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>.
19 #use warnings; FIXME - Bug 2505
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
27 use C4::Search::PazPar2;
30 use C4::Reserves; # GetReserveStatus
33 use Koha::AuthorisedValues;
41 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
44 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
49 C4::Search - Functions for searching the Koha catalog.
53 See opac/opac-search.pl or catalogue/search.pl for example of usage
57 This module provides searching functions for Koha's bibliographic databases
71 &enabled_staff_search_views
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, '' );
92 my ( $biblionumber, $title );
94 # search duplicate on ISBN, easy and fast..
96 if ( $result->{isbn} ) {
97 $result->{isbn} =~ s/\(.*$//;
98 $result->{isbn} =~ s/\s+$//;
99 $query = "isbn:$result->{isbn}";
103 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
109 $titleindex = 'title|exact';
110 $authorindex = 'author|exact';
112 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
114 $titleindex = 'ti,ext';
115 $authorindex = 'au,ext';
119 $result->{title} =~ s /\\//g;
120 $result->{title} =~ s /\"//g;
121 $result->{title} =~ s /\(//g;
122 $result->{title} =~ s /\)//g;
124 # FIXME: instead of removing operators, could just do
125 # quotes around the value
126 $result->{title} =~ s/(and|or|not)//g;
127 $query = "$titleindex:\"$result->{title}\"";
128 if ( $result->{author} ) {
129 $result->{author} =~ s /\\//g;
130 $result->{author} =~ s /\"//g;
131 $result->{author} =~ s /\(//g;
132 $result->{author} =~ s /\)//g;
134 # remove valid operators
135 $result->{author} =~ s/(and|or|not)//g;
136 $query .= " $op $authorindex:\"$result->{author}\"";
140 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
142 if (!defined $error) {
143 foreach my $possible_duplicate_record (@{$searchresults}) {
144 my $marcrecord = new_record_from_zebra(
146 $possible_duplicate_record
149 my $result = TransformMarcToKoha( $marcrecord, '' );
151 # FIXME :: why 2 $biblionumber ?
153 push @results, $result->{'biblionumber'};
154 push @results, $result->{'title'};
163 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
165 This function provides a simple search API on the bibliographic catalog
171 * $query can be a simple keyword or a complete CCL query
172 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
173 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
174 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
175 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
180 Returns an array consisting of three elements
181 * $error is undefined unless an error is detected
182 * $results is a reference to an array of records.
183 * $total_hits is the number of hits that would have been returned with no limit
185 If an error is returned the two other return elements are undefined. If error itself is undefined
186 the other two elements are always defined
188 =item C<usage in the script:>
192 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
194 if (defined $error) {
195 $template->param(query_error => $error);
196 warn "error: ".$error;
197 output_html_with_http_headers $input, $cookie, $template->output;
201 my $hits = @{$marcresults};
204 for my $r ( @{$marcresults} ) {
205 my $marcrecord = MARC::File::USMARC::decode($r);
206 my $biblio = TransformMarcToKoha($marcrecord,q{});
208 #build the iarray of hashs for the template.
210 title => $biblio->{'title'},
211 subtitle => $biblio->{'subtitle'},
212 biblionumber => $biblio->{'biblionumber'},
213 author => $biblio->{'author'},
214 publishercode => $biblio->{'publishercode'},
215 publicationyear => $biblio->{'publicationyear'},
220 $template->param(result=>\@results);
225 my ( $query, $offset, $max_results, $servers, %options ) = @_;
227 return ( 'No query entered', undef, undef ) unless $query;
228 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
229 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
237 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
239 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
242 # Initialize & Search Zebra
243 for ( my $i = 0 ; $i < @servers ; $i++ ) {
245 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
247 $query =~ s/=/:/g unless $options{skip_normalize};
248 $QParser->parse( $query );
249 $query = $QParser->target_syntax($servers[$i]);
250 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
252 $query =~ s/:/=/g unless $options{skip_normalize};
253 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
255 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
259 $zconns[$i]->errmsg() . " ("
260 . $zconns[$i]->errcode() . ") "
261 . $zconns[$i]->addinfo() . " "
262 . $zconns[$i]->diagset();
264 return ( $error, undef, undef ) if $zconns[$i]->errcode();
268 # caught a ZOOM::Exception
272 . $@->addinfo() . " "
274 warn $error." for query: $query";
275 return ( $error, undef, undef );
284 my $first_record = defined($offset) ? $offset + 1 : 1;
285 my $hits = $tmpresults[ $i - 1 ]->size();
286 $total_hits += $hits;
287 my $last_record = $hits;
288 if ( defined $max_results && $offset + $max_results < $hits ) {
289 $last_record = $offset + $max_results;
292 for my $j ( $first_record .. $last_record ) {
294 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
297 push @{$results}, $record if defined $record;
302 foreach my $zoom_query (@zoom_queries) {
303 $zoom_query->destroy();
306 return ( undef, $results, $total_hits );
