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>.
21 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
22 use C4::Koha; # getFacets
28 use C4::Reserves; # GetReserveStatus
31 use Koha::AuthorisedValues;
35 use Koha::RecordProcessor;
40 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
43 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
48 C4::Search - Functions for searching the Koha catalog.
52 See opac/opac-search.pl or catalogue/search.pl for example of usage
56 This module provides searching functions for Koha's bibliographic databases
70 &enabled_staff_search_views
73 # make all your functions, whether exported or not;
77 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
79 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
85 my $dbh = C4::Context->dbh;
86 my $result = TransformMarcToKoha( $record, '' );
90 # search duplicate on ISBN, easy and fast..
92 if ( $result->{isbn} ) {
93 $result->{isbn} =~ s/\(.*$//;
94 $result->{isbn} =~ s/\s+$//;
95 $query = "isbn:$result->{isbn}";
99 my $titleindex = 'ti,ext';
100 my $authorindex = 'au,ext';
103 $result->{title} =~ s /\\//g;
104 $result->{title} =~ s /\"//g;
105 $result->{title} =~ s /\(//g;
106 $result->{title} =~ s /\)//g;
108 $query = "$titleindex:\"$result->{title}\"";
109 if ( $result->{author} ) {
110 $result->{author} =~ s /\\//g;
111 $result->{author} =~ s /\"//g;
112 $result->{author} =~ s /\(//g;
113 $result->{author} =~ s /\)//g;
115 $query .= " $op $authorindex:\"$result->{author}\"";
119 my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
120 my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50);
122 if (!defined $error) {
123 foreach my $possible_duplicate_record (@{$searchresults}) {
124 my $marcrecord = new_record_from_zebra(
126 $possible_duplicate_record
129 my $result = TransformMarcToKoha( $marcrecord, '' );
131 # FIXME :: why 2 $biblionumber ?
133 push @results, $result->{'biblionumber'};
134 push @results, $result->{'title'};
143 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
145 This function provides a simple search API on the bibliographic catalog
151 * $query can be a simple keyword or a complete CCL query
152 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
153 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
154 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
155 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
160 Returns an array consisting of three elements
161 * $error is undefined unless an error is detected
162 * $results is a reference to an array of records.
163 * $total_hits is the number of hits that would have been returned with no limit
165 If an error is returned the two other return elements are undefined. If error itself is undefined
166 the other two elements are always defined
168 =item C<usage in the script:>
172 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
174 if (defined $error) {
175 $template->param(query_error => $error);
176 warn "error: ".$error;
177 output_html_with_http_headers $input, $cookie, $template->output;
181 my $hits = @{$marcresults};
184 for my $r ( @{$marcresults} ) {
185 my $marcrecord = MARC::File::USMARC::decode($r);
186 my $biblio = TransformMarcToKoha($marcrecord,q{});
188 #build the iarray of hashs for the template.
190 title => $biblio->{'title'},
191 subtitle => $biblio->{'subtitle'},
192 biblionumber => $biblio->{'biblionumber'},
193 author => $biblio->{'author'},
194 publishercode => $biblio->{'publishercode'},
195 publicationyear => $biblio->{'publicationyear'},
200 $template->param(result=>\@results);
205 my ( $query, $offset, $max_results, $servers, %options ) = @_;
207 return ( 'No query entered', undef, undef ) unless $query;
208 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
209 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
216 # Initialize & Search Zebra
217 for ( my $i = 0 ; $i < @servers ; $i++ ) {
219 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
220 $query =~ s/:/=/g unless $options{skip_normalize};
221 $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]);
222 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
226 $zconns[$i]->errmsg() . " ("
227 . $zconns[$i]->errcode() . ") "
228 . $zconns[$i]->addinfo() . " "
229 . $zconns[$i]->diagset();
231 return ( $error, undef, undef ) if $zconns[$i]->errcode();
235 # caught a ZOOM::Exception
239 . $@->addinfo() . " "
241 warn $error." for query: $query";
242 return ( $error, undef, undef );
251 my $first_record = defined($offset) ? $offset + 1 : 1;
252 my $hits = $tmpresults[ $i - 1 ]->size();
253 $total_hits += $hits;
254 my $last_record = $hits;
255 if ( defined $max_results && $offset + $max_results < $hits ) {
256 $last_record = $offset + $max_results;
259 for my $j ( $first_record .. $last_record ) {
261 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
264 push @{$results}, $record if defined $record;
269 foreach my $zoom_query (@zoom_queries) {
270 $zoom_query->destroy();
273 return ( undef, $results, $total_hits );
