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;
42 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
50 C4::Search - Functions for searching the Koha catalog.
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
58 This module provides searching functions for Koha's bibliographic databases
72 &enabled_staff_search_views
75 # make all your functions, whether exported or not;
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $record, '' );
93 my ( $biblionumber, $title );
95 # search duplicate on ISBN, easy and fast..
97 if ( $result->{isbn} ) {
98 $result->{isbn} =~ s/\(.*$//;
99 $result->{isbn} =~ s/\s+$//;
100 $query = "isbn:$result->{isbn}";
104 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
110 $titleindex = 'title|exact';
111 $authorindex = 'author|exact';
113 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
115 $titleindex = 'ti,ext';
116 $authorindex = 'au,ext';
120 $result->{title} =~ s /\\//g;
121 $result->{title} =~ s /\"//g;
122 $result->{title} =~ s /\(//g;
123 $result->{title} =~ s /\)//g;
125 # FIXME: instead of removing operators, could just do
126 # quotes around the value
127 $result->{title} =~ s/(and|or|not)//g;
128 $query = "$titleindex:\"$result->{title}\"";
129 if ( $result->{author} ) {
130 $result->{author} =~ s /\\//g;
131 $result->{author} =~ s /\"//g;
132 $result->{author} =~ s /\(//g;
133 $result->{author} =~ s /\)//g;
135 # remove valid operators
136 $result->{author} =~ s/(and|or|not)//g;
137 $query .= " $op $authorindex:\"$result->{author}\"";
141 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
143 if (!defined $error) {
144 foreach my $possible_duplicate_record (@{$searchresults}) {
145 my $marcrecord = new_record_from_zebra(
147 $possible_duplicate_record
150 my $result = TransformMarcToKoha( $marcrecord, '' );
152 # FIXME :: why 2 $biblionumber ?
154 push @results, $result->{'biblionumber'};
155 push @results, $result->{'title'};
164 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
166 This function provides a simple search API on the bibliographic catalog
172 * $query can be a simple keyword or a complete CCL query
173 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
174 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
175 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
176 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
181 Returns an array consisting of three elements
182 * $error is undefined unless an error is detected
183 * $results is a reference to an array of records.
184 * $total_hits is the number of hits that would have been returned with no limit
186 If an error is returned the two other return elements are undefined. If error itself is undefined
187 the other two elements are always defined
189 =item C<usage in the script:>
193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
195 if (defined $error) {
196 $template->param(query_error => $error);
197 warn "error: ".$error;
198 output_html_with_http_headers $input, $cookie, $template->output;
202 my $hits = @{$marcresults};
205 for my $r ( @{$marcresults} ) {
206 my $marcrecord = MARC::File::USMARC::decode($r);
207 my $biblio = TransformMarcToKoha($marcrecord,q{});
209 #build the iarray of hashs for the template.
211 title => $biblio->{'title'},
212 subtitle => $biblio->{'subtitle'},
213 biblionumber => $biblio->{'biblionumber'},
214 author => $biblio->{'author'},
215 publishercode => $biblio->{'publishercode'},
216 publicationyear => $biblio->{'publicationyear'},
221 $template->param(result=>\@results);
226 my ( $query, $offset, $max_results, $servers, %options ) = @_;
228 return ( 'No query entered', undef, undef ) unless $query;
229 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
230 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
238 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
240 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
243 # Initialize & Search Zebra
244 for ( my $i = 0 ; $i < @servers ; $i++ ) {
246 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
248 $query =~ s/=/:/g unless $options{skip_normalize};
249 $QParser->parse( $query );
250 $query = $QParser->target_syntax($servers[$i]);
251 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
253 $query =~ s/:/=/g unless $options{skip_normalize};
254 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
256 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
260 $zconns[$i]->errmsg() . " ("
261 . $zconns[$i]->errcode() . ") "
262 . $zconns[$i]->addinfo() . " "
263 . $zconns[$i]->diagset();
265 return ( $error, undef, undef ) if $zconns[$i]->errcode();
269 # caught a ZOOM::Exception
273 . $@->addinfo() . " "
275 warn $error." for query: $query";
276 return ( $error, undef, undef );
285 my $first_record = defined($offset) ? $offset + 1 : 1;
286 my $hits = $tmpresults[ $i - 1 ]->size();
287 $total_hits += $hits;
288 my $last_record = $hits;
289 if ( defined $max_results && $offset + $max_results < $hits ) {
290 $last_record = $offset + $max_results;
293 for my $j ( $first_record .. $last_record ) {
295 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
298 push @{$results}, $record if defined $record;
303 foreach my $zoom_query (@zoom_queries) {
304 $zoom_query->destroy();
307 return ( undef, $results, $total_hits );