311 ( undef, $results_hashref, \@facets_loop ) = getRecords (
313 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
314 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
318 The all singing, all dancing, multi-server, asynchronous, scanning,
319 searching, record nabbing, facet-building
321 See verbse embedded documentation.
327 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
328 $results_per_page, $offset, $expanded_facet, $branches,
329 $itemtypes, $query_type, $scan, $opac
332 my @servers = @$servers_ref;
333 my @sort_by = @$sort_by_ref;
335 # Initialize variables for the ZOOM connection and results object
339 my $results_hashref = ();
341 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
342 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
344 # Initialize variables for the faceted results objects
345 my $facets_counter = {};
346 my $facets_info = {};
347 my $facets = getFacets();
349 my @facets_loop; # stores the ref to array of hashes for template facets loop
351 ### LOOP THROUGH THE SERVERS
352 for ( my $i = 0 ; $i < @servers ; $i++ ) {
353 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
355 # perform the search, create the results objects
356 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
357 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
359 #$query_to_use = $simple_query if $scan;
360 warn $simple_query if ( $scan and $DEBUG );
362 # Check if we've got a query_type defined, if so, use it
365 if ($query_type =~ /^ccl/) {
366 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
367 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
368 } elsif ($query_type =~ /^cql/) {
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
370 } elsif ($query_type =~ /^pqf/) {
371 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
373 warn "Unknown query_type '$query_type'. Results undetermined.";
376 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
378 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
382 warn "WARNING: query problem with $query_to_use " . $@;
385 # Concatenate the sort_by limits and pass them to the results object
386 # Note: sort will override rank
388 foreach my $sort (@sort_by) {
389 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
390 $sort_by .= "1=1003 <i ";
392 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
393 $sort_by .= "1=1003 >i ";
395 elsif ( $sort eq "popularity_asc" ) {
396 $sort_by .= "1=9003 <i ";
398 elsif ( $sort eq "popularity_dsc" ) {
399 $sort_by .= "1=9003 >i ";
401 elsif ( $sort eq "call_number_asc" ) {
402 $sort_by .= "1=8007 <i ";
404 elsif ( $sort eq "call_number_dsc" ) {
405 $sort_by .= "1=8007 >i ";
407 elsif ( $sort eq "pubdate_asc" ) {
408 $sort_by .= "1=31 <i ";
410 elsif ( $sort eq "pubdate_dsc" ) {
411 $sort_by .= "1=31 >i ";
413 elsif ( $sort eq "acqdate_asc" ) {
414 $sort_by .= "1=32 <i ";
416 elsif ( $sort eq "acqdate_dsc" ) {
417 $sort_by .= "1=32 >i ";
419 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
420 $sort_by .= "1=4 <i ";
422 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
423 $sort_by .= "1=4 >i ";
426 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
429 if ( $sort_by && !$scan && $results[$i] ) {
430 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
431 warn "WARNING sort $sort_by failed";
434 } # finished looping through servers
436 # The big moment: asynchronously retrieve results from all servers
441 my ( $i, $size ) = @_;
444 # loop through the results
445 $results_hash->{'hits'} = $size;
447 if ( $offset + $results_per_page <= $size ) {
448 $times = $offset + $results_per_page;
454 for ( my $j = $offset ; $j < $times ; $j++ ) {
458 ## Check if it's an index scan
460 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
462 # here we create a minimal MARC record and hand it off to the
463 # template just like a normal result ... perhaps not ideal, but
465 my $tmprecord = MARC::Record->new();
466 $tmprecord->encoding('UTF-8');
470 # the minimal record in author/title (depending on MARC flavour)
471 if ( C4::Context->preference("marcflavour") eq
474 $tmptitle = MARC::Field->new(
479 $tmprecord->append_fields($tmptitle);
483 MARC::Field->new( '245', ' ', ' ', a => $term, );
485 MARC::Field->new( '100', ' ', ' ', a => $occ, );
486 $tmprecord->append_fields($tmptitle);
487 $tmprecord->append_fields($tmpauthor);
489 $results_hash->{'RECORDS'}[$j] =
490 $tmprecord->as_usmarc();
495 $record = $results[ $i - 1 ]->record($j)->raw();
496 # warn "RECORD $j:".$record;
497 $results_hash->{'RECORDS'}[$j] = $record;
501 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
503 # Fill the facets while we're looping, but only for the
504 # biblioserver and not for a scan
505 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
506 $facets_counter = GetFacets( $results[ $i - 1 ] );
507 $facets_info = _get_facets_info( $facets );
511 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
513 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
514 keys %$facets_counter
518 my $number_of_facets;
519 my @this_facets_array;
522 $facets_counter->{$link_value}
523 ->{$b} <=> $facets_counter->{$link_value}
525 } keys %{ $facets_counter->{$link_value} }
529 if ( ( $number_of_facets <= 5 )
530 || ( $expanded_facet eq $link_value )
531 || ( $facets_info->{$link_value}->{'expanded'} )
535 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
536 my $facet_link_value = $one_facet;
537 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
539 # fix the length that will display in the label,
540 my $facet_label_value = $one_facet;
541 my $facet_max_length = C4::Context->preference(
542 'FacetLabelTruncationLength')
545 substr( $one_facet, 0, $facet_max_length )
547 if length($facet_label_value) >
550 # if it's a branch, label by the name, not the code,
551 if ( $link_value =~ /branch/ ) {
552 if ( defined $branches
553 && ref($branches) eq "HASH"
554 && defined $branches->{$one_facet}
555 && ref( $branches->{$one_facet} ) eq
559 $branches->{$one_facet}
563 $facet_label_value = "*";
567 # if it's a itemtype, label by the name, not the code,
568 if ( $link_value =~ /itype/ ) {
569 if ( defined $itemtypes
570 && ref($itemtypes) eq "HASH"
571 && defined $itemtypes->{$one_facet}
572 && ref( $itemtypes->{$one_facet} ) eq
576 $itemtypes->{$one_facet}
577 ->{translated_description};
581 # also, if it's a location code, use the name instead of the code
582 if ( $link_value =~ /location/ ) {
583 # TODO Retrieve all authorised values at once, instead of 1 query per entry
584 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
585 $facet_label_value = $av->count ? $av->next->opac_description : '';
588 # but we're down with the whole label being in the link's title.
589 push @this_facets_array,
592 $facets_counter->{$link_value}
594 facet_label_value => $facet_label_value,
595 facet_title_value => $one_facet,
596 facet_link_value => $facet_link_value,
597 type_link_value => $link_value,
599 if ($facet_label_value);
603 # handle expanded option
604 unless ( $facets_info->{$link_value}->{'expanded'} ) {
606 if ( ( $number_of_facets > 5 )
607 && ( $expanded_facet ne $link_value ) );
611 type_link_value => $link_value,
612 type_id => $link_value . "_id",
614 . $facets_info->{$link_value}->{'label_value'} =>
616 facets => \@this_facets_array,
617 expandable => $expandable,
618 expand => $link_value,
622 $facets_info->{$link_value}->{'label_value'} =~
625 and ( Koha::Libraries->search->count == 1 )
632 # This sorts the facets into alphabetical order
634 foreach my $f (@facets_loop) {
635 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
637 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
640 return ( undef, $results_hashref, \@facets_loop );
648 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
649 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
651 if ( $indexing_mode eq 'dom' &&
652 $use_zebra_facets ) {
653 $facets = _get_facets_from_zebra( $rs );
655 $facets = _get_facets_from_records( $rs );
661 sub _get_facets_from_records {
665 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
666 my $facets_config = getFacets();
668 my $size = $rs->size();
669 my $jmax = $size > $facets_maxrecs
673 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
675 my $marc_record = new_record_from_zebra (
677 $rs->record( $j )->raw()
680 if ( ! defined $marc_record ) {
681 warn "ERROR DECODING RECORD - $@: " .
682 $rs->record( $j )->raw();
686 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
692 =head2 _get_facets_data_from_record
694 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
696 Internal function that extracts facets information from a MARC::Record object
697 and populates $facets_counter for using in getRecords.