278 ( undef, $results_hashref, \@facets_loop ) = getRecords (
280 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
281 $results_per_page, $offset, $branches, $itemtypes,
282 $query_type, $scan, $opac
285 The all singing, all dancing, multi-server, asynchronous, scanning,
286 searching, record nabbing, facet-building
288 See verbose embedded documentation.
294 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
295 $results_per_page, $offset, $branches, $itemtypes,
296 $query_type, $scan, $opac
299 my @servers = @$servers_ref;
300 my @sort_by = @$sort_by_ref;
301 $offset = 0 if $offset < 0;
303 # Initialize variables for the ZOOM connection and results object
306 my $results_hashref = ();
308 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
309 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
311 # Initialize variables for the faceted results objects
312 my $facets_counter = {};
313 my $facets_info = {};
314 my $facets = getFacets();
316 my @facets_loop; # stores the ref to array of hashes for template facets loop
318 ### LOOP THROUGH THE SERVERS
319 for ( my $i = 0 ; $i < @servers ; $i++ ) {
320 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
322 # perform the search, create the results objects
323 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
324 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
326 #$query_to_use = $simple_query if $scan;
327 warn $simple_query if ( $scan and $DEBUG );
329 # Check if we've got a query_type defined, if so, use it
332 if ($query_type =~ /^ccl/) {
333 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
334 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
335 } elsif ($query_type =~ /^cql/) {
336 $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i]));
337 } elsif ($query_type =~ /^pqf/) {
338 $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i]));
340 warn "Unknown query_type '$query_type'. Results undetermined.";
343 $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
345 $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i]));
349 warn "WARNING: query problem with $query_to_use " . $@;
352 # Concatenate the sort_by limits and pass them to the results object
353 # Note: sort will override rank
355 foreach my $sort (@sort_by) {
356 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
357 $sort_by .= "1=1003 <i ";
359 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
360 $sort_by .= "1=1003 >i ";
362 elsif ( $sort eq "popularity_asc" ) {
363 $sort_by .= "1=9003 <i ";
365 elsif ( $sort eq "popularity_dsc" ) {
366 $sort_by .= "1=9003 >i ";
368 elsif ( $sort eq "call_number_asc" ) {
369 $sort_by .= "1=8007 <i ";
371 elsif ( $sort eq "call_number_dsc" ) {
372 $sort_by .= "1=8007 >i ";
374 elsif ( $sort eq "pubdate_asc" ) {
375 $sort_by .= "1=31 <i ";
377 elsif ( $sort eq "pubdate_dsc" ) {
378 $sort_by .= "1=31 >i ";
380 elsif ( $sort eq "acqdate_asc" ) {
381 $sort_by .= "1=32 <i ";
383 elsif ( $sort eq "acqdate_dsc" ) {
384 $sort_by .= "1=32 >i ";
386 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
387 $sort_by .= "1=4 <i ";
389 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
390 $sort_by .= "1=4 >i ";
393 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
396 if ( $sort_by && !$scan && $results[$i] ) {
397 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
398 warn "WARNING sort $sort_by failed";
401 } # finished looping through servers
403 # The big moment: asynchronously retrieve results from all servers
408 my ( $i, $size ) = @_;
411 # loop through the results
412 $results_hash->{'hits'} = $size;
414 if ( $offset + $results_per_page <= $size ) {
415 $times = $offset + $results_per_page;
421 for ( my $j = $offset ; $j < $times ; $j++ ) {
424 ## Check if it's an index scan
426 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
428 # here we create a minimal MARC record and hand it off to the
429 # template just like a normal result ... perhaps not ideal, but
431 my $tmprecord = MARC::Record->new();
432 $tmprecord->encoding('UTF-8');
436 # the minimal record in author/title (depending on MARC flavour)
437 if ( C4::Context->preference("marcflavour") eq
