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 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
593 return ( undef, $results_hashref, \@facets_loop );
601 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
603 if ( $use_zebra_facets ) {
604 $facets = _get_facets_from_zebra( $rs );
606 $facets = _get_facets_from_records( $rs );
612 sub _get_facets_from_records {
616 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
617 my $facets_config = getFacets();
619 my $size = $rs->size();
620 my $jmax = $size > $facets_maxrecs
624 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
626 my $marc_record = new_record_from_zebra (
628 $rs->record( $j )->raw()
631 if ( ! defined $marc_record ) {
632 warn "ERROR DECODING RECORD - $@: " .
633 $rs->record( $j )->raw();
637 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
643 =head2 _get_facets_data_from_record
645 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
647 Internal function that extracts facets information from a MARC::Record object
648 and populates $facets_counter for using in getRecords.
650 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
655 sub _get_facets_data_from_record {
657 my ( $marc_record, $facets, $facets_counter ) = @_;
659 for my $facet (@$facets) {
663 foreach my $tag ( @{ $facet->{ tags } } ) {
665 # tag number is the first three digits
666 my $tag_num = substr( $tag, 0, 3 );
667 # subfields are the remainder
668 my $subfield_letters = substr( $tag, 3 );
670 my @fields = $marc_record->field( $tag_num );
671 foreach my $field (@fields) {
672 # If $field->indicator(1) eq 'z', it means it is a 'see from'
673 # field introduced because of IncludeSeeFromInSearches, so skip it
674 next if $field->indicator(1) eq 'z';
676 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
677 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
679 unless ( grep { $_ eq $data } @used_datas ) {
680 push @used_datas, $data;
681 $facets_counter->{ $facet->{ idx } }->{ $data }++;
688 =head2 _get_facets_from_zebra
690 my $facets = _get_facets_from_zebra( $result_set )
692 Retrieves facets for a specified result set. It loops through the facets defined
693 in C4::Koha::getFacets and returns a hash with the following structure:
703 sub _get_facets_from_zebra {
707 # save current elementSetName
708 my $elementSetName = $rs->option( 'elementSetName' );
710 my $facets_loop = getFacets();
711 my $facets_data = {};
712 # loop through defined facets and fill the facets hashref
713 foreach my $facet ( @$facets_loop ) {
715 my $idx = $facet->{ idx };
716 my $sep = $facet->{ sep };
717 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
718 if ( $facet_values ) {
719 # we've actually got a result
720 $facets_data->{ $idx } = $facet_values;
723 # set elementSetName to its previous value to avoid side effects
724 $rs->option( elementSetName => $elementSetName );
729 =head2 _get_facet_from_result_set
732 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
734 Internal function that extracts facet information for a specific index ($facet_idx) and
735 returns a hash containing facet values and count:
738 $facet_value => $count ,
742 Warning: this function has the side effect of changing the elementSetName for the result
743 set. It is a helper function for the main loop, which takes care of backing it up for
748 sub _get_facet_from_result_set {
750 my $facet_idx = shift;
754 my $internal_sep = '<*>';
755 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
757 return if ( ! defined $facet_idx || ! defined $rs );
758 # zebra's facet element, untokenized index
759 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
760 # configure zebra results for retrieving the desired facet
761 $rs->option( elementSetName => $facet_element );
762 # get the facet record from result set
763 my $facet = $rs->record( 0 )->raw;
764 # if the facet has no restuls...
765 return if !defined $facet;
766 # TODO: benchmark DOM vs. SAX performance
767 my $facet_dom = XML::LibXML->load_xml(
770 my @terms = $facet_dom->getElementsByTagName('term');
774 foreach my $term ( @terms ) {
775 my $facet_value = $term->textContent;
776 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
777 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
778 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
784 =head2 _get_facets_info
786 my $facets_info = C4::Search::_get_facets_info( $facets )
788 Internal function that extracts facets information and properly builds
789 the data structure needed to render facet labels.