312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
314 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
315 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
319 The all singing, all dancing, multi-server, asynchronous, scanning,
320 searching, record nabbing, facet-building
322 See verbse embedded documentation.
328 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
329 $results_per_page, $offset, $expanded_facet, $branches,
330 $itemtypes, $query_type, $scan, $opac
333 my @servers = @$servers_ref;
334 my @sort_by = @$sort_by_ref;
335 $offset = 0 if $offset < 0;
337 # Initialize variables for the ZOOM connection and results object
341 my $results_hashref = ();
343 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
344 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
346 # Initialize variables for the faceted results objects
347 my $facets_counter = {};
348 my $facets_info = {};
349 my $facets = getFacets();
351 my @facets_loop; # stores the ref to array of hashes for template facets loop
353 ### LOOP THROUGH THE SERVERS
354 for ( my $i = 0 ; $i < @servers ; $i++ ) {
355 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
357 # perform the search, create the results objects
358 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
359 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
361 #$query_to_use = $simple_query if $scan;
362 warn $simple_query if ( $scan and $DEBUG );
364 # Check if we've got a query_type defined, if so, use it
367 if ($query_type =~ /^ccl/) {
368 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
370 } elsif ($query_type =~ /^cql/) {
371 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
372 } elsif ($query_type =~ /^pqf/) {
373 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
375 warn "Unknown query_type '$query_type'. Results undetermined.";
378 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
380 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
384 warn "WARNING: query problem with $query_to_use " . $@;
387 # Concatenate the sort_by limits and pass them to the results object
388 # Note: sort will override rank
390 foreach my $sort (@sort_by) {
391 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
392 $sort_by .= "1=1003 <i ";
394 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
395 $sort_by .= "1=1003 >i ";
397 elsif ( $sort eq "popularity_asc" ) {
398 $sort_by .= "1=9003 <i ";
400 elsif ( $sort eq "popularity_dsc" ) {
401 $sort_by .= "1=9003 >i ";
403 elsif ( $sort eq "call_number_asc" ) {
404 $sort_by .= "1=8007 <i ";
406 elsif ( $sort eq "call_number_dsc" ) {
407 $sort_by .= "1=8007 >i ";
409 elsif ( $sort eq "pubdate_asc" ) {
410 $sort_by .= "1=31 <i ";
412 elsif ( $sort eq "pubdate_dsc" ) {
413 $sort_by .= "1=31 >i ";
415 elsif ( $sort eq "acqdate_asc" ) {
416 $sort_by .= "1=32 <i ";
418 elsif ( $sort eq "acqdate_dsc" ) {
419 $sort_by .= "1=32 >i ";
421 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
422 $sort_by .= "1=4 <i ";
424 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
425 $sort_by .= "1=4 >i ";
428 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
431 if ( $sort_by && !$scan && $results[$i] ) {
432 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
433 warn "WARNING sort $sort_by failed";
436 } # finished looping through servers
438 # The big moment: asynchronously retrieve results from all servers
443 my ( $i, $size ) = @_;
446 # loop through the results
447 $results_hash->{'hits'} = $size;
449 if ( $offset + $results_per_page <= $size ) {
450 $times = $offset + $results_per_page;
456 for ( my $j = $offset ; $j < $times ; $j++ ) {
460 ## Check if it's an index scan
462 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
464 # here we create a minimal MARC record and hand it off to the
465 # template just like a normal result ... perhaps not ideal, but
467 my $tmprecord = MARC::Record->new();
468 $tmprecord->encoding('UTF-8');
472 # the minimal record in author/title (depending on MARC flavour)
473 if ( C4::Context->preference("marcflavour") eq
476 $tmptitle = MARC::Field->new(
481 $tmprecord->append_fields($tmptitle);
485 MARC::Field->new( '245', ' ', ' ', a => $term, );
487 MARC::Field->new( '100', ' ', ' ', a => $occ, );
488 $tmprecord->append_fields($tmptitle);
489 $tmprecord->append_fields($tmpauthor);
491 $results_hash->{'RECORDS'}[$j] =
492 $tmprecord->as_usmarc();
497 $record = $results[ $i - 1 ]->record($j)->raw();
498 # warn "RECORD $j:".$record;
499 $results_hash->{'RECORDS'}[$j] = $record;
503 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
505 # Fill the facets while we're looping, but only for the
506 # biblioserver and not for a scan
507 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
508 $facets_counter = GetFacets( $results[ $i - 1 ] );
509 $facets_info = _get_facets_info( $facets );
513 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
515 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
516 keys %$facets_counter
520 my $number_of_facets;
521 my @this_facets_array;
524 $facets_counter->{$link_value}
525 ->{$b} <=> $facets_counter->{$link_value}
527 } keys %{ $facets_counter->{$link_value} }
531 if ( ( $number_of_facets <= 5 )
532 || ( $expanded_facet eq $link_value )
533 || ( $facets_info->{$link_value}->{'expanded'} )
537 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
538 my $facet_link_value = $one_facet;
539 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
541 # fix the length that will display in the label,
542 my $facet_label_value = $one_facet;
543 my $facet_max_length = C4::Context->preference(
544 'FacetLabelTruncationLength')
547 substr( $one_facet, 0, $facet_max_length )
549 if length($facet_label_value) >
552 # if it's a branch, label by the name, not the code,
553 if ( $link_value =~ /branch/ ) {
554 if ( defined $branches
555 && ref($branches) eq "HASH"
556 && defined $branches->{$one_facet}
557 && ref( $branches->{$one_facet} ) eq
561 $branches->{$one_facet}
565 $facet_label_value = "*";
569 # if it's a itemtype, label by the name, not the code,
570 if ( $link_value =~ /itype/ ) {
571 if ( defined $itemtypes
572 && ref($itemtypes) eq "HASH"
573 && defined $itemtypes->{$one_facet}
574 && ref( $itemtypes->{$one_facet} ) eq
578 $itemtypes->{$one_facet}
579 ->{translated_description};
583 # also, if it's a location code, use the name instead of the code
584 if ( $link_value =~ /location/ ) {
585 # TODO Retrieve all authorised values at once, instead of 1 query per entry
586 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
587 $facet_label_value = $av->count ? $av->next->opac_description : '';
590 # but we're down with the whole label being in the link's title.
591 push @this_facets_array,
594 $facets_counter->{$link_value}
596 facet_label_value => $facet_label_value,
597 facet_title_value => $one_facet,
598 facet_link_value => $facet_link_value,
599 type_link_value => $link_value,
601 if ($facet_label_value);
605 # handle expanded option
606 unless ( $facets_info->{$link_value}->{'expanded'} ) {
608 if ( ( $number_of_facets > 5 )
609 && ( $expanded_facet ne $link_value ) );
613 type_link_value => $link_value,
614 type_id => $link_value . "_id",
616 . $facets_info->{$link_value}->{'label_value'} =>
618 facets => \@this_facets_array,
619 expandable => $expandable,
620 expand => $link_value,
624 $facets_info->{$link_value}->{'label_value'} =~
627 and ( Koha::Libraries->search->count == 1 )
634 # This sorts the facets into alphabetical order
636 foreach my $f (@facets_loop) {
637 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
639 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
642 return ( undef, $results_hashref, \@facets_loop );
650 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
651 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
653 if ( $indexing_mode eq 'dom' &&
654 $use_zebra_facets ) {
655 $facets = _get_facets_from_zebra( $rs );
657 $facets = _get_facets_from_records( $rs );
663 sub _get_facets_from_records {
667 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
668 my $facets_config = getFacets();
670 my $size = $rs->size();
671 my $jmax = $size > $facets_maxrecs
675 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
677 my $marc_record = new_record_from_zebra (
679 $rs->record( $j )->raw()
682 if ( ! defined $marc_record ) {
683 warn "ERROR DECODING RECORD - $@: " .
684 $rs->record( $j )->raw();
688 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
694 =head2 _get_facets_data_from_record
696 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
698 Internal function that extracts facets information from a MARC::Record object
699 and populates $facets_counter for using in getRecords.
