgi => $query_cgi,
1526 query_desc => $query_desc,
1527 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1528 parsed_operand => $operand,
1529 original_operand => $operands[$i] // '',
1531 index_plus => $index_plus,
1532 indexes_set => $indexes_set,
1533 previous_operand => $previous_operand,
1539 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1542 # Normalize the query and limit strings
1543 # This is flawed , means we can't search anything with : in it
1544 # if user wants to do ccl or cql, start the query with that
1545 # $query =~ s/:/=/g;
1546 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1547 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1548 $query =~ s/(?<=(wrdl)):/=/g;
1549 $query =~ s/(?<=(trn|phr)):/=/g;
1550 $query =~ s/(?<=(st-numeric)):/=/g;
1551 $query =~ s/(?<=(st-year)):/=/g;
1552 $query =~ s/(?<=(st-date-normalized)):/=/g;
1554 # Removing warnings for later substitutions
1556 $query_desc //= q{};
1559 $limit_desc //= q{};
1561 $simple_query //= q{};
1563 for ( $query, $query_desc, $limit, $limit_desc ) {
1564 s/ +/ /g; # remove extra spaces
1565 s/^ //g; # remove any beginning spaces
1566 s/ $//g; # remove any ending spaces
1567 s/==/=/g; # remove double == from query
1569 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1571 for ($query_cgi,$simple_query) {
1574 # append the limit to the query
1575 $query .= " " . $limit;
1577 Koha::Logger->get->debug(
1578 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1579 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1582 undef, $query, $simple_query, $query_cgi,
1583 $query_desc, $limit, $limit_cgi, $limit_desc,
1588 =head2 _build_initial_query
1590 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1592 Build a section of the initial query containing indexes, operators, and operands.
1596 sub _build_initial_query {
1600 if ($params->{previous_operand}){
1601 #If there is a previous operand, add a supplied operator or the default 'and'
1602 $operator = ($params->{operator}) ? ($params->{operator}) : 'AND';
1605 #NOTE: indexes_set is typically set when doing truncation or field weighting
1606 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1608 #e.g. "kw,wrdl:test"
1609 #e.g. " and kw,wrdl:test"
1610 $params->{query} .= " " . $operator . " " . $operand;
1612 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1613 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1614 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if ( $params->{original_operand} ne '' );
1616 #e.g. " and kw,wrdl: test"
1617 $params->{query_desc} .= " " . $operator . " " . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1619 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1621 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1624 =head2 searchResults
1626 my @search_results = searchResults($search_context, $searchdesc, $hits,
1627 $results_per_page, $offset, $scan,
1630 Format results in a form suitable for passing to the template
1634 # IMO this subroutine is pretty messy still -- it's responsible for
1635 # building the HTML output for the template
1637 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1638 my $dbh = C4::Context->dbh;
1643 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1644 my ($is_opac, $hidelostitems);
1645 if ($search_context->{'interface'} eq 'opac') {
1646 $hidelostitems = C4::Context->preference('hidelostitems');
1650 my $record_processor = Koha::RecordProcessor->new({
1651 filters => 'ViewPolicy'
1654 #Build branchnames hash
1655 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' })->as_list;
1657 # FIXME - We build an authorised values hash here, using the default framework
1658 # though it is possible to have different authvals for different fws.
1660 my $shelflocations =
1661 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1663 # get notforloan authorised value list (see $shelflocations FIXME)
1664 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1665 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1668 my $itemtypes = Koha::ItemTypes->search_with_localization;
1669 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1671 #search item field code
1672 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1674 ## find column names of items related to MARC
1675 my %subfieldstosearch;
1676 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1677 for my $column ( @columns ) {
1678 my ( $tagfield, $tagsubfield ) =
1679 &GetMarcFromKohaField( "items." . $column );
1680 if ( defined $tagsubfield ) {
1681 $subfieldstosearch{$column} = $tagsubfield;
1685 # handle which records to actually retrieve
1686 my $times; # Times is which record to process up to
1687 if ( $hits && $offset + $results_per_page <= $hits ) {
1688 $times = $offset + $results_per_page;
1691 $times = $hits; # If less hits than results_per_page+offset we go to the end
1694 my $marcflavour = C4::Context->preference("marcflavour");
1695 # We get the biblionumber position in MARC
1696 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1698 # set stuff for XSLT processing here once, not later again for every record we retrieved
1700 my $userenv = C4::Context->userenv;
1702 = ( defined $userenv and $userenv->{number} )
1703 ? Koha::Patrons->find( $userenv->{number} )
1705 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1707 # loop through all of the records we've retrieved
1708 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1712 # For Scan searches we built USMARC data
1713 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1715 # Normal search, render from Zebra's output
1716 $marcrecord = new_record_from_zebra(
1721 if ( ! defined $marcrecord ) {
1722 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1730 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1731 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1733 SetUTF8Flag($marcrecord);