793 sub _get_facets_info {
797 my $facets_info = {};
799 for my $facet ( @$facets ) {
800 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
807 sub _detect_truncation {
808 my ( $operand, $index ) = @_;
809 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
812 my @wordlist = split( /\s/, $operand );
813 foreach my $word (@wordlist) {
814 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
815 push @rightlefttruncated, $word;
817 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
818 push @lefttruncated, $word;
820 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
821 push @righttruncated, $word;
823 elsif ( index( $word, "*" ) < 0 ) {
824 push @nontruncated, $word;
827 push @regexpr, $word;
831 \@nontruncated, \@righttruncated, \@lefttruncated,
832 \@rightlefttruncated, \@regexpr
837 sub _build_stemmed_operand {
838 my ($operand,$lang) = @_;
839 require Lingua::Stem::Snowball ;
840 my $stemmed_operand=q{};
842 # Stemmer needs language
843 return $operand unless $lang;
845 # If operand contains a digit, it is almost certainly an identifier, and should
846 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
847 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
848 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
849 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
850 return $operand if $operand =~ /\d/;
852 # FIXME: the locale should be set based on the user's language and/or search choice
854 # Make sure we only use the first two letters from the language code
855 $lang = lc(substr($lang, 0, 2));
856 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
857 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
858 if ($lang eq 'nb' || $lang eq 'nn') {
861 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
862 encoding => "UTF-8" );
864 my @words = split( / /, $operand );
865 my @stems = $stemmer->stem(\@words);
866 for my $stem (@stems) {
867 $stemmed_operand .= "$stem";
868 $stemmed_operand .= "?"
869 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
870 $stemmed_operand .= " ";
873 Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand");
874 return $stemmed_operand;
878 sub _build_weighted_query {
880 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
881 # pretty well but could work much better if we had a smarter query parser
882 my ( $operand, $stemmed_operand, $index ) = @_;
883 my $stemming = C4::Context->preference("QueryStemming") || 0;
884 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
885 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
886 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
888 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
890 # Keyword, or, no index specified
891 if ( ( $index eq 'kw' ) || ( !$index ) ) {
893 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
894 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
895 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
896 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
897 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
898 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
899 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
900 if $fuzzy_enabled; # add fuzzy, word list
901 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
902 if ( $stemming and $stemmed_operand )
903 ; # add stemming, right truncation
904 $weighted_query .= " or wrdl,r9=\"$operand\"";
906 # embedded sorting: 0 a-z; 1 z-a
907 # $weighted_query .= ") or (sort1,aut=1";
910 # Barcode searches should skip this process
911 elsif ( $index eq 'bc' ) {
912 $weighted_query .= "bc=\"$operand\"";
915 # Authority-number searches should skip this process
916 elsif ( $index eq 'an' ) {
917 $weighted_query .= "an=\"$operand\"";
920 # If the index is numeric, don't autoquote it.
921 elsif ( $index =~ /,st-numeric$/ ) {
922 $weighted_query .= " $index=$operand";
925 # If the index already has more than one qualifier, wrap the operand
926 # in quotes and pass it back (assumption is that the user knows what they
927 # are doing and won't appreciate us mucking up their query
928 elsif ( $index =~ ',' ) {
929 $weighted_query .= " $index=\"$operand\"";
932 #TODO: build better cases based on specific search indexes
934 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
935 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
936 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
937 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
938 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
939 if $fuzzy_enabled; # add fuzzy, word list
940 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
941 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
944 $weighted_query .= "))"; # close rank specification
945 return $weighted_query;
950 Return an array with available indexes.
974 'Author-personal-bibliography',
984 'Chronological-subdivision',
994 'Conference-name-heading',
995 'Conference-name-see',
996 'Conference-name-seealso',
1001 'Corporate-name-heading',
1002 'Corporate-name-see',
1003 'Corporate-name-seealso',
1004 'Country-publication',
1007 'date-entered-on-file',
1008 'Date-of-acquisition',
1009 'Date-of-publication',
1010 'Date-time-last-modified',
1011 'Dewey-classification',
1012 'Dissertation-information',
1021 'Geographic-subdivision',
1024 'Heading-use-main-or-added-entry',
1025 'Heading-use-series-added-entry ',
1026 'Heading-use-subject-added-entry',
1030 'Identifier-standard',
1031 'Illustration-code',
1033 'Index-term-uncontrolled',
1034 'Interest-age-level',
1035 'Interest-grade-level',
1045 'language-original',
1055 'Local-classification',
1058 'Match-heading-see-from',
1066 'Name-geographic-heading',
1067 'Name-geographic-see',
1068 'Name-geographic-seealso',
1074 'Other-control-number',
1077 'Personal-name-heading',
1078 'Personal-name-see',
1079 'Personal-name-seealso',
1081 'Place-publication',
1088 'Reading-grade-level',
1089 'Record-control-number',
1100 'Subject-heading-thesaurus',
1101 'Subject-name-personal',
1102 'Subject-subdivision',
1111 'Term-genre-form-heading',
1112 'Term-genre-form-see',
1113 'Term-genre-form-seealso',
1119 'Title-uniform-heading',
1120 'Title-uniform-see',
1121 'Title-uniform-seealso',
1131 'classification-source',
1133 'coded-location-qualifier',
1144 'Local-classification',
1147 'materials-specified',
1152 'Number-local-acquisition',
1157 'replacementpricedate',
1175 $simple_query, $query_cgi,
1176 $query_desc, $limit,
1177 $limit_cgi, $limit_desc,
1178 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1180 Build queries and limits in CCL, CGI, Human,
1181 handle truncation, stemming, field weighting, fuzziness, etc.