701 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
706 sub _get_facets_data_from_record {
708 my ( $marc_record, $facets, $facets_counter ) = @_;
710 for my $facet (@$facets) {
714 foreach my $tag ( @{ $facet->{ tags } } ) {
716 # tag number is the first three digits
717 my $tag_num = substr( $tag, 0, 3 );
718 # subfields are the remainder
719 my $subfield_letters = substr( $tag, 3 );
721 my @fields = $marc_record->field( $tag_num );
722 foreach my $field (@fields) {
723 # If $field->indicator(1) eq 'z', it means it is a 'see from'
724 # field introduced because of IncludeSeeFromInSearches, so skip it
725 next if $field->indicator(1) eq 'z';
727 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
729 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
730 push @used_datas, $data;
731 $facets_counter->{ $facet->{ idx } }->{ $data }++;
738 =head2 _get_facets_from_zebra
740 my $facets = _get_facets_from_zebra( $result_set )
742 Retrieves facets for a specified result set. It loops through the facets defined
743 in C4::Koha::getFacets and returns a hash with the following structure:
753 sub _get_facets_from_zebra {
757 # save current elementSetName
758 my $elementSetName = $rs->option( 'elementSetName' );
760 my $facets_loop = getFacets();
761 my $facets_data = {};
762 # loop through defined facets and fill the facets hashref
763 foreach my $facet ( @$facets_loop ) {
765 my $idx = $facet->{ idx };
766 my $sep = $facet->{ sep };
767 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
768 if ( $facet_values ) {
769 # we've actually got a result
770 $facets_data->{ $idx } = $facet_values;
773 # set elementSetName to its previous value to avoid side effects
774 $rs->option( elementSetName => $elementSetName );
779 =head2 _get_facet_from_result_set
782 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
784 Internal function that extracts facet information for a specific index ($facet_idx) and
785 returns a hash containing facet values and count:
788 $facet_value => $count ,
792 Warning: this function has the side effect of changing the elementSetName for the result
793 set. It is a helper function for the main loop, which takes care of backing it up for
798 sub _get_facet_from_result_set {
800 my $facet_idx = shift;
804 my $internal_sep = '<*>';
805 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
807 return if ( ! defined $facet_idx || ! defined $rs );
808 # zebra's facet element, untokenized index
809 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
810 # configure zebra results for retrieving the desired facet
811 $rs->option( elementSetName => $facet_element );
812 # get the facet record from result set
813 my $facet = $rs->record( 0 )->raw;
814 # if the facet has no restuls...
815 return if !defined $facet;
816 # TODO: benchmark DOM vs. SAX performance
817 my $facet_dom = XML::LibXML->load_xml(
820 my @terms = $facet_dom->getElementsByTagName('term');
824 foreach my $term ( @terms ) {
825 my $facet_value = $term->textContent;
826 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
827 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
833 =head2 _get_facets_info
835 my $facets_info = C4::Search::_get_facets_info( $facets )
837 Internal function that extracts facets information and properly builds
838 the data structure needed to render facet labels.
842 sub _get_facets_info {
846 my $facets_info = {};
848 for my $facet ( @$facets ) {
849 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
850 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
858 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
859 $results_per_page, $offset, $expanded_facet, $branches,
862 #NOTE: Parameter $branches is not used here !
864 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
866 $paz->search($simple_query);
867 sleep 1; # FIXME: WHY?
870 my $results_hashref = {};
871 my $stats = XMLin($paz->stat);
872 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
874 # for a grouped search result, the number of hits
875 # is the number of groups returned; 'bib_hits' will have
876 # the total number of bibs.
877 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
878 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
880 HIT: foreach my $hit (@{ $results->{'hit'} }) {
881 my $recid = $hit->{recid}->[0];
883 my $work_title = $hit->{'md-work-title'}->[0];
885 if (exists $hit->{'md-work-author'}) {
886 $work_author = $hit->{'md-work-author'}->[0];
888 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
890 my $result_group = {};
891 $result_group->{'group_label'} = $group_label;
892 $result_group->{'group_merge_key'} = $recid;
895 if (exists $hit->{count}) {
896 $count = $hit->{count}->[0];
898 $result_group->{'group_count'} = $count;
900 for (my $i = 0; $i < $count; $i++) {
901 # FIXME -- may need to worry about diacritics here
902 my $rec = $paz->record($recid, $i);
903 push @{ $result_group->{'RECORDS'} }, $rec;
906 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
909 # pass through facets
910 my $termlist_xml = $paz->termlist('author,subject');
911 my $terms = XMLin($termlist_xml, forcearray => 1);
912 my @facets_loop = ();
913 #die Dumper($results);
914 # foreach my $list (sort keys %{ $terms->{'list'} }) {
916 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
918 # facet_label_value => $facet->{'name'}->[0],
921 # push @facets_loop, ( {
922 # type_label => $list,
923 # facets => \@facets,
927 return ( undef, $results_hashref, \@facets_loop );
931 sub _detect_truncation {
932 my ( $operand, $index ) = @_;
933 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
936 my @wordlist = split( /\s/, $operand );
937 foreach my $word (@wordlist) {
938 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
939 push @rightlefttruncated, $word;
941 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
942 push @lefttruncated, $word;
944 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
945 push @righttruncated, $word;
947 elsif ( index( $word, "*" ) < 0 ) {
948 push @nontruncated, $word;
951 push @regexpr, $word;
955 \@nontruncated, \@righttruncated, \@lefttruncated,
956 \@rightlefttruncated, \@regexpr
961 sub _build_stemmed_operand {
962 my ($operand,$lang) = @_;
963 require Lingua::Stem::Snowball ;
964 my $stemmed_operand=q{};
966 # Stemmer needs language
967 return $operand unless $lang;
969 # If operand contains a digit, it is almost certainly an identifier, and should
970 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
971 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
972 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
973 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
974 return $operand if $operand =~ /\d/;
976 # FIXME: the locale should be set based on the user's language and/or search choice
978 # Make sure we only use the first two letters from the language code
979 $lang = lc(substr($lang, 0, 2));
980 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
981 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
982 if ($lang eq 'nb' || $lang eq 'nn') {
985 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
986 encoding => "UTF-8" );
988 my @words = split( / /, $operand );
989 my @stems = $stemmer->stem(\@words);
990 for my $stem (@stems) {
991 $stemmed_operand .= "$stem";
992 $stemmed_operand .= "?"