699 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
704 sub _get_facets_data_from_record {
706 my ( $marc_record, $facets, $facets_counter ) = @_;
708 for my $facet (@$facets) {
712 foreach my $tag ( @{ $facet->{ tags } } ) {
714 # tag number is the first three digits
715 my $tag_num = substr( $tag, 0, 3 );
716 # subfields are the remainder
717 my $subfield_letters = substr( $tag, 3 );
719 my @fields = $marc_record->field( $tag_num );
720 foreach my $field (@fields) {
721 # If $field->indicator(1) eq 'z', it means it is a 'see from'
722 # field introduced because of IncludeSeeFromInSearches, so skip it
723 next if $field->indicator(1) eq 'z';
725 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
727 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
728 push @used_datas, $data;
729 $facets_counter->{ $facet->{ idx } }->{ $data }++;
736 =head2 _get_facets_from_zebra
738 my $facets = _get_facets_from_zebra( $result_set )
740 Retrieves facets for a specified result set. It loops through the facets defined
741 in C4::Koha::getFacets and returns a hash with the following structure:
751 sub _get_facets_from_zebra {
755 # save current elementSetName
756 my $elementSetName = $rs->option( 'elementSetName' );
758 my $facets_loop = getFacets();
759 my $facets_data = {};
760 # loop through defined facets and fill the facets hashref
761 foreach my $facet ( @$facets_loop ) {
763 my $idx = $facet->{ idx };
764 my $sep = $facet->{ sep };
765 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
766 if ( $facet_values ) {
767 # we've actually got a result
768 $facets_data->{ $idx } = $facet_values;
771 # set elementSetName to its previous value to avoid side effects
772 $rs->option( elementSetName => $elementSetName );
777 =head2 _get_facet_from_result_set
780 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
782 Internal function that extracts facet information for a specific index ($facet_idx) and
783 returns a hash containing facet values and count:
786 $facet_value => $count ,
790 Warning: this function has the side effect of changing the elementSetName for the result
791 set. It is a helper function for the main loop, which takes care of backing it up for
796 sub _get_facet_from_result_set {
798 my $facet_idx = shift;
802 my $internal_sep = '<*>';
803 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
805 return if ( ! defined $facet_idx || ! defined $rs );
806 # zebra's facet element, untokenized index
807 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
808 # configure zebra results for retrieving the desired facet
809 $rs->option( elementSetName => $facet_element );
810 # get the facet record from result set
811 my $facet = $rs->record( 0 )->raw;
812 # if the facet has no restuls...
813 return if !defined $facet;
814 # TODO: benchmark DOM vs. SAX performance
815 my $facet_dom = XML::LibXML->load_xml(
818 my @terms = $facet_dom->getElementsByTagName('term');
822 foreach my $term ( @terms ) {
823 my $facet_value = $term->textContent;
824 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
825 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
831 =head2 _get_facets_info
833 my $facets_info = C4::Search::_get_facets_info( $facets )
835 Internal function that extracts facets information and properly builds
836 the data structure needed to render facet labels.
840 sub _get_facets_info {
844 my $facets_info = {};
846 for my $facet ( @$facets ) {
847 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
848 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
856 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
857 $results_per_page, $offset, $expanded_facet, $branches,
860 #NOTE: Parameter $branches is not used here !
862 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
864 $paz->search($simple_query);
865 sleep 1; # FIXME: WHY?
868 my $results_hashref = {};
869 my $stats = XMLin($paz->stat);
870 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
872 # for a grouped search result, the number of hits
873 # is the number of groups returned; 'bib_hits' will have
874 # the total number of bibs.
875 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
876 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
878 HIT: foreach my $hit (@{ $results->{'hit'} }) {
879 my $recid = $hit->{recid}->[0];
881 my $work_title = $hit->{'md-work-title'}->[0];
883 if (exists $hit->{'md-work-author'}) {
884 $work_author = $hit->{'md-work-author'}->[0];
886 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
888 my $result_group = {};
889 $result_group->{'group_label'} = $group_label;
890 $result_group->{'group_merge_key'} = $recid;
893 if (exists $hit->{count}) {
894 $count = $hit->{count}->[0];
896 $result_group->{'group_count'} = $count;
898 for (my $i = 0; $i < $count; $i++) {
899 # FIXME -- may need to worry about diacritics here
900 my $rec = $paz->record($recid, $i);
901 push @{ $result_group->{'RECORDS'} }, $rec;
904 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
907 # pass through facets
908 my $termlist_xml = $paz->termlist('author,subject');
909 my $terms = XMLin($termlist_xml, forcearray => 1);
910 my @facets_loop = ();
911 #die Dumper($results);
912 # foreach my $list (sort keys %{ $terms->{'list'} }) {
914 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
916 # facet_label_value => $facet->{'name'}->[0],
919 # push @facets_loop, ( {
920 # type_label => $list,
921 # facets => \@facets,
925 return ( undef, $results_hashref, \@facets_loop );
929 sub _detect_truncation {
930 my ( $operand, $index ) = @_;
931 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
934 my @wordlist = split( /\s/, $operand );
935 foreach my $word (@wordlist) {
936 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
937 push @rightlefttruncated, $word;
939 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
940 push @lefttruncated, $word;
942 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
943 push @righttruncated, $word;
945 elsif ( index( $word, "*" ) < 0 ) {
946 push @nontruncated, $word;
949 push @regexpr, $word;
953 \@nontruncated, \@righttruncated, \@lefttruncated,
954 \@rightlefttruncated, \@regexpr
959 sub _build_stemmed_operand {
960 my ($operand,$lang) = @_;
961 require Lingua::Stem::Snowball ;
962 my $stemmed_operand=q{};
964 # If operand contains a digit, it is almost certainly an identifier, and should
965 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
966 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
967 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
968 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
969 return $operand if $operand =~ /\d/;
971 # FIXME: the locale should be set based on the user's language and/or search choice
973 # Make sure we only use the first two letters from the language code
974 $lang = lc(substr($lang, 0, 2));
975 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
976 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
977 if ($lang eq 'nb' || $lang eq 'nn') {
980 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
981 encoding => "UTF-8" );
983 my @words = split( / /, $operand );
984 my @stems = $stemmer->stem(\@words);
985 for my $stem (@stems) {
986 $stemmed_operand .= "$stem";
987 $stemmed_operand .= "?"