1183 See verbose embedded documentation.
1189 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1194 my @operators = $operators ? @$operators : ();
1195 my @indexes = $indexes ? @$indexes : ();
1196 my @operands = $operands ? @$operands : ();
1197 my @limits = $limits ? @$limits : ();
1198 my @sort_by = $sort_by ? @$sort_by : ();
1200 my $stemming = C4::Context->preference("QueryStemming") || 0;
1201 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1202 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1203 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1205 my $query = $operands[0] // "";
1206 my $simple_query = $operands[0];
1208 # initialize the variables we're passing back
1217 my $cclindexes = getIndexes();
1218 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1219 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1221 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1223 $query = "ccl=$query" if $cclq;
1226 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1228 if ( $query =~ /^ccl=/ ) {
1230 # This is needed otherwise ccl= and &limit won't work together, and
1231 # this happens when selecting a subject on the opac-detail page
1232 @limits = grep {!/^$/} @limits;
1233 my $original_q = $q; # without available part
1234 unless ( grep { $_ eq 'available' } @limits ) {
1235 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1239 if ( grep { $_ eq 'available' } @limits ) {
1240 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1241 @limits = grep {!/^available$/} @limits;
1243 $q .= ' and '.join(' and ', @limits) if @limits;
1245 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1247 if ( $query =~ /^cql=/ ) {
1248 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1250 if ( $query =~ /^pqf=/ ) {
1252 $query_cgi = "q=pqf=".uri_escape_utf8($');
1253 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1256 # pass nested queries directly
1257 # FIXME: need better handling of some of these variables in this case
1258 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1259 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1260 # if ( $query =~ /(\(|\))/ ) {
1262 # undef, $query, $simple_query, $query_cgi,
1263 # $query, $limit, $limit_cgi, $limit_desc,
1268 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1269 # query operands and indexes and add stemming, truncation, field weighting, etc.
1270 # Once we do so, we'll end up with a value in $query, just like if we had an
1271 # incoming $query from the user
1274 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1275 my $previous_operand
1276 ; # a flag used to keep track if there was a previous query
1277 # if there was, we can apply the current operator
1279 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1281 # COMBINE OPERANDS, INDEXES AND OPERATORS
1282 if ( ($operands[$i] // '') ne '' ) {
1283 $operands[$i]=~s/^\s+//;
1285 # A flag to determine whether or not to add the index to the query
1288 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1289 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1293 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1295 my $operand = $operands[$i];
1296 my $index = $indexes[$i] || 'kw';
1298 # Add index-specific attributes
1300 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1301 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1302 #which is processed higher up in this sub. Other than that, year searches are typically
1303 #handled as limits which are not processed her either.
1305 # Search ranges: Date of Publication, st-numeric
1306 if ( $index =~ /(yr|st-numeric)/ ) {
1307 #weight_fields/relevance search causes errors with date ranges
1308 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1309 #In the case of YYYY-YYYY, it will return no results
1310 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1313 # Date of Acquisition
1314 elsif ( $index =~ /acqdate/ ) {
1315 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1316 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1317 #top of the results just because they have lots of item records matching that date.