993 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
994 $stemmed_operand .= " ";
996 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
997 return $stemmed_operand;
1001 sub _build_weighted_query {
1003 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1004 # pretty well but could work much better if we had a smarter query parser
1005 my ( $operand, $stemmed_operand, $index ) = @_;
1006 my $stemming = C4::Context->preference("QueryStemming") || 0;
1007 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1008 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1009 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1011 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1013 # Keyword, or, no index specified
1014 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1016 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1017 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1018 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1019 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1020 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1021 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1022 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1023 if $fuzzy_enabled; # add fuzzy, word list
1024 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1025 if ( $stemming and $stemmed_operand )
1026 ; # add stemming, right truncation
1027 $weighted_query .= " or wrdl,r9=\"$operand\"";
1029 # embedded sorting: 0 a-z; 1 z-a
1030 # $weighted_query .= ") or (sort1,aut=1";
1033 # Barcode searches should skip this process
1034 elsif ( $index eq 'bc' ) {
1035 $weighted_query .= "bc=\"$operand\"";
1038 # Authority-number searches should skip this process
1039 elsif ( $index eq 'an' ) {
1040 $weighted_query .= "an=\"$operand\"";
1043 # If the index is numeric, don't autoquote it.
1044 elsif ( $index =~ /,st-numeric$/ ) {
1045 $weighted_query .= " $index=$operand";
1048 # If the index already has more than one qualifier, wrap the operand
1049 # in quotes and pass it back (assumption is that the user knows what they
1050 # are doing and won't appreciate us mucking up their query
1051 elsif ( $index =~ ',' ) {
1052 $weighted_query .= " $index=\"$operand\"";
1055 #TODO: build better cases based on specific search indexes
1057 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1058 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1059 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1060 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1061 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1062 if $fuzzy_enabled; # add fuzzy, word list
1063 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1064 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1067 $weighted_query .= "))"; # close rank specification
1068 return $weighted_query;
1073 Return an array with available indexes.
1095 'Author-personal-bibliography',
1105 'Chronological-subdivision',
1115 'Conference-name-heading',
1116 'Conference-name-see',
1117 'Conference-name-seealso',
1122 'Corporate-name-heading',
1123 'Corporate-name-see',
1124 'Corporate-name-seealso',
1125 'Country-publication',
1128 'date-entered-on-file',
1129 'Date-of-acquisition',
1130 'Date-of-publication',
1131 'Date-time-last-modified',
1132 'Dewey-classification',
1133 'Dissertation-information',
1142 'Geographic-subdivision',
1145 'Heading-use-main-or-added-entry',
1146 'Heading-use-series-added-entry ',
1147 'Heading-use-subject-added-entry',
1151 'Identifier-standard',
1152 'Illustration-code',
1154 'Index-term-uncontrolled',
1155 'Interest-age-level',
1156 'Interest-grade-level',
1166 'language-original',
1176 'Local-classification',
1179 'Match-heading-see-from',
1187 'Name-geographic-heading',
1188 'Name-geographic-see',
1189 'Name-geographic-seealso',
1195 'Other-control-number',
1198 'Personal-name-heading',
1199 'Personal-name-see',
1200 'Personal-name-seealso',
1202 'Place-publication',
1209 'Reading-grade-level',
1210 'Record-control-number',
1221 'Subject-heading-thesaurus',
1222 'Subject-name-personal',
1223 'Subject-subdivision',
1232 'Term-genre-form-heading',
1233 'Term-genre-form-see',
1234 'Term-genre-form-seealso',
1240 'Title-uniform-heading',
1241 'Title-uniform-see',
1242 'Title-uniform-seealso',
1252 'classification-source',
1254 'coded-location-qualifier',
1265 'Local-classification',
1268 'materials-specified',
1273 'Number-local-acquisition',
1278 'replacementpricedate',
1293 =head2 _handle_exploding_index
1295 my $query = _handle_exploding_index($index, $term)
1297 Callback routine to generate the search for "exploding" indexes (i.e.
1298 those indexes which are turned into multiple or-connected searches based
1303 sub _handle_exploding_index {
1304 my ($QParser, $filter, $params, $negate, $server) = @_;
1305 my $index = $filter;
1306 my $term = join(' ', @$params);
1308 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1310 my $marcflavour = C4::Context->preference('marcflavour');
1312 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1313 my $wantedcodes = '';
1314 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1315 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1316 foreach my $auth (@$results) {
1317 my $record = MARC::Record->new_from_usmarc($auth);
1318 my @references = $record->field('5..');
1320 if ($index eq 'su-br') {
1322 } elsif ($index eq 'su-na') {
1324 } elsif ($index eq 'su-rl') {
1327 foreach my $reference (@references) {
1328 my $codes = $reference->subfield($codesubfield);
1329 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1333 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1339 ( $operators, $operands, $indexes, $limits,
1340 $sort_by, $scan, $lang ) =
1341 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1343 Shim function to ease the transition from buildQuery to a new QueryParser.
1344 This function is called at the beginning of buildQuery, and modifies
1345 buildQuery's input. If it can handle the input, it returns a query that
1346 buildQuery will not try to parse.
1351 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1353 my @operators = $operators ? @$operators : ();
1354 my @indexes = $indexes ? @$indexes : ();
1355 my @operands = $operands ? @$operands : ();
1356 my @limits = $limits ? @$limits : ();
1357 my @sort_by = $sort_by ? @$sort_by : ();
1359 my $query = $operands[0];
1365 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1366 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1367 undef $QParser if (scalar @limits > 0);
1371 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1373 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1374 next unless $operands[$ii];
1375 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1377 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1378 $query .= $operands[$ii];
1380 elsif ( $indexes[$ii] =~ m/su-/ ) {
1381 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1385 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1388 foreach my $limit (@limits) {
1390 if ( scalar(@sort_by) > 0 ) {
1392 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1393 $query =~ s/$modifier_re//g;
1394 foreach my $modifier (@sort_by) {
1395 $query .= " #$modifier";
1399 $query_desc = $query;
1400 $query_desc =~ s/\s+/ /g;
1401 if ( C4::Context->preference("QueryWeightFields") ) {
1403 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1404 { 'target_syntax_callback' => \&_handle_exploding_index } );
1405 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1406 { 'target_syntax_callback' => \&_handle_exploding_index } );
1407 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1408 { 'target_syntax_callback' => \&_handle_exploding_index } );
1409 $QParser->parse($query);
1410 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1413 require Koha::QueryParser::Driver::PQF;
1414 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1415 s/$modifier_re//g for @operands;
1418 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1424 $simple_query, $query_cgi,
1425 $query_desc, $limit,
1426 $limit_cgi, $limit_desc,
1427 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1429 Build queries and limits in CCL, CGI, Human,
1430 handle truncation, stemming, field weighting, fuzziness, etc.