988 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
989 $stemmed_operand .= " ";
991 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
992 return $stemmed_operand;
996 sub _build_weighted_query {
998 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
999 # pretty well but could work much better if we had a smarter query parser
1000 my ( $operand, $stemmed_operand, $index ) = @_;
1001 my $stemming = C4::Context->preference("QueryStemming") || 0;
1002 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1003 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1004 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1006 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1008 # Keyword, or, no index specified
1009 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1011 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1012 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1013 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1014 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1015 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1016 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1017 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1018 if $fuzzy_enabled; # add fuzzy, word list
1019 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1020 if ( $stemming and $stemmed_operand )
1021 ; # add stemming, right truncation
1022 $weighted_query .= " or wrdl,r9=\"$operand\"";
1024 # embedded sorting: 0 a-z; 1 z-a
1025 # $weighted_query .= ") or (sort1,aut=1";
1028 # Barcode searches should skip this process
1029 elsif ( $index eq 'bc' ) {
1030 $weighted_query .= "bc=\"$operand\"";
1033 # Authority-number searches should skip this process
1034 elsif ( $index eq 'an' ) {
1035 $weighted_query .= "an=\"$operand\"";
1038 # If the index is numeric, don't autoquote it.
1039 elsif ( $index =~ /,st-numeric$/ ) {
1040 $weighted_query .= " $index=$operand";
1043 # If the index already has more than one qualifier, wrap the operand
1044 # in quotes and pass it back (assumption is that the user knows what they
1045 # are doing and won't appreciate us mucking up their query
1046 elsif ( $index =~ ',' ) {
1047 $weighted_query .= " $index=\"$operand\"";
1050 #TODO: build better cases based on specific search indexes
1052 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1053 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1054 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1055 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1056 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1057 if $fuzzy_enabled; # add fuzzy, word list
1058 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1059 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1062 $weighted_query .= "))"; # close rank specification
1063 return $weighted_query;
1068 Return an array with available indexes.
1090 'Author-personal-bibliography',
1100 'Chronological-subdivision',
1110 'Conference-name-heading',
1111 'Conference-name-see',
1112 'Conference-name-seealso',
1117 'Corporate-name-heading',
1118 'Corporate-name-see',
1119 'Corporate-name-seealso',
1120 'Country-publication',
1123 'date-entered-on-file',
1124 'Date-of-acquisition',
1125 'Date-of-publication',
1126 'Date-time-last-modified',
1127 'Dewey-classification',
1128 'Dissertation-information',
1137 'Geographic-subdivision',
1140 'Heading-use-main-or-added-entry',
1141 'Heading-use-series-added-entry ',
1142 'Heading-use-subject-added-entry',
1146 'Identifier-standard',
1147 'Illustration-code',
1149 'Index-term-uncontrolled',
1150 'Interest-age-level',
1151 'Interest-grade-level',
1161 'language-original',
1171 'Local-classification',
1174 'Match-heading-see-from',
1182 'Name-geographic-heading',
1183 'Name-geographic-see',
1184 'Name-geographic-seealso',
1190 'Other-control-number',
1193 'Personal-name-heading',
1194 'Personal-name-see',
1195 'Personal-name-seealso',
1197 'Place-publication',
1204 'Reading-grade-level',
1205 'Record-control-number',
1216 'Subject-heading-thesaurus',
1217 'Subject-name-personal',
1218 'Subject-subdivision',
1227 'Term-genre-form-heading',
1228 'Term-genre-form-see',
1229 'Term-genre-form-seealso',
1235 'Title-uniform-heading',
1236 'Title-uniform-see',
1237 'Title-uniform-seealso',
1247 'classification-source',
1249 'coded-location-qualifier',
1260 'Local-classification',
1263 'materials-specified',
1268 'Number-local-acquisition',
1273 'replacementpricedate',
1288 =head2 _handle_exploding_index
1290 my $query = _handle_exploding_index($index, $term)
1292 Callback routine to generate the search for "exploding" indexes (i.e.
1293 those indexes which are turned into multiple or-connected searches based
1298 sub _handle_exploding_index {
1299 my ($QParser, $filter, $params, $negate, $server) = @_;
1300 my $index = $filter;
1301 my $term = join(' ', @$params);
1303 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1305 my $marcflavour = C4::Context->preference('marcflavour');
1307 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1308 my $wantedcodes = '';
1309 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1310 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1311 foreach my $auth (@$results) {
1312 my $record = MARC::Record->new_from_usmarc($auth);
1313 my @references = $record->field('5..');
1315 if ($index eq 'su-br') {
1317 } elsif ($index eq 'su-na') {
1319 } elsif ($index eq 'su-rl') {
1322 foreach my $reference (@references) {
1323 my $codes = $reference->subfield($codesubfield);
1324 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1328 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1334 ( $operators, $operands, $indexes, $limits,
1335 $sort_by, $scan, $lang ) =
1336 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1338 Shim function to ease the transition from buildQuery to a new QueryParser.
1339 This function is called at the beginning of buildQuery, and modifies
1340 buildQuery's input. If it can handle the input, it returns a query that
1341 buildQuery will not try to parse.
1346 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1348 my @operators = $operators ? @$operators : ();
1349 my @indexes = $indexes ? @$indexes : ();
1350 my @operands = $operands ? @$operands : ();
1351 my @limits = $limits ? @$limits : ();
1352 my @sort_by = $sort_by ? @$sort_by : ();
1354 my $query = $operands[0];
1360 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1361 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1362 undef $QParser if (scalar @limits > 0);
1366 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1368 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1369 next unless $operands[$ii];
1370 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1372 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1373 $query .= $operands[$ii];
1375 elsif ( $indexes[$ii] =~ m/su-/ ) {
1376 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1380 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1383 foreach my $limit (@limits) {
1385 if ( scalar(@sort_by) > 0 ) {
1387 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1388 $query =~ s/$modifier_re//g;
1389 foreach my $modifier (@sort_by) {
1390 $query .= " #$modifier";
1394 $query_desc = $query;
1395 $query_desc =~ s/\s+/ /g;
1396 if ( C4::Context->preference("QueryWeightFields") ) {
1398 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1399 { 'target_syntax_callback' => \&_handle_exploding_index } );
1400 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1401 { 'target_syntax_callback' => \&_handle_exploding_index } );
1402 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1403 { 'target_syntax_callback' => \&_handle_exploding_index } );
1404 $QParser->parse($query);
1405 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1408 require Koha::QueryParser::Driver::PQF;
1409 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1410 s/$modifier_re//g for @operands;
1413 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1419 $simple_query, $query_cgi,
1420 $query_desc, $limit,
1421 $limit_cgi, $limit_desc,
1422 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1424 Build queries and limits in CCL, CGI, Human,
1425 handle truncation, stemming, field weighting, fuzziness, etc.