1318 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1320 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1322 # ISBN,ISSN,Standard Number, don't need special treatment
1323 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1325 $stemming, $auto_truncation,
1326 $weight_fields, $fuzzy_enabled
1329 if ( $index eq 'nb' ) {
1330 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1331 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1332 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1333 $indexes[$i] = $index = 'kw';
1338 # Set default structure attribute (word list)
1339 my $struct_attr = q{};
1340 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1341 $struct_attr = ",wrdl";
1344 # Some helpful index variants
1345 my $index_plus = $index . $struct_attr . ':';
1346 my $index_plus_comma = $index . $struct_attr . ',';
1348 if ($auto_truncation){
1349 unless ( $index =~ /,(st-|phr|ext)/ ) {
1350 #FIXME only valid with LTR scripts
1351 $operand=join(" ",map{
1352 (index($_,"*")>0?"$_":"$_*")
1353 }split (/\s+/,$operand));
1358 my $truncated_operand = q{};
1359 my( $nontruncated, $righttruncated, $lefttruncated,
1360 $rightlefttruncated, $regexpr
1361 ) = _detect_truncation( $operand, $index );
1363 Koha::Logger->get->debug(
1364 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<");
1368 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1369 scalar(@$rightlefttruncated) > 0 )
1372 # Don't field weight or add the index to the query, we do it here
1374 undef $weight_fields;
1375 my $previous_truncation_operand;
1376 if (scalar @$nontruncated) {
1377 $truncated_operand .= "$index_plus @$nontruncated ";
1378 $previous_truncation_operand = 1;
1380 if (scalar @$righttruncated) {
1381 $truncated_operand .= "and " if $previous_truncation_operand;
1382 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1383 $previous_truncation_operand = 1;
1385 if (scalar @$lefttruncated) {
1386 $truncated_operand .= "and " if $previous_truncation_operand;
1387 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1388 $previous_truncation_operand = 1;
1390 if (scalar @$rightlefttruncated) {
1391 $truncated_operand .= "and " if $previous_truncation_operand;
1392 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1393 $previous_truncation_operand = 1;
1396 $operand = $truncated_operand if $truncated_operand;
1397 Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<");
1400 my $stemmed_operand = q{};
1401 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1404 Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<");
1406 # Handle Field Weighting
1407 my $weighted_operand = q{};
1408 if ($weight_fields) {
1409 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1410 $operand = $weighted_operand;
1414 Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<");
1416 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1418 #N.B. Truncation is mutually exclusive with Weighted Queries,
1419 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1420 #the need for this relevance wrapper.
1421 $operand = "(rk=($operand))" unless $weight_fields;
1423 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1425 query_cgi => $query_cgi,
1426 query_desc => $query_desc,
1427 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1428 parsed_operand => $operand,
1429 original_operand => $operands[$i] // '',
1431 index_plus => $index_plus,
1432 indexes_set => $indexes_set,
1433 previous_operand => $previous_operand,
1439 Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<");
1442 my %group_OR_limits;
1443 my $availability_limit;
1444 foreach my $this_limit (@limits) {
1445 next unless $this_limit;
1446 if ( $this_limit =~ /available/ ) {
1448 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1450 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1451 $availability_limit .=
1452 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1453 $limit_cgi .= "&limit=available";
1457 # group_OR_limits, prefixed by mc-
1458 # OR every member of the group
1459 elsif ( $this_limit =~ /mc/ ) {
1460 my ($k,$v) = split(/:/, $this_limit,2);
1461 if ( $k !~ /mc-i(tem)?type/ ) {
1462 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1463 $this_limit =~ tr/"//d;
1464 $this_limit = $k.':"'.$v.'"';
1467 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1468 $limit_desc .= " or " if $group_OR_limits{$k};
1469 $group_OR_limits{$k} .= "$this_limit";
1470 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1471 $limit_desc .= " $this_limit";
1474 # Regular old limits
1476 $limit .= " and " if $limit || $query;
1477 $limit .= "$this_limit";
1478 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1479 if ($this_limit =~ /^branch:(.+)/) {
1480 my $branchcode = $1;
1481 my $library = Koha::Libraries->find( $branchcode );
1482 if (defined $library) {
1483 $limit_desc .= " branch:" . $library->branchname;
1485 $limit_desc .= " $this_limit";
1488 $limit_desc .= " $this_limit";
1492 foreach my $k (keys (%group_OR_limits)) {
1493 $limit .= " and " if ( $query || $limit );
1494 $limit .= "($group_OR_limits{$k})";
1496 if ($availability_limit) {
1497 $limit .= " and " if ( $query || $limit );
1498 $limit .= "($availability_limit)";
1501 # Normalize the query and limit strings
1502 # This is flawed , means we can't search anything with : in it
1503 # if user wants to do ccl or cql, start the query with that
1504 # $query =~ s/:/=/g;
1505 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1506 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1507 $query =~ s/(?<=(wrdl)):/=/g;
1508 $query =~ s/(?<=(trn|phr)):/=/g;
1509 $query =~ s/(?<=(st-numeric)):/=/g;
1510 $query =~ s/(?<=(st-year)):/=/g;
1511 $query =~ s/(?<=(st-date-normalized)):/=/g;
1513 # Removing warnings for later substitutions
1515 $query_desc //= q{};
1518 $limit_desc //= q{};
1520 $simple_query //= q{};
1522 for ( $query, $query_desc, $limit, $limit_desc ) {
1523 s/ +/ /g; # remove extra spaces
1524 s/^ //g; # remove any beginning spaces
1525 s/ $//g; # remove any ending spaces
1526 s/==/=/g; # remove double == from query
1528 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1530 for ($query_cgi,$simple_query) {
1533 # append the limit to the query
1534 $query .= " " . $limit;
1536 Koha::Logger->get->debug(
1537 sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s",
1538 $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc );
1541 undef, $query, $simple_query, $query_cgi,
1542 $query_desc, $limit, $limit_cgi, $limit_desc,
1547 =head2 _build_initial_query
1549 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1551 Build a section of the initial query containing indexes, operators, and operands.