440 $tmptitle = MARC::Field->new(
445 $tmprecord->append_fields($tmptitle);
449 MARC::Field->new( '245', ' ', ' ', a => $term, );
451 MARC::Field->new( '100', ' ', ' ', a => $occ, );
452 $tmprecord->append_fields($tmptitle);
453 $tmprecord->append_fields($tmpauthor);
455 $results_hash->{'RECORDS'}[$j] =
456 $tmprecord->as_usmarc();
461 $record = $results[ $i - 1 ]->record($j)->raw();
462 # warn "RECORD $j:".$record;
463 $results_hash->{'RECORDS'}[$j] = $record;
467 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
469 # Fill the facets while we're looping, but only for the
470 # biblioserver and not for a scan
471 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
472 $facets_counter = GetFacets( $results[ $i - 1 ] );
473 $facets_info = _get_facets_info( $facets );
477 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
479 sort { $a cmp $b } keys %$facets_counter
482 my @this_facets_array;
485 $facets_counter->{$link_value}
486 ->{$b} <=> $facets_counter->{$link_value}
488 } keys %{ $facets_counter->{$link_value} }
491 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
492 my $facet_link_value = $one_facet;
493 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
495 # fix the length that will display in the label,
496 my $facet_label_value = $one_facet;
497 my $facet_max_length = C4::Context->preference(
498 'FacetLabelTruncationLength')
501 substr( $one_facet, 0, $facet_max_length )
503 if length($facet_label_value) >
506 # if it's a branch, label by the name, not the code,
507 if ( $link_value =~ /branch/ ) {
508 if ( defined $branches
509 && ref($branches) eq "HASH"
510 && defined $branches->{$one_facet}
511 && ref( $branches->{$one_facet} ) eq
515 $branches->{$one_facet}
519 $facet_label_value = "*";
523 # if it's a itemtype, label by the name, not the code,
524 if ( $link_value =~ /itype/ ) {
525 if ( defined $itemtypes
526 && ref($itemtypes) eq "HASH"
527 && defined $itemtypes->{$one_facet}
528 && ref( $itemtypes->{$one_facet} ) eq
532 $itemtypes->{$one_facet}
533 ->{translated_description};
537 # also, if it's a location code, use the name instead of the code
538 if ( $link_value =~ /location/ ) {
539 # TODO Retrieve all authorised values at once, instead of 1 query per entry
540 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
541 $facet_label_value = $av->count ? $av->next->opac_description : '';
544 # also, if it's a collection code, use the name instead of the code
545 if ( $link_value =~ /ccode/ ) {
546 # TODO Retrieve all authorised values at once, instead of 1 query per entry
547 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
548 $facet_label_value = $av->count ? $av->next->opac_description : '';
551 # but we're down with the whole label being in the link's title.
552 push @this_facets_array,
555 $facets_counter->{$link_value}
557 facet_label_value => $facet_label_value,
558 facet_title_value => $one_facet,
559 facet_link_value => $facet_link_value,
560 type_link_value => $link_value,
562 if ($facet_label_value);
567 type_link_value => $link_value,
568 type_id => $link_value . "_id",
570 . $facets_info->{$link_value}->{'label_value'} =>
572 facets => \@this_facets_array,
576 $facets_info->{$link_value}->{'label_value'} =~
579 and ( Koha::Libraries->search->count == 1 )
586 # This sorts the facets into alphabetical order
588 foreach my $f (@facets_loop) {
589 if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){
591 [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
596 return ( undef, $results_hashref, \@facets_loop );
604 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
606 if ( $use_zebra_facets ) {
607 $facets = _get_facets_from_zebra( $rs );
609 $facets = _get_facets_from_records( $rs );
615 sub _get_facets_from_records {
619 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
620 my $facets_config = getFacets();
622 my $size = $rs->size();
623 my $jmax = $size > $facets_maxrecs
627 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
629 my $marc_record = new_record_from_zebra (
631 $rs->record( $j )->raw()
634 if ( ! defined $marc_record ) {
635 warn "ERROR DECODING RECORD - $@: " .