1427 See verbose embedded documentation.
1433 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1435 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1438 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1441 my @operators = $operators ? @$operators : ();
1442 my @indexes = $indexes ? @$indexes : ();
1443 my @operands = $operands ? @$operands : ();
1444 my @limits = $limits ? @$limits : ();
1445 my @sort_by = $sort_by ? @$sort_by : ();
1447 my $stemming = C4::Context->preference("QueryStemming") || 0;
1448 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1449 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1450 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1452 my $query = $operands[0];
1453 my $simple_query = $operands[0];
1455 # initialize the variables we're passing back
1464 my $cclindexes = getIndexes();
1465 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1466 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1468 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1470 $query = "ccl=$query" if $cclq;
1473 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1475 if ( $query =~ /^ccl=/ ) {
1477 # This is needed otherwise ccl= and &limit won't work together, and
1478 # this happens when selecting a subject on the opac-detail page
1479 @limits = grep {!/^$/} @limits;
1480 my $original_q = $q; # without available part
1481 unless ( grep { /^available$/ } @limits ) {
1482 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1486 if ( grep { /^available$/ } @limits ) {
1487 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1488 delete $limits['available'];
1490 $q .= ' and '.join(' and ', @limits) if @limits;
1492 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1494 if ( $query =~ /^cql=/ ) {
1495 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1497 if ( $query =~ /^pqf=/ ) {
1499 $query_cgi = "q=".uri_escape_utf8($query_desc);
1502 $query_cgi = "q=pqf=".uri_escape_utf8($');
1504 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1507 # pass nested queries directly
1508 # FIXME: need better handling of some of these variables in this case
1509 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1510 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1511 # if ( $query =~ /(\(|\))/ ) {
1513 # undef, $query, $simple_query, $query_cgi,
1514 # $query, $limit, $limit_cgi, $limit_desc,
1519 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1520 # query operands and indexes and add stemming, truncation, field weighting, etc.
1521 # Once we do so, we'll end up with a value in $query, just like if we had an
1522 # incoming $query from the user
1525 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1526 my $previous_operand
1527 ; # a flag used to keep track if there was a previous query
1528 # if there was, we can apply the current operator
1530 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1532 # COMBINE OPERANDS, INDEXES AND OPERATORS
1533 if ( $operands[$i] ) {
1534 $operands[$i]=~s/^\s+//;
1536 # A flag to determine whether or not to add the index to the query
1539 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1540 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1544 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1546 my $operand = $operands[$i];
1547 my $index = $indexes[$i];
1549 # Add index-specific attributes
1551 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1552 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1553 #which is processed higher up in this sub. Other than that, year searches are typically
1554 #handled as limits which are not processed her either.
1556 # Search ranges: Date of Publication, st-numeric
1557 if ( $index =~ /(yr|st-numeric)/ ) {
1558 #weight_fields/relevance search causes errors with date ranges
1559 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1560 #In the case of YYYY-YYYY, it will return no results
1561 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1564 # Date of Acquisition
1565 elsif ( $index =~ /acqdate/ ) {
1566 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1567 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1568 #top of the results just because they have lots of item records matching that date.
1569 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1571 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1573 # ISBN,ISSN,Standard Number, don't need special treatment
1574 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1576 $stemming, $auto_truncation,
1577 $weight_fields, $fuzzy_enabled
1580 if ( $index eq 'nb' ) {
1581 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1582 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1583 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1584 $indexes[$i] = $index = '';
1593 # Set default structure attribute (word list)
1594 my $struct_attr = q{};
1595 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1596 $struct_attr = ",wrdl";
1599 # Some helpful index variants
1600 my $index_plus = $index . $struct_attr . ':';
1601 my $index_plus_comma = $index . $struct_attr . ',';
1603 if ($auto_truncation){
1604 unless ( $index =~ /,(st-|phr|ext)/ ) {
1605 #FIXME only valid with LTR scripts
1606 $operand=join(" ",map{
1607 (index($_,"*")>0?"$_":"$_*")
1608 }split (/\s+/,$operand));
1609 warn $operand if $DEBUG;
1614 my $truncated_operand;
1615 my( $nontruncated, $righttruncated, $lefttruncated,
1616 $rightlefttruncated, $regexpr
1617 ) = _detect_truncation( $operand, $index );
1619 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1624 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1625 scalar(@$rightlefttruncated) > 0 )
1628 # Don't field weight or add the index to the query, we do it here
1630 undef $weight_fields;
1631 my $previous_truncation_operand;
1632 if (scalar @$nontruncated) {
1633 $truncated_operand .= "$index_plus @$nontruncated ";
1634 $previous_truncation_operand = 1;
1636 if (scalar @$righttruncated) {
1637 $truncated_operand .= "and " if $previous_truncation_operand;
1638 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1639 $previous_truncation_operand = 1;
1641 if (scalar @$lefttruncated) {
1642 $truncated_operand .= "and " if $previous_truncation_operand;
1643 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1644 $previous_truncation_operand = 1;
1646 if (scalar @$rightlefttruncated) {
1647 $truncated_operand .= "and " if $previous_truncation_operand;
1648 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1649 $previous_truncation_operand = 1;
1652 $operand = $truncated_operand if $truncated_operand;
1653 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1656 my $stemmed_operand;
1657 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1660 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1662 # Handle Field Weighting
1663 my $weighted_operand;
1664 if ($weight_fields) {
1665 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1666 $operand = $weighted_operand;
1670 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1672 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1674 query_cgi => $query_cgi,
1675 query_desc => $query_desc,
1676 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1677 parsed_operand => $operand,
1678 original_operand => ($operands[$i]) ? $operands[$i] : '',
1680 index_plus => $index_plus,
1681 indexes_set => $indexes_set,
1682 previous_operand => $previous_operand,
1688 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1691 my %group_OR_limits;
1692 my $availability_limit;
1693 foreach my $this_limit (@limits) {
1694 next unless $this_limit;
1695 if ( $this_limit =~ /available/ ) {
1697 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1699 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1700 $availability_limit .=
1701 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1702 $limit_cgi .= "&limit=available";
1706 # group_OR_limits, prefixed by mc-
1707 # OR every member of the group
1708 elsif ( $this_limit =~ /mc/ ) {