1555 sub _build_initial_query {
1559 if ($params->{previous_operand}){
1560 #If there is a previous operand, add a supplied operator or the default 'and'
1561 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1564 #NOTE: indexes_set is typically set when doing truncation or field weighting
1565 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1567 #e.g. "kw,wrdl:test"
1568 #e.g. " and kw,wrdl:test"
1569 $params->{query} .= $operator . $operand;
1571 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1572 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1573 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1575 #e.g. " and kw,wrdl: test"
1576 $params->{query_desc} .= $operator . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1578 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1580 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1583 =head2 searchResults
1585 my @search_results = searchResults($search_context, $searchdesc, $hits,
1586 $results_per_page, $offset, $scan,
1589 Format results in a form suitable for passing to the template
1593 # IMO this subroutine is pretty messy still -- it's responsible for
1594 # building the HTML output for the template
1596 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1597 my $dbh = C4::Context->dbh;
1602 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1603 my ($is_opac, $hidelostitems);
1604 if ($search_context->{'interface'} eq 'opac') {
1605 $hidelostitems = C4::Context->preference('hidelostitems');
1609 my $record_processor = Koha::RecordProcessor->new({
1610 filters => 'ViewPolicy'
1613 #Build branchnames hash
1614 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1616 # FIXME - We build an authorised values hash here, using the default framework
1617 # though it is possible to have different authvals for different fws.
1619 my $shelflocations =
1620 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1622 # get notforloan authorised value list (see $shelflocations FIXME)
1623 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1624 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1627 my $itemtypes = Koha::ItemTypes->search_with_localization;
1628 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1630 #search item field code
1631 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1633 ## find column names of items related to MARC
1634 my %subfieldstosearch;
1635 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1636 for my $column ( @columns ) {
1637 my ( $tagfield, $tagsubfield ) =
1638 &GetMarcFromKohaField( "items." . $column );
1639 if ( defined $tagsubfield ) {
1640 $subfieldstosearch{$column} = $tagsubfield;
1644 # handle which records to actually retrieve
1645 my $times; # Times is which record to process up to
1646 if ( $hits && $offset + $results_per_page <= $hits ) {
1647 $times = $offset + $results_per_page;
1650 $times = $hits; # If less hits than results_per_page+offset we go to the end
1653 my $marcflavour = C4::Context->preference("marcflavour");
1654 # We get the biblionumber position in MARC
1655 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1657 # set stuff for XSLT processing here once, not later again for every record we retrieved
1661 $xslsyspref = "OPACXSLTResultsDisplay";
1662 $xslfile = C4::Context->preference( $xslsyspref );
1664 $xslsyspref = "XSLTResultsDisplay";
1665 $xslfile = C4::Context->preference( $xslsyspref ) || "default";
1667 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1668 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1670 my $userenv = C4::Context->userenv;
1672 = ( defined $userenv and $userenv->{number} )
1673 ? Koha::Patrons->find( $userenv->{number} )
1675 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1677 # loop through all of the records we've retrieved
1678 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1682 # For Scan searches we built USMARC data
1683 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1685 # Normal search, render from Zebra's output
1686 $marcrecord = new_record_from_zebra(
1691 if ( ! defined $marcrecord ) {
1692 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1700 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1701 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1703 SetUTF8Flag($marcrecord);
1704 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw, 'no_items' );
1705 $oldbiblio->{result_number} = $i + 1;
1707 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1708 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1709 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1710 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1711 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1713 # edition information, if any
1714 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1716 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1717 # add imageurl to itemtype if there is one
1718 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1719 # Build summary if there is one (the summary is defined in the itemtypes table)
1720 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1722 # FIXME: this is only used in the deprecated non-XLST opac results
1723 if ( !$xslfile && $is_opac && $itemtype && $itemtype->{summary} ) {
1724 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1725 my @fields = $marcrecord->fields();
1728 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1730 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1731 $tag =~ /(.{3})(.)/;
1732 if($marcrecord->field($1)){
1733 my @abc = $marcrecord->field($1)->subfield($2);