636 $rs->record( $j )->raw();
640 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
646 =head2 _get_facets_data_from_record
648 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
650 Internal function that extracts facets information from a MARC::Record object
651 and populates $facets_counter for using in getRecords.
653 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
658 sub _get_facets_data_from_record {
660 my ( $marc_record, $facets, $facets_counter ) = @_;
662 for my $facet (@$facets) {
666 foreach my $tag ( @{ $facet->{ tags } } ) {
668 # tag number is the first three digits
669 my $tag_num = substr( $tag, 0, 3 );
670 # subfields are the remainder
671 my $subfield_letters = substr( $tag, 3 );
673 my @fields = $marc_record->field( $tag_num );
674 foreach my $field (@fields) {
675 # If $field->indicator(1) eq 'z', it means it is a 'see from'
676 # field introduced because of IncludeSeeFromInSearches, so skip it
677 next if $field->indicator(1) eq 'z';
679 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
680 $data =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
682 unless ( grep { $_ eq $data } @used_datas ) {
683 push @used_datas, $data;
684 $facets_counter->{ $facet->{ idx } }->{ $data }++;
691 =head2 _get_facets_from_zebra
693 my $facets = _get_facets_from_zebra( $result_set )
695 Retrieves facets for a specified result set. It loops through the facets defined
696 in C4::Koha::getFacets and returns a hash with the following structure:
706 sub _get_facets_from_zebra {
710 # save current elementSetName
711 my $elementSetName = $rs->option( 'elementSetName' );
713 my $facets_loop = getFacets();
714 my $facets_data = {};
715 # loop through defined facets and fill the facets hashref
716 foreach my $facet ( @$facets_loop ) {
718 my $idx = $facet->{ idx };
719 my $sep = $facet->{ sep };
720 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
721 if ( $facet_values ) {
722 # we've actually got a result
723 $facets_data->{ $idx } = $facet_values;
726 # set elementSetName to its previous value to avoid side effects
727 $rs->option( elementSetName => $elementSetName );
732 =head2 _get_facet_from_result_set
735 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
737 Internal function that extracts facet information for a specific index ($facet_idx) and
738 returns a hash containing facet values and count:
741 $facet_value => $count ,
745 Warning: this function has the side effect of changing the elementSetName for the result
746 set. It is a helper function for the main loop, which takes care of backing it up for
751 sub _get_facet_from_result_set {
753 my $facet_idx = shift;
757 my $internal_sep = '<*>';
758 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
760 return if ( ! defined $facet_idx || ! defined $rs );
761 # zebra's facet element, untokenized index
762 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
763 # configure zebra results for retrieving the desired facet
764 $rs->option( elementSetName => $facet_element );
765 # get the facet record from result set
766 my $facet = $rs->record( 0 )->raw;
767 # if the facet has no restuls...
768 return if !defined $facet;
769 # TODO: benchmark DOM vs. SAX performance
770 my $facet_dom = XML::LibXML->load_xml(
773 my @terms = $facet_dom->getElementsByTagName('term');
777 foreach my $term ( @terms ) {
778 my $facet_value = $term->textContent;
779 $facet_value =~ s/\s*(?<!\p{Uppercase})[.\-,;]*\s*$//;
780 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
781 $facets->{ $facet_value } += $term->getAttribute( 'occur' );
787 =head2 _get_facets_info
789 my $facets_info = C4::Search::_get_facets_info( $facets )
791 Internal function that extracts facets information and properly builds
792 the data structure needed to render facet labels.