1432 See verbose embedded documentation.
1438 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1440 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1443 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1446 my @operators = $operators ? @$operators : ();
1447 my @indexes = $indexes ? @$indexes : ();
1448 my @operands = $operands ? @$operands : ();
1449 my @limits = $limits ? @$limits : ();
1450 my @sort_by = $sort_by ? @$sort_by : ();
1452 my $stemming = C4::Context->preference("QueryStemming") || 0;
1453 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1454 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1455 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1457 my $query = $operands[0];
1458 my $simple_query = $operands[0];
1460 # initialize the variables we're passing back
1469 my $cclindexes = getIndexes();
1470 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1471 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1473 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1475 $query = "ccl=$query" if $cclq;
1478 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1480 if ( $query =~ /^ccl=/ ) {
1482 # This is needed otherwise ccl= and &limit won't work together, and
1483 # this happens when selecting a subject on the opac-detail page
1484 @limits = grep {!/^$/} @limits;
1485 my $original_q = $q; # without available part
1486 unless ( grep { /^available$/ } @limits ) {
1487 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1491 if ( grep { /^available$/ } @limits ) {
1492 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1493 delete $limits['available'];
1495 $q .= ' and '.join(' and ', @limits) if @limits;
1497 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1499 if ( $query =~ /^cql=/ ) {
1500 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1502 if ( $query =~ /^pqf=/ ) {
1504 $query_cgi = "q=".uri_escape_utf8($query_desc);
1507 $query_cgi = "q=pqf=".uri_escape_utf8($');
1509 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1512 # pass nested queries directly
1513 # FIXME: need better handling of some of these variables in this case
1514 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1515 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1516 # if ( $query =~ /(\(|\))/ ) {
1518 # undef, $query, $simple_query, $query_cgi,
1519 # $query, $limit, $limit_cgi, $limit_desc,
1524 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1525 # query operands and indexes and add stemming, truncation, field weighting, etc.
1526 # Once we do so, we'll end up with a value in $query, just like if we had an
1527 # incoming $query from the user
1530 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1531 my $previous_operand
1532 ; # a flag used to keep track if there was a previous query
1533 # if there was, we can apply the current operator
1535 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1537 # COMBINE OPERANDS, INDEXES AND OPERATORS
1538 if ( ($operands[$i] // '') ne '' ) {
1539 $operands[$i]=~s/^\s+//;
1541 # A flag to determine whether or not to add the index to the query
1544 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1545 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1549 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1551 my $operand = $operands[$i];
1552 my $index = $indexes[$i];
1554 # Add index-specific attributes
1556 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1557 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1558 #which is processed higher up in this sub. Other than that, year searches are typically
1559 #handled as limits which are not processed her either.
1561 # Search ranges: Date of Publication, st-numeric
1562 if ( $index =~ /(yr|st-numeric)/ ) {
1563 #weight_fields/relevance search causes errors with date ranges
1564 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1565 #In the case of YYYY-YYYY, it will return no results
1566 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1569 # Date of Acquisition
1570 elsif ( $index =~ /acqdate/ ) {
1571 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1572 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1573 #top of the results just because they have lots of item records matching that date.
1574 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1576 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1578 # ISBN,ISSN,Standard Number, don't need special treatment
1579 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1581 $stemming, $auto_truncation,
1582 $weight_fields, $fuzzy_enabled
1585 if ( $index eq 'nb' ) {
1586 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1587 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1588 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1589 $indexes[$i] = $index = '';
1598 # Set default structure attribute (word list)
1599 my $struct_attr = q{};
1600 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1601 $struct_attr = ",wrdl";
1604 # Some helpful index variants
1605 my $index_plus = $index . $struct_attr . ':';
1606 my $index_plus_comma = $index . $struct_attr . ',';
1608 if ($auto_truncation){
1609 unless ( $index =~ /,(st-|phr|ext)/ ) {
1610 #FIXME only valid with LTR scripts
1611 $operand=join(" ",map{
1612 (index($_,"*")>0?"$_":"$_*")
1613 }split (/\s+/,$operand));
1614 warn $operand if $DEBUG;
1619 my $truncated_operand;
1620 my( $nontruncated, $righttruncated, $lefttruncated,
1621 $rightlefttruncated, $regexpr
1622 ) = _detect_truncation( $operand, $index );
1624 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1629 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1630 scalar(@$rightlefttruncated) > 0 )
1633 # Don't field weight or add the index to the query, we do it here
1635 undef $weight_fields;
1636 my $previous_truncation_operand;
1637 if (scalar @$nontruncated) {
1638 $truncated_operand .= "$index_plus @$nontruncated ";
1639 $previous_truncation_operand = 1;
1641 if (scalar @$righttruncated) {
1642 $truncated_operand .= "and " if $previous_truncation_operand;
1643 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1644 $previous_truncation_operand = 1;
1646 if (scalar @$lefttruncated) {
1647 $truncated_operand .= "and " if $previous_truncation_operand;
1648 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1649 $previous_truncation_operand = 1;
1651 if (scalar @$rightlefttruncated) {
1652 $truncated_operand .= "and " if $previous_truncation_operand;
1653 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1654 $previous_truncation_operand = 1;
1657 $operand = $truncated_operand if $truncated_operand;
1658 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1661 my $stemmed_operand;
1662 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1665 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1667 # Handle Field Weighting
1668 my $weighted_operand;
1669 if ($weight_fields) {
1670 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1671 $operand = $weighted_operand;
1675 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1677 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1679 query_cgi => $query_cgi,
1680 query_desc => $query_desc,
1681 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1682 parsed_operand => $operand,
1683 original_operand => $operands[$i] // '',
1685 index_plus => $index_plus,
1686 indexes_set => $indexes_set,
1687 previous_operand => $previous_operand,
1693 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1696 my %group_OR_limits;
1697 my $availability_limit;
1698 foreach my $this_limit (@limits) {
1699 next unless $this_limit;
1700 if ( $this_limit =~ /available/ ) {
1702 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1704 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1705 $availability_limit .=
1706 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1707 $limit_cgi .= "&limit=available";
1711 # group_OR_limits, prefixed by mc-
1712 # OR every member of the group
1713 elsif ( $this_limit =~ /mc/ ) {