1734 $tags->{$tag} = $#abc + 1 ;
1738 # We catch how many times to repeat this line
1740 foreach my $tag (keys(%$tags)){
1741 $max = $tags->{$tag} if($tags->{$tag} > $max);
1744 # we replace, and repeat each line
1745 for (my $i = 0 ; $i < $max ; $i++){
1746 my $newline = $line;
1748 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1749 $tag =~ /(.{3})(.)/;
1751 if($marcrecord->field($1)){
1752 my @repl = $marcrecord->field($1)->subfield($2);
1753 my $subfieldvalue = $repl[$i];
1754 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1757 $newsummary .= "$newline\n";
1761 $newsummary =~ s/\[(.*?)]//g;
1762 $newsummary =~ s/\n/<br\/>/g;
1763 $oldbiblio->{summary} = $newsummary;
1766 # Pull out the items fields
1767 my @fields = $marcrecord->field($itemtag);
1768 my $marcflavor = C4::Context->preference("marcflavour");
1770 # adding linked items that belong to host records
1771 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1772 my $analyticsfield = '773';
1773 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1774 $analyticsfield = '773';
1775 } elsif ($marcflavor eq 'UNIMARC') {
1776 $analyticsfield = '461';
1778 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1779 my $hostbiblionumber = $hostfield->subfield("0");
1780 my $linkeditemnumber = $hostfield->subfield("9");
1781 if( $hostbiblionumber ) {
1782 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1783 if ($linkeditemmarc) {
1784 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1785 if ($linkeditemfield) {
1786 push( @fields, $linkeditemfield );
1793 # Setting item statuses for display
1794 my @available_items_loop;
1795 my @onloan_items_loop;
1796 my @other_items_loop;
1798 my $available_items;
1802 my $ordered_count = 0;
1803 my $available_count = 0;
1804 my $onloan_count = 0;
1805 my $longoverdue_count = 0;
1806 my $other_count = 0;
1807 my $withdrawn_count = 0;
1808 my $itemlost_count = 0;
1809 my $hideatopac_count = 0;
1810 my $itembinding_count = 0;
1811 my $itemdamaged_count = 0;
1812 my $item_in_transit_count = 0;
1813 my $can_place_holds = 0;
1814 my $item_onhold_count = 0;
1815 my $notforloan_count = 0;
1816 my $items_count = scalar(@fields);
1817 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1818 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1819 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1821 # loop through every item
1822 foreach my $field (@fields) {
1825 # populate the items hash
1826 foreach my $code ( keys %subfieldstosearch ) {
1827 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1829 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1833 # hidden because lost
1834 if ($hidelostitems && $item->{itemlost}) {
1835 $hideatopac_count++;
1838 # hidden based on OpacHiddenItems syspref
1839 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
1841 push @hiddenitems, @hi;
1842 $hideatopac_count++;
1847 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1848 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1850 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1851 if ($item->{$hbranch}) {
1852 $item->{'branchname'} = $branches{$item->{$hbranch}};
1854 elsif ($item->{$otherbranch}) { # Last resort
1855 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1859 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1860 . ( $item->{location} ? $item->{location} : q{} )
1861 . ( $item->{itype} ? $item->{itype} : q{} )
1862 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1863 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1864 if ( $item->{onloan}
1866 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1869 my $key = $prefix . $item->{onloan} . $item->{barcode};
1870 $onloan_items->{$key}->{due_date} = $item->{onloan};
1871 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1872 $onloan_items->{$key}->{branchname} = $item->{branchname};
1873 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1874 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1875 $onloan_items->{$key}->{description} = $item->{description};
1876 $onloan_items->{$key}->{imageurl} =
1877 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1879 # if something's checked out and lost, mark it as 'long overdue'
1880 if ( $item->{itemlost} ) {
1881 $onloan_items->{$key}->{longoverdue}++;
1882 $longoverdue_count++;
1884 else { # can place holds as long as item isn't lost
1885 $can_place_holds = 1;
1889 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1892 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1893 $item->{notforloan} = 1 if !$item->{notforloan} &&
1894 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1897 if ( $item->{notforloan} < 0 ) {
1899 } elsif ( $item->{notforloan} > 0 ) {
1900 $notforloan_count++;
1903 # is item in transit?
1904 my $transfertwhen = '';
1905 my ($transfertfrom, $transfertto);
1907 # is item on the reserve shelf?
1908 my $reservestatus = '';
1910 unless ($item->{withdrawn}
1911 || $item->{itemlost}
1913 || $item->{notforloan}
1914 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1915 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1917 # A couple heuristics to limit how many times
1918 # we query the database for item transfer information, sacrificing
1919 # accuracy in some cases for speed;
1921 # 1. don't query if item has one of the other statuses
1922 # 2. don't check transit status if the bib has
1923 # more than 20 items
1925 # FIXME: to avoid having the query the database like this, and to make
1926 # the in transit status count as unavailable for search limiting,
1927 # should map transit status to record indexed in Zebra.