1734 my $oldbiblio = TransformMarcToKoha({ record => $marcrecord, limit_table => 'no_items' });
1735 $oldbiblio->{result_number} = $i + 1;
1737 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1738 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1739 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1740 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1741 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1743 # edition information, if any
1744 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1746 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1747 # add imageurl to itemtype if there is one
1748 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1749 # Build summary if there is one (the summary is defined in the itemtypes table)
1750 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1752 # Pull out the items fields
1753 my @fields = $marcrecord->field($itemtag);
1754 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1755 my $marcflavor = C4::Context->preference("marcflavour");
1757 # adding linked items that belong to host records
1758 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1759 my $analyticsfield = '773';
1760 if ($marcflavor eq 'MARC21') {
1761 $analyticsfield = '773';
1762 } elsif ($marcflavor eq 'UNIMARC') {
1763 $analyticsfield = '461';
1765 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1766 my $hostbiblionumber = $hostfield->subfield("0");
1767 my $linkeditemnumber = $hostfield->subfield("9");
1768 if( $hostbiblionumber ) {
1769 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1770 if ($linkeditemmarc) {
1771 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1772 if ($linkeditemfield) {
1773 push( @fields, $linkeditemfield );
1780 # Setting item statuses for display
1781 my @available_items_loop;
1782 my @onloan_items_loop;
1783 my @other_items_loop;
1785 my $available_items;
1789 my $ordered_count = 0;
1790 my $available_count = 0;
1791 my $onloan_count = 0;
1792 my $longoverdue_count = 0;
1793 my $other_count = 0;
1794 my $withdrawn_count = 0;
1795 my $itemlost_count = 0;
1796 my $hideatopac_count = 0;
1797 my $itembinding_count = 0;
1798 my $itemdamaged_count = 0;
1799 my $item_in_transit_count = 0;
1800 my $item_onhold_count = 0;
1801 my $notforloan_count = 0;
1802 my $item_recalled_count = 0;
1803 my $items_count = scalar(@fields);
1804 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1805 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1806 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1808 # loop through every item
1809 foreach my $field (@fields) {
1812 # populate the items hash
1813 foreach my $code ( keys %subfieldstosearch ) {
1814 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1817 unless ( $item->{itemnumber} ) {
1818 warn "MARC item without itemnumber retrieved for biblio ($oldbiblio->{biblionumber})";
1822 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1826 # hidden based on OpacHiddenItems syspref or because lost
1827 my $hi = Koha::Items->search( { itemnumber => $item->{itemnumber} } )
1828 ->filter_by_visible_in_opac({ patron => $search_context->{patron} });
1829 unless ( $hi->count ) {
1830 push @hiddenitems, $item->{itemnumber};
1831 $hideatopac_count++;
1836 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1837 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1839 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1840 if ($item->{$hbranch}) {
1841 $item->{'branchname'} = $branches{$item->{$hbranch}};
1843 elsif ($item->{$otherbranch}) { # Last resort
1844 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1848 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1849 . ( $item->{location} ? $item->{location} : q{} )
1850 . ( $item->{itype} ? $item->{itype} : q{} )
1851 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1852 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1853 if ( $item->{onloan}
1855 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1858 my $key = $prefix . $item->{onloan} . $item->{barcode};
1859 $onloan_items->{$key}->{due_date} = $item->{onloan};
1860 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1861 $onloan_items->{$key}->{branchname} = $item->{branchname};
1862 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1863 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1864 $onloan_items->{$key}->{description} = $item->{description};
1865 $onloan_items->{$key}->{imageurl} =
1866 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1868 # if something's checked out and lost, mark it as 'long overdue'
1869 if ( $item->{itemlost} ) {
1870 $onloan_items->{$key}->{longoverdue}++;
1871 $longoverdue_count++;
1875 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1878 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1879 $item->{notforloan} = 1 if !$item->{notforloan} &&
1880 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1883 if ( $item->{notforloan} < 0 ) {
1885 } elsif ( $item->{notforloan} > 0 ) {
1886 $notforloan_count++;
1889 # is item in transit?
1890 my $transfertwhen = '';
1891 my ($transfertfrom, $transfertto);
1893 # is item on the reserve shelf?
1894 my $reservestatus = '';
1896 # is item a waiting recall?
1897 my $recallstatus = '';
1899 unless ($item->{withdrawn}
1900 || $item->{itemlost}
1902 || $item->{notforloan}
1903 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1904 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1906 # A couple heuristics to limit how many times
1907 # we query the database for item transfer information, sacrificing
1908 # accuracy in some cases for speed;
1910 # 1. don't query if item has one of the other statuses
1911 # 2. don't check transit status if the bib has
1912 # more than 20 items
1914 # FIXME: to avoid having the query the database like this, and to make
1915 # the in transit status count as unavailable for search limiting,
1916 # should map transit status to record indexed in Zebra.