1714 my ($k,$v) = split(/:/, $this_limit,2);
1715 if ( $k !~ /mc-i(tem)?type/ ) {
1716 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1717 $this_limit =~ tr/"//d;
1718 $this_limit = $k.":'".$v."'";
1721 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1722 $limit_desc .= " or " if $group_OR_limits{$k};
1723 $group_OR_limits{$k} .= "$this_limit";
1724 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1725 $limit_desc .= " $this_limit";
1728 # Regular old limits
1730 $limit .= " and " if $limit || $query;
1731 $limit .= "$this_limit";
1732 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1733 if ($this_limit =~ /^branch:(.+)/) {
1734 my $branchcode = $1;
1735 my $library = Koha::Libraries->find( $branchcode );
1736 if (defined $library) {
1737 $limit_desc .= " branch:" . $library->branchname;
1739 $limit_desc .= " $this_limit";
1742 $limit_desc .= " $this_limit";
1746 foreach my $k (keys (%group_OR_limits)) {
1747 $limit .= " and " if ( $query || $limit );
1748 $limit .= "($group_OR_limits{$k})";
1750 if ($availability_limit) {
1751 $limit .= " and " if ( $query || $limit );
1752 $limit .= "($availability_limit)";
1755 # Normalize the query and limit strings
1756 # This is flawed , means we can't search anything with : in it
1757 # if user wants to do ccl or cql, start the query with that
1758 # $query =~ s/:/=/g;
1759 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1760 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1761 $query =~ s/(?<=(wrdl)):/=/g;
1762 $query =~ s/(?<=(trn|phr)):/=/g;
1763 $query =~ s/(?<=(st-numeric)):/=/g;
1764 $query =~ s/(?<=(st-year)):/=/g;
1765 $query =~ s/(?<=(st-date-normalized)):/=/g;
1767 for ( $query, $query_desc, $limit, $limit_desc ) {
1768 s/ +/ /g; # remove extra spaces
1769 s/^ //g; # remove any beginning spaces
1770 s/ $//g; # remove any ending spaces
1771 s/==/=/g; # remove double == from query
1773 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1775 for ($query_cgi,$simple_query) {
1778 # append the limit to the query
1779 $query .= " " . $limit;
1783 warn "QUERY:" . $query;
1784 warn "QUERY CGI:" . $query_cgi;
1785 warn "QUERY DESC:" . $query_desc;
1786 warn "LIMIT:" . $limit;
1787 warn "LIMIT CGI:" . $limit_cgi;
1788 warn "LIMIT DESC:" . $limit_desc;
1789 warn "---------\nLeave buildQuery\n---------";
1793 undef, $query, $simple_query, $query_cgi,
1794 $query_desc, $limit, $limit_cgi, $limit_desc,
1799 =head2 _build_initial_query
1801 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1803 Build a section of the initial query containing indexes, operators, and operands.
1807 sub _build_initial_query {
1811 if ($params->{previous_operand}){
1812 #If there is a previous operand, add a supplied operator or the default 'and'
1813 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1816 #NOTE: indexes_set is typically set when doing truncation or field weighting
1817 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1819 #e.g. "kw,wrdl:test"
1820 #e.g. " and kw,wrdl:test"
1821 $params->{query} .= $operator . $operand;
1823 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1824 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1825 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1827 #e.g. " and kw,wrdl: test"
1828 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1830 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1832 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1835 =head2 searchResults
1837 my @search_results = searchResults($search_context, $searchdesc, $hits,
1838 $results_per_page, $offset, $scan,
1841 Format results in a form suitable for passing to the template
1845 # IMO this subroutine is pretty messy still -- it's responsible for
1846 # building the HTML output for the template
1848 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1849 my $dbh = C4::Context->dbh;
1854 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1855 my ($is_opac, $hidelostitems);
1856 if ($search_context eq 'opac') {
1857 $hidelostitems = C4::Context->preference('hidelostitems');
1861 #Build branchnames hash
1862 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1864 # FIXME - We build an authorised values hash here, using the default framework
1865 # though it is possible to have different authvals for different fws.
1867 my $shelflocations =
1868 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1870 # get notforloan authorised value list (see $shelflocations FIXME)
1871 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1872 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1875 my $itemtypes = Koha::ItemTypes->search_with_localization;
1876 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1878 #search item field code
1879 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1881 ## find column names of items related to MARC
1882 my %subfieldstosearch;
1883 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1884 for my $column ( @columns ) {
1885 my ( $tagfield, $tagsubfield ) =
1886 &GetMarcFromKohaField( "items." . $column, "" );
1887 if ( defined $tagsubfield ) {
1888 $subfieldstosearch{$column} = $tagsubfield;
1892 # handle which records to actually retrieve
1894 if ( $hits && $offset + $results_per_page <= $hits ) {
1895 $times = $offset + $results_per_page;
1898 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1901 my $marcflavour = C4::Context->preference("marcflavour");
1902 # We get the biblionumber position in MARC
1903 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1905 # set stuff for XSLT processing here once, not later again for every record we retrieved
1906 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1907 my $xslsyspref = $interface . "XSLTResultsDisplay";
1908 my $xslfile = C4::Context->preference($xslsyspref);
1909 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1910 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1912 # loop through all of the records we've retrieved
1913 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1917 # For Scan searches we built USMARC data
1918 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1920 # Normal search, render from Zebra's output
1921 $marcrecord = new_record_from_zebra(
1926 if ( ! defined $marcrecord ) {
1927 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1935 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1936 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1938 SetUTF8Flag($marcrecord);
1939 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1940 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1941 $oldbiblio->{result_number} = $i + 1;
1943 # add imageurl to itemtype if there is one
1944 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1946 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1947 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1948 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1949 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1950 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1952 # edition information, if any
1953 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1954 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1955 # Build summary if there is one (the summary is defined in the itemtypes table)
1956 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1957 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1958 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1959 my @fields = $marcrecord->fields();
1962 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1964 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1965 $tag =~ /(.{3})(.)/;
1966 if($marcrecord->field($1)){
1967 my @abc = $marcrecord->field($1)->subfield($2);
1968 $tags->{$tag} = $#abc + 1 ;
1972 # We catch how many times to repeat this line
1974 foreach my $tag (keys(%$tags)){
1975 $max = $tags->{$tag} if($tags->{$tag} > $max);
1978 # we replace, and repeat each line
1979 for (my $i = 0 ; $i < $max ; $i++){
1980 my $newline = $line;