1929 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1930 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1933 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1934 if ( $item->{withdrawn}
1935 || $item->{itemlost}
1937 || $item->{notforloan}
1938 || $reservestatus eq 'Waiting'
1939 || ($transfertwhen && $transfertwhen ne ''))
1941 $withdrawn_count++ if $item->{withdrawn};
1942 $itemlost_count++ if $item->{itemlost};
1943 $itemdamaged_count++ if $item->{damaged};
1944 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1945 $item_onhold_count++ if $reservestatus eq 'Waiting';
1946 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1948 # can place a hold on a item if
1949 # not lost nor withdrawn
1950 # not damaged unless AllowHoldsOnDamagedItems is true
1951 # item is either for loan or on order (notforloan < 0)
1952 $can_place_holds = 1
1955 && !$item->{withdrawn}
1956 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
1957 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
1962 my $key = $prefix . $item->{status};
1963 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1964 $other_items->{$key}->{$_} = $item->{$_};
1966 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1967 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1968 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1969 $other_items->{$key}->{count}++ if $item->{$hbranch};
1970 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1971 $other_items->{$key}->{description} = $item->{description};
1972 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1976 $can_place_holds = 1;
1978 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1979 foreach (qw(branchname itemcallnumber description)) {
1980 $available_items->{$prefix}->{$_} = $item->{$_};
1982 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1983 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1986 } # notforloan, item level and biblioitem level
1988 # if all items are hidden, do not show the record
1989 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
1993 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1994 for my $key ( sort keys %$onloan_items ) {
1995 (++$onloanitemscount > $maxitems) and last;
1996 push @onloan_items_loop, $onloan_items->{$key};
1998 for my $key ( sort keys %$other_items ) {
1999 (++$otheritemscount > $maxitems) and last;
2000 push @other_items_loop, $other_items->{$key};
2002 for my $key ( sort keys %$available_items ) {
2003 (++$availableitemscount > $maxitems) and last;
2004 push @available_items_loop, $available_items->{$key}
2007 # XSLT processing of some stuff
2008 # we fetched the sysprefs already before the loop through all retrieved record!
2009 if (!$scan && $xslfile) {
2010 $record_processor->options({
2011 frameworkcode => $fw,
2012 interface => $search_context->{'interface'}
2015 $record_processor->process($marcrecord);
2016 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang, $xslt_variables);
2019 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2020 if (!C4::Context->preference("item-level_itypes")) {
2021 if ($itemtype && $itemtype->{notforloan}) {
2022 $can_place_holds = 0;
2025 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2026 $oldbiblio->{items_count} = $items_count;
2027 $oldbiblio->{available_items_loop} = \@available_items_loop;
2028 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2029 $oldbiblio->{other_items_loop} = \@other_items_loop;
2030 $oldbiblio->{availablecount} = $available_count;
2031 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2032 $oldbiblio->{onloancount} = $onloan_count;
2033 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2034 $oldbiblio->{othercount} = $other_count;
2035 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2036 $oldbiblio->{withdrawncount} = $withdrawn_count;
2037 $oldbiblio->{itemlostcount} = $itemlost_count;
2038 $oldbiblio->{damagedcount} = $itemdamaged_count;
2039 $oldbiblio->{intransitcount} = $item_in_transit_count;
2040 $oldbiblio->{onholdcount} = $item_onhold_count;
2041 $oldbiblio->{orderedcount} = $ordered_count;
2042 $oldbiblio->{notforloancount} = $notforloan_count;
2044 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2045 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2046 my $subfields = substr $fieldspec, 3;
2047 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2048 my @alternateholdingsinfo = ();
2049 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2050 my $alternateholdingscount = 0;
2052 for my $field (@holdingsfields) {
2053 my %holding = ( holding => '' );
2054 my $havesubfield = 0;
2055 for my $subfield ($field->subfields()) {
2056 if ((index $subfields, $$subfield[0]) >= 0) {
2057 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2058 $holding{'holding'} .= $$subfield[1];
2062 if ($havesubfield) {
2063 push(@alternateholdingsinfo, \%holding);
2064 $alternateholdingscount++;
2068 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2069 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2072 $oldbiblio->{biblio_object} = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2074 push( @newresults, $oldbiblio );
2080 =head2 enabled_staff_search_views
2082 %hash = enabled_staff_search_views()
2084 This function returns a hash that contains three flags obtained from the system
2085 preferences, used to determine whether a particular staff search results view
2090 =item C<Output arg:>
2092 * $hash{can_view_MARC} is true only if the MARC view is enabled
2093 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2094 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2096 =item C<usage in the script:>
2100 $template->param ( C4::Search::enabled_staff_search_views );
2104 sub enabled_staff_search_views
2107 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2108 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2109 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2113 =head2 z3950_search_args
2115 $arrayref = z3950_search_args($matchpoints)
2117 This function returns an array reference that contains the search parameters to be
2118 passed to the Z39.50 search script (z3950_search.pl). The array elements
2119 are hash refs whose keys are name and value, and whose values are the
2120 name of a search parameter, the value of that search parameter and the URL encoded
2121 value of that parameter.