1709 my ($k,$v) = split(/:/, $this_limit,2);
1710 if ( $k !~ /mc-i(tem)?type/ ) {
1711 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1712 $this_limit =~ tr/"//d;
1713 $this_limit = $k.":'".$v."'";
1716 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1717 $limit_desc .= " or " if $group_OR_limits{$k};
1718 $group_OR_limits{$k} .= "$this_limit";
1719 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1720 $limit_desc .= " $this_limit";
1723 # Regular old limits
1725 $limit .= " and " if $limit || $query;
1726 $limit .= "$this_limit";
1727 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1728 if ($this_limit =~ /^branch:(.+)/) {
1729 my $branchcode = $1;
1730 my $library = Koha::Libraries->find( $branchcode );
1731 if (defined $library) {
1732 $limit_desc .= " branch:" . $library->branchname;
1734 $limit_desc .= " $this_limit";
1737 $limit_desc .= " $this_limit";
1741 foreach my $k (keys (%group_OR_limits)) {
1742 $limit .= " and " if ( $query || $limit );
1743 $limit .= "($group_OR_limits{$k})";
1745 if ($availability_limit) {
1746 $limit .= " and " if ( $query || $limit );
1747 $limit .= "($availability_limit)";
1750 # Normalize the query and limit strings
1751 # This is flawed , means we can't search anything with : in it
1752 # if user wants to do ccl or cql, start the query with that
1753 # $query =~ s/:/=/g;
1754 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1755 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1756 $query =~ s/(?<=(wrdl)):/=/g;
1757 $query =~ s/(?<=(trn|phr)):/=/g;
1758 $query =~ s/(?<=(st-numeric)):/=/g;
1759 $query =~ s/(?<=(st-year)):/=/g;
1760 $query =~ s/(?<=(st-date-normalized)):/=/g;
1762 for ( $query, $query_desc, $limit, $limit_desc ) {
1763 s/ +/ /g; # remove extra spaces
1764 s/^ //g; # remove any beginning spaces
1765 s/ $//g; # remove any ending spaces
1766 s/==/=/g; # remove double == from query
1768 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1770 for ($query_cgi,$simple_query) {
1773 # append the limit to the query
1774 $query .= " " . $limit;
1778 warn "QUERY:" . $query;
1779 warn "QUERY CGI:" . $query_cgi;
1780 warn "QUERY DESC:" . $query_desc;
1781 warn "LIMIT:" . $limit;
1782 warn "LIMIT CGI:" . $limit_cgi;
1783 warn "LIMIT DESC:" . $limit_desc;
1784 warn "---------\nLeave buildQuery\n---------";
1788 undef, $query, $simple_query, $query_cgi,
1789 $query_desc, $limit, $limit_cgi, $limit_desc,
1794 =head2 _build_initial_query
1796 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1798 Build a section of the initial query containing indexes, operators, and operands.
1802 sub _build_initial_query {
1806 if ($params->{previous_operand}){
1807 #If there is a previous operand, add a supplied operator or the default 'and'
1808 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1811 #NOTE: indexes_set is typically set when doing truncation or field weighting
1812 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1814 #e.g. "kw,wrdl:test"
1815 #e.g. " and kw,wrdl:test"
1816 $params->{query} .= $operator . $operand;
1818 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1819 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1820 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1822 #e.g. " and kw,wrdl: test"
1823 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1825 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1827 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1830 =head2 searchResults
1832 my @search_results = searchResults($search_context, $searchdesc, $hits,
1833 $results_per_page, $offset, $scan,
1836 Format results in a form suitable for passing to the template
1840 # IMO this subroutine is pretty messy still -- it's responsible for
1841 # building the HTML output for the template
1843 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1844 my $dbh = C4::Context->dbh;
1849 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1850 my ($is_opac, $hidelostitems);
1851 if ($search_context eq 'opac') {
1852 $hidelostitems = C4::Context->preference('hidelostitems');
1856 #Build branchnames hash
1857 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1859 # FIXME - We build an authorised values hash here, using the default framework
1860 # though it is possible to have different authvals for different fws.
1862 my $shelflocations =
1863 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1865 # get notforloan authorised value list (see $shelflocations FIXME)
1866 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => { not => undef } });
1867 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1870 my %itemtypes = %{ GetItemTypes() };
1872 #search item field code
1873 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1875 ## find column names of items related to MARC
1876 my %subfieldstosearch;
1877 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1878 for my $column ( @columns ) {
1879 my ( $tagfield, $tagsubfield ) =
1880 &GetMarcFromKohaField( "items." . $column, "" );
1881 if ( defined $tagsubfield ) {
1882 $subfieldstosearch{$column} = $tagsubfield;
1886 # handle which records to actually retrieve
1888 if ( $hits && $offset + $results_per_page <= $hits ) {
1889 $times = $offset + $results_per_page;
1892 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1895 my $marcflavour = C4::Context->preference("marcflavour");
1896 # We get the biblionumber position in MARC
1897 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1899 # set stuff for XSLT processing here once, not later again for every record we retrieved
1900 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1901 my $xslsyspref = $interface . "XSLTResultsDisplay";
1902 my $xslfile = C4::Context->preference($xslsyspref);
1903 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1904 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1906 # loop through all of the records we've retrieved
1907 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1911 # For Scan searches we built USMARC data
1912 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1914 # Normal search, render from Zebra's output
1915 $marcrecord = new_record_from_zebra(
1920 if ( ! defined $marcrecord ) {
1921 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1929 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1930 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1932 SetUTF8Flag($marcrecord);
1933 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1934 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1935 $oldbiblio->{result_number} = $i + 1;
1937 # add imageurl to itemtype if there is one
1938 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1940 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1941 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1942 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1943 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1944 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1946 # edition information, if any
1947 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1948 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1949 # Build summary if there is one (the summary is defined in the itemtypes table)
1950 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1951 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1952 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1953 my @fields = $marcrecord->fields();
1956 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1958 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1959 $tag =~ /(.{3})(.)/;
1960 if($marcrecord->field($1)){
1961 my @abc = $marcrecord->field($1)->subfield($2);
1962 $tags->{$tag} = $#abc + 1 ;
1966 # We catch how many times to repeat this line
1968 foreach my $tag (keys(%$tags)){
1969 $max = $tags->{$tag} if($tags->{$tag} > $max);
1972 # we replace, and repeat each line
1973 for (my $i = 0 ; $i < $max ; $i++){
1974 my $newline = $line;
1976 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1977 $tag =~ /(.{3})(.)/;