1982 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1983 $tag =~ /(.{3})(.)/;
1985 if($marcrecord->field($1)){
1986 my @repl = $marcrecord->field($1)->subfield($2);
1987 my $subfieldvalue = $repl[$i];
1988 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1991 $newsummary .= "$newline\n";
1995 $newsummary =~ s/\[(.*?)]//g;
1996 $newsummary =~ s/\n/<br\/>/g;
1997 $oldbiblio->{summary} = $newsummary;
2000 # Pull out the items fields
2001 my @fields = $marcrecord->field($itemtag);
2002 my $marcflavor = C4::Context->preference("marcflavour");
2003 # adding linked items that belong to host records
2004 my $analyticsfield = '773';
2005 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2006 $analyticsfield = '773';
2007 } elsif ($marcflavor eq 'UNIMARC') {
2008 $analyticsfield = '461';
2010 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2011 my $hostbiblionumber = $hostfield->subfield("0");
2012 my $linkeditemnumber = $hostfield->subfield("9");
2013 if( $hostbiblionumber ) {
2014 my $hostbiblio = GetMarcBiblio({
2015 biblionumber => $hostbiblionumber,
2016 embed_items => 1 });
2017 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2019 my @hostitems = $hostbiblio->field($itemfield);
2020 foreach my $hostitem (@hostitems){
2021 if ($hostitem->subfield("9") eq $linkeditemnumber){
2022 my $linkeditem =$hostitem;
2023 # append linked items if they exist
2024 push @fields, $linkeditem if $linkeditem;
2031 # Setting item statuses for display
2032 my @available_items_loop;
2033 my @onloan_items_loop;
2034 my @other_items_loop;
2036 my $available_items;
2040 my $ordered_count = 0;
2041 my $available_count = 0;
2042 my $onloan_count = 0;
2043 my $longoverdue_count = 0;
2044 my $other_count = 0;
2045 my $withdrawn_count = 0;
2046 my $itemlost_count = 0;
2047 my $hideatopac_count = 0;
2048 my $itembinding_count = 0;
2049 my $itemdamaged_count = 0;
2050 my $item_in_transit_count = 0;
2051 my $can_place_holds = 0;
2052 my $item_onhold_count = 0;
2053 my $notforloan_count = 0;
2054 my $items_count = scalar(@fields);
2055 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2056 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2057 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2059 # loop through every item
2060 foreach my $field (@fields) {
2063 # populate the items hash
2064 foreach my $code ( keys %subfieldstosearch ) {
2065 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2067 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2071 # hidden because lost
2072 if ($hidelostitems && $item->{itemlost}) {
2073 $hideatopac_count++;
2076 # hidden based on OpacHiddenItems syspref
2077 my @hi = C4::Items::GetHiddenItemnumbers($item);
2079 push @hiddenitems, @hi;
2080 $hideatopac_count++;
2085 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2086 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2088 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2089 if ($item->{$hbranch}) {
2090 $item->{'branchname'} = $branches{$item->{$hbranch}};
2092 elsif ($item->{$otherbranch}) { # Last resort
2093 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2096 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2097 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2098 my $userenv = C4::Context->userenv;
2099 if ( $item->{onloan}
2101 && $userenv->{number}
2102 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2105 my $key = $prefix . $item->{onloan} . $item->{barcode};
2106 $onloan_items->{$key}->{due_date} = $item->{onloan};
2107 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2108 $onloan_items->{$key}->{branchname} = $item->{branchname};
2109 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2110 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2111 $onloan_items->{$key}->{description} = $item->{description};
2112 $onloan_items->{$key}->{imageurl} =
2113 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2115 # if something's checked out and lost, mark it as 'long overdue'
2116 if ( $item->{itemlost} ) {
2117 $onloan_items->{$key}->{longoverdue}++;
2118 $longoverdue_count++;
2120 else { # can place holds as long as item isn't lost
2121 $can_place_holds = 1;
2125 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2128 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2131 if ( $item->{notforloan} < 0 ) {
2133 } elsif ( $item->{notforloan} > 0 ) {
2134 $notforloan_count++;
2137 # is item in transit?
2138 my $transfertwhen = '';
2139 my ($transfertfrom, $transfertto);
2141 # is item on the reserve shelf?
2142 my $reservestatus = '';
2144 unless ($item->{withdrawn}
2145 || $item->{itemlost}
2147 || $item->{notforloan}
2148 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2149 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2151 # A couple heuristics to limit how many times
2152 # we query the database for item transfer information, sacrificing
2153 # accuracy in some cases for speed;
2155 # 1. don't query if item has one of the other statuses
2156 # 2. don't check transit status if the bib has
2157 # more than 20 items
2159 # FIXME: to avoid having the query the database like this, and to make
2160 # the in transit status count as unavailable for search limiting,
2161 # should map transit status to record indexed in Zebra.
2163 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2164 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2167 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2168 if ( $item->{withdrawn}
2169 || $item->{itemlost}
2171 || $item->{notforloan}
2172 || $reservestatus eq 'Waiting'
2173 || ($transfertwhen ne ''))
2175 $withdrawn_count++ if $item->{withdrawn};
2176 $itemlost_count++ if $item->{itemlost};
2177 $itemdamaged_count++ if $item->{damaged};
2178 $item_in_transit_count++ if $transfertwhen ne '';
2179 $item_onhold_count++ if $reservestatus eq 'Waiting';
2180 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2182 # can place a hold on a item if
2183 # not lost nor withdrawn
2184 # not damaged unless AllowHoldsOnDamagedItems is true
2185 # item is either for loan or on order (notforloan < 0)
2186 $can_place_holds = 1
2189 && !$item->{withdrawn}
2190 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2191 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2196 my $key = $prefix . $item->{status};
2197 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2198 $other_items->{$key}->{$_} = $item->{$_};
2200 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2201 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2202 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2203 $other_items->{$key}->{count}++ if $item->{$hbranch};
2204 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2205 $other_items->{$key}->{description} = $item->{description};
2206 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2210 $can_place_holds = 1;
2212 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2213 foreach (qw(branchname itemcallnumber description)) {
2214 $available_items->{$prefix}->{$_} = $item->{$_};
2216 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2217 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2220 } # notforloan, item level and biblioitem level
2222 # if all items are hidden, do not show the record
2223 if ($items_count > 0 && $hideatopac_count == $items_count) {
2227 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2228 for my $key ( sort keys %$onloan_items ) {
2229 (++$onloanitemscount > $maxitems) and last;
2230 push @onloan_items_loop, $onloan_items->{$key};
2232 for my $key ( sort keys %$other_items ) {
2233 (++$otheritemscount > $maxitems) and last;
2234 push @other_items_loop, $other_items->{$key};
2236 for my $key ( sort keys %$available_items ) {
2237 (++$availableitemscount > $maxitems) and last;
2238 push @available_items_loop, $available_items->{$key}
2241 # XSLT processing of some stuff
2242 # we fetched the sysprefs already before the loop through all retrieved record!