1918 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1919 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1920 if ( C4::Context->preference('UseRecalls') ) {
1921 if ( Koha::Recalls->search({ item_id => $item->{itemnumber}, status => 'waiting' })->count ) {
1922 $recallstatus = 'Waiting';
1927 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1928 if ( $item->{withdrawn}
1929 || $item->{itemlost}
1931 || $item->{notforloan}
1932 || $reservestatus eq 'Waiting'
1933 || $recallstatus eq 'Waiting'
1934 || ($transfertwhen && $transfertwhen ne ''))
1936 $withdrawn_count++ if $item->{withdrawn};
1937 $itemlost_count++ if $item->{itemlost};
1938 $itemdamaged_count++ if $item->{damaged};
1939 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1940 $item_onhold_count++ if $reservestatus eq 'Waiting';
1941 $item_recalled_count++ if $recallstatus eq 'Waiting';
1942 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1946 my $key = $prefix . $item->{status};
1947 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1948 $other_items->{$key}->{$_} = $item->{$_};
1950 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1951 $other_items->{$key}->{recalled} = ($recallstatus) ? 1 : 0;
1952 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1953 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1954 $other_items->{$key}->{count}++ if $item->{$hbranch};
1955 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1956 $other_items->{$key}->{description} = $item->{description};
1957 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1962 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1963 foreach (qw(branchname itemcallnumber description)) {
1964 $available_items->{$prefix}->{$_} = $item->{$_};
1966 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1967 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1970 } # notforloan, item level and biblioitem level
1972 # if all items are hidden, do not show the record
1973 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1977 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1978 for my $key ( sort keys %$onloan_items ) {
1979 (++$onloanitemscount > $maxitems) and last;
1980 push @onloan_items_loop, $onloan_items->{$key};
1982 for my $key ( sort keys %$other_items ) {
1983 (++$otheritemscount > $maxitems) and last;
1984 push @other_items_loop, $other_items->{$key};
1986 for my $key ( sort keys %$available_items ) {
1987 (++$availableitemscount > $maxitems) and last;
1988 push @available_items_loop, $available_items->{$key}
1991 # XSLT processing of some stuff
1992 # we fetched the sysprefs already before the loop through all retrieved record!
1994 $record_processor->options({
1995 frameworkcode => $fw,
1996 interface => $search_context->{'interface'}
1999 $record_processor->process($marcrecord);
2001 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
2003 biblionumber => $oldbiblio->{biblionumber},
2004 record => $marcrecord,
2007 ? 'OPACXSLTResultsDisplay'
2008 : 'XSLTResultsDisplay'
2011 hidden_items => \@hiddenitems,
2012 xslt_variables => $xslt_variables
2017 my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2018 $oldbiblio->{biblio_object} = $biblio_object;
2020 my $can_place_holds = 1;
2021 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2022 if (!C4::Context->preference("item-level_itypes")) {
2023 if ($itemtype && $itemtype->{notforloan}) {
2024 $can_place_holds = 0;
2027 $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object;
2029 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2030 $oldbiblio->{items_count} = $items_count;
2031 $oldbiblio->{available_items_loop} = \@available_items_loop;
2032 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2033 $oldbiblio->{other_items_loop} = \@other_items_loop;
2034 $oldbiblio->{availablecount} = $available_count;
2035 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2036 $oldbiblio->{onloancount} = $onloan_count;
2037 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2038 $oldbiblio->{othercount} = $other_count;
2039 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2040 $oldbiblio->{withdrawncount} = $withdrawn_count;
2041 $oldbiblio->{itemlostcount} = $itemlost_count;
2042 $oldbiblio->{damagedcount} = $itemdamaged_count;
2043 $oldbiblio->{intransitcount} = $item_in_transit_count;
2044 $oldbiblio->{onholdcount} = $item_onhold_count;
2045 $oldbiblio->{recalledcount} = $item_recalled_count;
2046 $oldbiblio->{orderedcount} = $ordered_count;
2047 $oldbiblio->{notforloancount} = $notforloan_count;
2049 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2050 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2051 my $subfields = substr $fieldspec, 3;
2052 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2053 my @alternateholdingsinfo = ();
2054 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2055 my $alternateholdingscount = 0;
2057 for my $field (@holdingsfields) {
2058 my %holding = ( holding => '' );
2059 my $havesubfield = 0;
2060 for my $subfield ($field->subfields()) {
2061 if ((index $subfields, $$subfield[0]) >= 0) {
2062 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2063 $holding{'holding'} .= $$subfield[1];
2067 if ($havesubfield) {
2068 push(@alternateholdingsinfo, \%holding);
2069 $alternateholdingscount++;
2073 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2074 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2077 push( @newresults, $oldbiblio );
2083 =head2 enabled_staff_search_views
2085 %hash = enabled_staff_search_views()
2087 This function returns a hash that contains three flags obtained from the system
2088 preferences, used to determine whether a particular staff search results view
2093 =item C<Output arg:>
2095 * $hash{can_view_MARC} is true only if the MARC view is enabled
2096 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2097 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2099 =item C<usage in the script:>
2103 $template->param ( C4::Search::enabled_staff_search_views );
2107 sub enabled_staff_search_views
2110 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2111 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2112 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2116 =head2 z3950_search_args
2118 $arrayref = z3950_search_args($matchpoints)
2120 This function returns an array reference that contains the search parameters to be
2121 passed to the Z39.50 search script (z3950_search.pl). The array elements
2122 are hash refs whose keys are name and value, and whose values are the
2123 name of a search parameter, the value of that search parameter and the URL encoded
2124 value of that parameter.