2123 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2125 The search parameter values are obtained from the bibliographic record whose
2126 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2128 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2129 a general purpose search argument. In this case, the returned array contains only
2130 entry: the key is 'title' and the value is derived from $matchpoints.
2132 If a search parameter value is undefined or empty, it is not included in the returned
2135 The returned array reference may be passed directly to the template parameters.
2139 =item C<Output arg:>
2141 * $array containing hash refs as described above
2143 =item C<usage in the script:>
2147 $data = Biblio::GetBiblioData($bibno);
2148 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2152 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2156 sub z3950_search_args {
2159 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2160 my $isbn = Business::ISBN->new( $isbn_string );
2162 if (defined $isbn && $isbn->is_valid)
2164 if ( ref($bibrec) ) {
2165 $bibrec->{isbn} = $isbn_string;
2166 $bibrec->{title} = undef;
2168 $bibrec = { isbn => $isbn_string };
2172 $bibrec = { title => $bibrec } if !ref $bibrec;
2175 for my $field (qw/ lccn isbn issn title author dewey subject /)
2177 push @$array, { name => $field, value => $bibrec->{$field} }
2178 if defined $bibrec->{$field};
2183 =head2 GetDistinctValues($field);
2185 C<$field> is a reference to the fields array
2189 sub GetDistinctValues {
2190 my ($fieldname,$string)=@_;
2191 # returns a reference to a hash of references to branches...
2192 if ($fieldname=~/\./){
2193 my ($table,$column)=split /\./, $fieldname;
2194 my $dbh = C4::Context->dbh;
2195 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 ");
2197 my $elements=$sth->fetchall_arrayref({});
2202 my @servers=qw<biblioserver authorityserver>;
2203 my (@zconns,@results);
2204 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2205 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2208 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2211 # The big moment: asynchronously retrieve results from all servers
2217 my ( $i, $size ) = @_;
2218 for ( my $j = 0 ; $j < $size ; $j++ ) {
2220 @hashscan{qw(value cnt)} =
2221 $results[ $i - 1 ]->display_term($j);
2222 push @elements, \%hashscan;
2230 =head2 _ZOOM_event_loop
2232 _ZOOM_event_loop(\@zconns, \@results, sub {
2233 my ( $i, $size ) = @_;
2237 Processes a ZOOM event loop and passes control to a closure for
2238 processing the results, and destroying the resultsets.
2242 sub _ZOOM_event_loop {
2243 my ($zconns, $results, $callback) = @_;
2244 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2245 my $ev = $zconns->[ $i - 1 ]->last_event();
2246 if ( $ev == ZOOM::Event::ZEND ) {
2247 next unless $results->[ $i - 1 ];
2248 my $size = $results->[ $i - 1 ]->size();
2250 $callback->($i, $size);
2255 foreach my $result (@$results) {
2260 =head2 new_record_from_zebra
2262 Given raw data from a searchengine result set, return a MARC::Record object
2264 This helper function is needed to take into account all the involved
2265 system preferences and configuration variables to properly create the
2266 MARC::Record object.
2268 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2269 data. If we are using DOM, then it has to be MARCXML.
2271 If we are using elasticsearch, it'll already be a MARC::Record and this
2272 function needs a new name.
2276 sub new_record_from_zebra {
2279 my $raw_data = shift;
2280 # Set the default indexing modes
2281 my $search_engine = C4::Context->preference("SearchEngine");
2282 if ($search_engine eq 'Elasticsearch') {
2283 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2285 my $index_mode = ( $server eq 'biblioserver' )
2286 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2287 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2289 my $marc_record = eval {
2290 if ( $index_mode eq 'dom' ) {
2291 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2293 MARC::Record->new_from_usmarc( $raw_data );
2300 return $marc_record;
2305 END { } # module clean-up code here (global destructor)
2312 Koha Development Team <http://koha-community.org/>