2243 if (!$scan && $xslfile) {
2244 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2245 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2248 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2249 if (!C4::Context->preference("item-level_itypes")) {
2250 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2251 $can_place_holds = 0;
2254 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2255 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2256 $oldbiblio->{items_count} = $items_count;
2257 $oldbiblio->{available_items_loop} = \@available_items_loop;
2258 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2259 $oldbiblio->{other_items_loop} = \@other_items_loop;
2260 $oldbiblio->{availablecount} = $available_count;
2261 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2262 $oldbiblio->{onloancount} = $onloan_count;
2263 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2264 $oldbiblio->{othercount} = $other_count;
2265 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2266 $oldbiblio->{withdrawncount} = $withdrawn_count;
2267 $oldbiblio->{itemlostcount} = $itemlost_count;
2268 $oldbiblio->{damagedcount} = $itemdamaged_count;
2269 $oldbiblio->{intransitcount} = $item_in_transit_count;
2270 $oldbiblio->{onholdcount} = $item_onhold_count;
2271 $oldbiblio->{orderedcount} = $ordered_count;
2272 $oldbiblio->{notforloancount} = $notforloan_count;
2274 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2275 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2276 my $subfields = substr $fieldspec, 3;
2277 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2278 my @alternateholdingsinfo = ();
2279 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2280 my $alternateholdingscount = 0;
2282 for my $field (@holdingsfields) {
2283 my %holding = ( holding => '' );
2284 my $havesubfield = 0;
2285 for my $subfield ($field->subfields()) {
2286 if ((index $subfields, $$subfield[0]) >= 0) {
2287 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2288 $holding{'holding'} .= $$subfield[1];
2292 if ($havesubfield) {
2293 push(@alternateholdingsinfo, \%holding);
2294 $alternateholdingscount++;
2298 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2299 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2302 push( @newresults, $oldbiblio );
2308 =head2 enabled_staff_search_views
2310 %hash = enabled_staff_search_views()
2312 This function returns a hash that contains three flags obtained from the system
2313 preferences, used to determine whether a particular staff search results view
2318 =item C<Output arg:>
2320 * $hash{can_view_MARC} is true only if the MARC view is enabled
2321 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2322 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2324 =item C<usage in the script:>
2328 $template->param ( C4::Search::enabled_staff_search_views );
2332 sub enabled_staff_search_views
2335 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2336 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2337 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2341 =head2 z3950_search_args
2343 $arrayref = z3950_search_args($matchpoints)
2345 This function returns an array reference that contains the search parameters to be
2346 passed to the Z39.50 search script (z3950_search.pl). The array elements
2347 are hash refs whose keys are name and value, and whose values are the
2348 name of a search parameter, the value of that search parameter and the URL encoded
2349 value of that parameter.
2351 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2353 The search parameter values are obtained from the bibliographic record whose
2354 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2356 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2357 a general purpose search argument. In this case, the returned array contains only
2358 entry: the key is 'title' and the value is derived from $matchpoints.
2360 If a search parameter value is undefined or empty, it is not included in the returned
2363 The returned array reference may be passed directly to the template parameters.
2367 =item C<Output arg:>
2369 * $array containing hash refs as described above
2371 =item C<usage in the script:>
2375 $data = Biblio::GetBiblioData($bibno);
2376 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2380 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2384 sub z3950_search_args {
2387 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2388 my $isbn = Business::ISBN->new( $isbn_string );
2390 if (defined $isbn && $isbn->is_valid)
2392 if ( ref($bibrec) ) {
2393 $bibrec->{isbn} = $isbn_string;
2394 $bibrec->{title} = undef;
2396 $bibrec = { isbn => $isbn_string };
2400 $bibrec = { title => $bibrec } if !ref $bibrec;
2403 for my $field (qw/ lccn isbn issn title author dewey subject /)
2405 push @$array, { name => $field, value => $bibrec->{$field} }
2406 if defined $bibrec->{$field};
2411 =head2 GetDistinctValues($field);
2413 C<$field> is a reference to the fields array
2417 sub GetDistinctValues {
2418 my ($fieldname,$string)=@_;
2419 # returns a reference to a hash of references to branches...
2420 if ($fieldname=~/\./){
2421 my ($table,$column)=split /\./, $fieldname;
2422 my $dbh = C4::Context->dbh;
2423 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2424 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 ");
2426 my $elements=$sth->fetchall_arrayref({});
2431 my @servers=qw<biblioserver authorityserver>;
2432 my (@zconns,@results);
2433 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2434 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2437 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2440 # The big moment: asynchronously retrieve results from all servers
2446 my ( $i, $size ) = @_;
2447 for ( my $j = 0 ; $j < $size ; $j++ ) {
2449 @hashscan{qw(value cnt)} =
2450 $results[ $i - 1 ]->display_term($j);
2451 push @elements, \%hashscan;
2459 =head2 _ZOOM_event_loop
2461 _ZOOM_event_loop(\@zconns, \@results, sub {
2462 my ( $i, $size ) = @_;
2466 Processes a ZOOM event loop and passes control to a closure for
2467 processing the results, and destroying the resultsets.
2471 sub _ZOOM_event_loop {
2472 my ($zconns, $results, $callback) = @_;
2473 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2474 my $ev = $zconns->[ $i - 1 ]->last_event();
2475 if ( $ev == ZOOM::Event::ZEND ) {
2476 next unless $results->[ $i - 1 ];
2477 my $size = $results->[ $i - 1 ]->size();
2479 $callback->($i, $size);
2484 foreach my $result (@$results) {
2489 =head2 new_record_from_zebra
2491 Given raw data from a searchengine result set, return a MARC::Record object
2493 This helper function is needed to take into account all the involved
2494 system preferences and configuration variables to properly create the
2495 MARC::Record object.
2497 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2498 data. If we are using DOM, then it has to be MARCXML.
2500 If we are using elasticsearch, it'll already be a MARC::Record and this
2501 function needs a new name.
2505 sub new_record_from_zebra {
2508 my $raw_data = shift;
2509 # Set the default indexing modes
2510 my $search_engine = C4::Context->preference("SearchEngine");
2511 if ($search_engine eq 'Elasticsearch') {
2512 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2514 my $index_mode = ( $server eq 'biblioserver' )
2515 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2516 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2518 my $marc_record = eval {
2519 if ( $index_mode eq 'dom' ) {
2520 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2522 MARC::Record->new_from_usmarc( $raw_data );
2529 return $marc_record;
2534 END { } # module clean-up code here (global destructor)
2541 Koha Development Team <http://koha-community.org/>