2126 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2128 The search parameter values are obtained from the bibliographic record whose
2129 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2131 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2132 a general purpose search argument. In this case, the returned array contains only
2133 entry: the key is 'title' and the value is derived from $matchpoints.
2135 If a search parameter value is undefined or empty, it is not included in the returned
2138 The returned array reference may be passed directly to the template parameters.
2142 =item C<Output arg:>
2144 * $array containing hash refs as described above
2146 =item C<usage in the script:>
2150 $data = Biblio::GetBiblioData($bibno);
2151 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2155 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2159 sub z3950_search_args {
2162 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2163 my $isbn = Business::ISBN->new( $isbn_string );
2165 if (defined $isbn && $isbn->is_valid)
2167 if ( ref($bibrec) ) {
2168 $bibrec->{isbn} = $isbn_string;
2169 $bibrec->{title} = undef;
2171 $bibrec = { isbn => $isbn_string };
2175 $bibrec = { title => $bibrec } if !ref $bibrec;
2178 for my $field (qw/ lccn isbn issn title author dewey subject /)
2180 push @$array, { name => $field, value => $bibrec->{$field} }
2181 if defined $bibrec->{$field};
2186 =head2 GetDistinctValues($field);
2188 C<$field> is a reference to the fields array
2192 sub GetDistinctValues {
2193 my ($fieldname,$string)=@_;
2194 # returns a reference to a hash of references to branches...
2195 if ($fieldname=~/\./){
2196 my ($table,$column)=split /\./, $fieldname;
2197 my $dbh = C4::Context->dbh;
2198 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 ");
2200 my $elements=$sth->fetchall_arrayref({});
2205 my @servers=qw<biblioserver authorityserver>;
2206 my (@zconns,@results);
2207 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2208 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2211 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2214 # The big moment: asynchronously retrieve results from all servers
2220 my ( $i, $size ) = @_;
2221 for ( my $j = 0 ; $j < $size ; $j++ ) {
2223 @hashscan{qw(value cnt)} =
2224 $results[ $i - 1 ]->display_term($j);
2225 push @elements, \%hashscan;
2233 =head2 _ZOOM_event_loop
2235 _ZOOM_event_loop(\@zconns, \@results, sub {
2236 my ( $i, $size ) = @_;
2240 Processes a ZOOM event loop and passes control to a closure for
2241 processing the results, and destroying the resultsets.
2245 sub _ZOOM_event_loop {
2246 my ($zconns, $results, $callback) = @_;
2247 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2248 my $ev = $zconns->[ $i - 1 ]->last_event();
2249 if ( $ev == ZOOM::Event::ZEND ) {
2250 next unless $results->[ $i - 1 ];
2251 my $size = $results->[ $i - 1 ]->size();
2253 $callback->($i, $size);
2258 foreach my $result (@$results) {
2263 =head2 new_record_from_zebra
2265 Given raw data from a searchengine result set, return a MARC::Record object
2267 This helper function is needed to take into account all the involved
2268 system preferences and configuration variables to properly create the
2269 MARC::Record object.
2271 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2272 data. If we are using DOM, then it has to be MARCXML.
2274 If we are using elasticsearch, it'll already be a MARC::Record and this
2275 function needs a new name.
2279 sub new_record_from_zebra {
2282 my $raw_data = shift;
2283 # Set the default indexing modes
2284 my $search_engine = C4::Context->preference("SearchEngine");
2285 if ($search_engine eq 'Elasticsearch') {
2286 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2288 my $index_mode = ( $server eq 'biblioserver' )
2289 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2290 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2292 my $marc_record = eval {
2293 if ( $index_mode eq 'dom' ) {
2294 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2296 MARC::Record->new_from_usmarc( $raw_data );
2303 return $marc_record;
2308 END { } # module clean-up code here (global destructor)
2315 Koha Development Team <http://koha-community.org/>