1979 if($marcrecord->field($1)){
1980 my @repl = $marcrecord->field($1)->subfield($2);
1981 my $subfieldvalue = $repl[$i];
1982 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1985 $newsummary .= "$newline\n";
1989 $newsummary =~ s/\[(.*?)]//g;
1990 $newsummary =~ s/\n/<br\/>/g;
1991 $oldbiblio->{summary} = $newsummary;
1994 # Pull out the items fields
1995 my @fields = $marcrecord->field($itemtag);
1996 my $marcflavor = C4::Context->preference("marcflavour");
1997 # adding linked items that belong to host records
1998 my $analyticsfield = '773';
1999 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2000 $analyticsfield = '773';
2001 } elsif ($marcflavor eq 'UNIMARC') {
2002 $analyticsfield = '461';
2004 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2005 my $hostbiblionumber = $hostfield->subfield("0");
2006 my $linkeditemnumber = $hostfield->subfield("9");
2007 if(!$hostbiblionumber eq undef){
2008 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2009 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2010 if(!$hostbiblio eq undef){
2011 my @hostitems = $hostbiblio->field($itemfield);
2012 foreach my $hostitem (@hostitems){
2013 if ($hostitem->subfield("9") eq $linkeditemnumber){
2014 my $linkeditem =$hostitem;
2015 # append linked items if they exist
2016 if (!$linkeditem eq undef){
2017 push (@fields, $linkeditem);}
2024 # Setting item statuses for display
2025 my @available_items_loop;
2026 my @onloan_items_loop;
2027 my @other_items_loop;
2029 my $available_items;
2033 my $ordered_count = 0;
2034 my $available_count = 0;
2035 my $onloan_count = 0;
2036 my $longoverdue_count = 0;
2037 my $other_count = 0;
2038 my $withdrawn_count = 0;
2039 my $itemlost_count = 0;
2040 my $hideatopac_count = 0;
2041 my $itembinding_count = 0;
2042 my $itemdamaged_count = 0;
2043 my $item_in_transit_count = 0;
2044 my $can_place_holds = 0;
2045 my $item_onhold_count = 0;
2046 my $notforloan_count = 0;
2047 my $items_count = scalar(@fields);
2048 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2049 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2050 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2052 # loop through every item
2053 foreach my $field (@fields) {
2056 # populate the items hash
2057 foreach my $code ( keys %subfieldstosearch ) {
2058 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2060 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2064 # hidden because lost
2065 if ($hidelostitems && $item->{itemlost}) {
2066 $hideatopac_count++;
2069 # hidden based on OpacHiddenItems syspref
2070 my @hi = C4::Items::GetHiddenItemnumbers($item);
2072 push @hiddenitems, @hi;
2073 $hideatopac_count++;
2078 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2079 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2081 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2082 if ($item->{$hbranch}) {
2083 $item->{'branchname'} = $branches{$item->{$hbranch}};
2085 elsif ($item->{$otherbranch}) { # Last resort
2086 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2089 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2090 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2091 my $userenv = C4::Context->userenv;
2092 if ( $item->{onloan}
2094 && $userenv->{number}
2095 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2098 my $key = $prefix . $item->{onloan} . $item->{barcode};
2099 $onloan_items->{$key}->{due_date} = output_pref( { str => $item->{onloan}, dateonly => 1 } );
2100 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2101 $onloan_items->{$key}->{branchname} = $item->{branchname};
2102 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2103 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2104 $onloan_items->{$key}->{description} = $item->{description};
2105 $onloan_items->{$key}->{imageurl} =
2106 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2108 # if something's checked out and lost, mark it as 'long overdue'
2109 if ( $item->{itemlost} ) {
2110 $onloan_items->{$key}->{longoverdue}++;
2111 $longoverdue_count++;
2113 else { # can place holds as long as item isn't lost
2114 $can_place_holds = 1;
2118 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2121 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2124 if ( $item->{notforloan} < 0 ) {
2126 } elsif ( $item->{notforloan} > 0 ) {
2127 $notforloan_count++;
2130 # is item in transit?
2131 my $transfertwhen = '';
2132 my ($transfertfrom, $transfertto);
2134 # is item on the reserve shelf?
2135 my $reservestatus = '';
2137 unless ($item->{withdrawn}
2138 || $item->{itemlost}
2140 || $item->{notforloan}
2141 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2142 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2144 # A couple heuristics to limit how many times
2145 # we query the database for item transfer information, sacrificing
2146 # accuracy in some cases for speed;
2148 # 1. don't query if item has one of the other statuses
2149 # 2. don't check transit status if the bib has
2150 # more than 20 items
2152 # FIXME: to avoid having the query the database like this, and to make
2153 # the in transit status count as unavailable for search limiting,
2154 # should map transit status to record indexed in Zebra.
2156 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2157 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2160 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2161 if ( $item->{withdrawn}
2162 || $item->{itemlost}
2164 || $item->{notforloan}
2165 || $reservestatus eq 'Waiting'
2166 || ($transfertwhen ne ''))
2168 $withdrawn_count++ if $item->{withdrawn};
2169 $itemlost_count++ if $item->{itemlost};
2170 $itemdamaged_count++ if $item->{damaged};
2171 $item_in_transit_count++ if $transfertwhen ne '';
2172 $item_onhold_count++ if $reservestatus eq 'Waiting';
2173 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2175 # can place a hold on a item if
2176 # not lost nor withdrawn
2177 # not damaged unless AllowHoldsOnDamagedItems is true
2178 # item is either for loan or on order (notforloan < 0)
2179 $can_place_holds = 1
2182 && !$item->{withdrawn}
2183 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2184 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2189 my $key = $prefix . $item->{status};
2190 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2191 $other_items->{$key}->{$_} = $item->{$_};
2193 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2194 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2195 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2196 $other_items->{$key}->{count}++ if $item->{$hbranch};
2197 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2198 $other_items->{$key}->{description} = $item->{description};
2199 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2203 $can_place_holds = 1;
2205 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2206 foreach (qw(branchname itemcallnumber description)) {
2207 $available_items->{$prefix}->{$_} = $item->{$_};
2209 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2210 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2213 } # notforloan, item level and biblioitem level
2215 # if all items are hidden, do not show the record
2216 if ($items_count > 0 && $hideatopac_count == $items_count) {
2220 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2221 for my $key ( sort keys %$onloan_items ) {
2222 (++$onloanitemscount > $maxitems) and last;
2223 push @onloan_items_loop, $onloan_items->{$key};
2225 for my $key ( sort keys %$other_items ) {
2226 (++$otheritemscount > $maxitems) and last;
2227 push @other_items_loop, $other_items->{$key};
2229 for my $key ( sort keys %$available_items ) {
2230 (++$availableitemscount > $maxitems) and last;
2231 push @available_items_loop, $available_items->{$key}
2234 # XSLT processing of some stuff
2235 # we fetched the sysprefs already before the loop through all retrieved record!
2236 if (!$scan && $xslfile) {
2237 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2238 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2241 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2242 if (!C4::Context->preference("item-level_itypes")) {
2243 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2244 $can_place_holds = 0;
2247 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2248 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2249 $oldbiblio->{items_count} = $items_count;
2250 $oldbiblio->{available_items_loop} = \@available_items_loop;
2251 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2252 $oldbiblio->{other_items_loop} = \@other_items_loop;
2253 $oldbiblio->{availablecount} = $available_count;
2254 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2255 $oldbiblio->{onloancount} = $onloan_count;
2256 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2257 $oldbiblio->{othercount} = $other_count;
2258 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2259 $oldbiblio->{withdrawncount} = $withdrawn_count;
2260 $oldbiblio->{itemlostcount} = $itemlost_count;
2261 $oldbiblio->{damagedcount} = $itemdamaged_count;
2262 $oldbiblio->{intransitcount} = $item_in_transit_count;
2263 $oldbiblio->{onholdcount} = $item_onhold_count;
2264 $oldbiblio->{orderedcount} = $ordered_count;
2265 $oldbiblio->{notforloancount} = $notforloan_count;
2267 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2268 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2269 my $subfields = substr $fieldspec, 3;
2270 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2271 my @alternateholdingsinfo = ();
2272 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2273 my $alternateholdingscount = 0;
2275 for my $field (@holdingsfields) {
2276 my %holding = ( holding => '' );
2277 my $havesubfield = 0;
2278 for my $subfield ($field->subfields()) {
2279 if ((index $subfields, $$subfield[0]) >= 0) {
2280 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2281 $holding{'holding'} .= $$subfield[1];
2285 if ($havesubfield) {
2286 push(@alternateholdingsinfo, \%holding);
2287 $alternateholdingscount++;
2291 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2292 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2295 push( @newresults, $oldbiblio );
2301 =head2 enabled_staff_search_views
2303 %hash = enabled_staff_search_views()
2305 This function returns a hash that contains three flags obtained from the system
2306 preferences, used to determine whether a particular staff search results view
2311 =item C<Output arg:>
2313 * $hash{can_view_MARC} is true only if the MARC view is enabled
2314 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2315 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2317 =item C<usage in the script:>
2321 $template->param ( C4::Search::enabled_staff_search_views );
2325 sub enabled_staff_search_views
2328 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2329 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2330 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2334 =head2 z3950_search_args
2336 $arrayref = z3950_search_args($matchpoints)
2338 This function returns an array reference that contains the search parameters to be
2339 passed to the Z39.50 search script (z3950_search.pl). The array elements
2340 are hash refs whose keys are name and value, and whose values are the
2341 name of a search parameter, the value of that search parameter and the URL encoded
2342 value of that parameter.
2344 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2346 The search parameter values are obtained from the bibliographic record whose
2347 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2349 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2350 a general purpose search argument. In this case, the returned array contains only
2351 entry: the key is 'title' and the value is derived from $matchpoints.
2353 If a search parameter value is undefined or empty, it is not included in the returned
2356 The returned array reference may be passed directly to the template parameters.
2360 =item C<Output arg:>
2362 * $array containing hash refs as described above
2364 =item C<usage in the script:>
2368 $data = Biblio::GetBiblioData($bibno);
2369 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2373 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2377 sub z3950_search_args {
2380 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2381 my $isbn = Business::ISBN->new( $isbn_string );
2383 if (defined $isbn && $isbn->is_valid)
2385 if ( ref($bibrec) ) {
2386 $bibrec->{isbn} = $isbn_string;
2387 $bibrec->{title} = undef;
2389 $bibrec = { isbn => $isbn_string };
2393 $bibrec = { title => $bibrec } if !ref $bibrec;
2396 for my $field (qw/ lccn isbn issn title author dewey subject /)
2398 push @$array, { name => $field, value => $bibrec->{$field} }
2399 if defined $bibrec->{$field};
2404 =head2 GetDistinctValues($field);
2406 C<$field> is a reference to the fields array
2410 sub GetDistinctValues {
2411 my ($fieldname,$string)=@_;
2412 # returns a reference to a hash of references to branches...
2413 if ($fieldname=~/\./){
2414 my ($table,$column)=split /\./, $fieldname;
2415 my $dbh = C4::Context->dbh;
2416 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2417 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 ");
2419 my $elements=$sth->fetchall_arrayref({});
2424 my @servers=qw<biblioserver authorityserver>;
2425 my (@zconns,@results);
2426 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2427 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2430 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2433 # The big moment: asynchronously retrieve results from all servers
2439 my ( $i, $size ) = @_;
2440 for ( my $j = 0 ; $j < $size ; $j++ ) {
2442 @hashscan{qw(value cnt)} =
2443 $results[ $i - 1 ]->display_term($j);
2444 push @elements, \%hashscan;
2452 =head2 _ZOOM_event_loop
2454 _ZOOM_event_loop(\@zconns, \@results, sub {
2455 my ( $i, $size ) = @_;
2459 Processes a ZOOM event loop and passes control to a closure for
2460 processing the results, and destroying the resultsets.
2464 sub _ZOOM_event_loop {
2465 my ($zconns, $results, $callback) = @_;
2466 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2467 my $ev = $zconns->[ $i - 1 ]->last_event();
2468 if ( $ev == ZOOM::Event::ZEND ) {
2469 next unless $results->[ $i - 1 ];
2470 my $size = $results->[ $i - 1 ]->size();
2472 $callback->($i, $size);
2477 foreach my $result (@$results) {
2482 =head2 new_record_from_zebra
2484 Given raw data from a searchengine result set, return a MARC::Record object
2486 This helper function is needed to take into account all the involved
2487 system preferences and configuration variables to properly create the
2488 MARC::Record object.
2490 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2491 data. If we are using DOM, then it has to be MARCXML.
2493 If we are using elasticsearch, it'll already be a MARC::Record and this
2494 function needs a new name.
2498 sub new_record_from_zebra {
2501 my $raw_data = shift;
2502 # Set the default indexing modes
2503 my $search_engine = C4::Context->preference("SearchEngine");
2504 if ($search_engine eq 'Elasticsearch') {
2507 my $index_mode = ( $server eq 'biblioserver' )
2508 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2509 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2511 my $marc_record = eval {
2512 if ( $index_mode eq 'dom' ) {
2513 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2515 MARC::Record->new_from_usmarc( $raw_data );
2522 return $marc_record;
2527 END { } # module clean-up code here (global destructor)
2534 Koha Development Team <http://koha-community.org/>