796 sub _get_facets_info {
800 my $facets_info = {};
802 for my $facet ( @$facets ) {
803 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
810 sub _detect_truncation {
811 my ( $operand, $index ) = @_;
812 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
815 my @wordlist = split( /\s/, $operand );
816 foreach my $word (@wordlist) {
817 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
818 push @rightlefttruncated, $word;
820 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
821 push @lefttruncated, $word;
823 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
824 push @righttruncated, $word;
826 elsif ( index( $word, "*" ) < 0 ) {
827 push @nontruncated, $word;
830 push @regexpr, $word;
834 \@nontruncated, \@righttruncated, \@lefttruncated,
835 \@rightlefttruncated, \@regexpr
840 sub _build_stemmed_operand {
841 my ($operand,$lang) = @_;
842 require Lingua::Stem::Snowball ;
843 my $stemmed_operand=q{};
845 # Stemmer needs language
846 return $operand unless $lang;
848 # If operand contains a digit, it is almost certainly an identifier, and should
849 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
850 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
851 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
852 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
853 return $operand if $operand =~ /\d/;
855 # FIXME: the locale should be set based on the user's language and/or search choice
857 # Make sure we only use the first two letters from the language code
858 $lang = lc(substr($lang, 0, 2));
859 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
860 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
861 if ($lang eq 'nb' || $lang eq 'nn') {
864 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
865 encoding => "UTF-8" );
867 my @words = split( / /, $operand );
868 my @stems = $stemmer->stem(\@words);
869 for my $stem (@stems) {
870 $stemmed_operand .= "$stem";
871 $stemmed_operand .= "?"
872 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
873 $stemmed_operand .= " ";
875 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
876 return $stemmed_operand;
880 sub _build_weighted_query {
882 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
883 # pretty well but could work much better if we had a smarter query parser
884 my ( $operand, $stemmed_operand, $index ) = @_;
885 my $stemming = C4::Context->preference("QueryStemming") || 0;
886 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
887 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
888 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
890 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
892 # Keyword, or, no index specified
893 if ( ( $index eq 'kw' ) || ( !$index ) ) {
895 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
896 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
897 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
898 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
899 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
900 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
901 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
902 if $fuzzy_enabled; # add fuzzy, word list
903 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
904 if ( $stemming and $stemmed_operand )
905 ; # add stemming, right truncation
906 $weighted_query .= " or wrdl,r9=\"$operand\"";
908 # embedded sorting: 0 a-z; 1 z-a
909 # $weighted_query .= ") or (sort1,aut=1";
912 # Barcode searches should skip this process
913 elsif ( $index eq 'bc' ) {
914 $weighted_query .= "bc=\"$operand\"";
917 # Authority-number searches should skip this process
918 elsif ( $index eq 'an' ) {
919 $weighted_query .= "an=\"$operand\"";
922 # If the index is numeric, don't autoquote it.
923 elsif ( $index =~ /,st-numeric$/ ) {
924 $weighted_query .= " $index=$operand";
927 # If the index already has more than one qualifier, wrap the operand
928 # in quotes and pass it back (assumption is that the user knows what they
929 # are doing and won't appreciate us mucking up their query
930 elsif ( $index =~ ',' ) {
931 $weighted_query .= " $index=\"$operand\"";
934 #TODO: build better cases based on specific search indexes
936 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
937 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
938 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
939 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
940 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
941 if $fuzzy_enabled; # add fuzzy, word list
942 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
943 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
946 $weighted_query .= "))"; # close rank specification
947 return $weighted_query;
952 Return an array with available indexes.
976 'Author-personal-bibliography',
986 'Chronological-subdivision',
996 'Conference-name-heading',
997 'Conference-name-see',
998 'Conference-name-seealso',
1003 'Corporate-name-heading',
1004 'Corporate-name-see',
1005 'Corporate-name-seealso',
1006 'Country-publication',
1009 'date-entered-on-file',
1010 'Date-of-acquisition',
1011 'Date-of-publication',
1012 'Date-time-last-modified',
1013 'Dewey-classification',
1014 'Dissertation-information',
1023 'Geographic-subdivision',
1026 'Heading-use-main-or-added-entry',
1027 'Heading-use-series-added-entry ',
1028 'Heading-use-subject-added-entry',
1032 'Identifier-standard',
1033 'Illustration-code',
1035 'Index-term-uncontrolled',
1036 'Interest-age-level',
1037 'Interest-grade-level',
1047 'language-original',
1057 'Local-classification',
1060 'Match-heading-see-from',
1068 'Name-geographic-heading',
1069 'Name-geographic-see',
1070 'Name-geographic-seealso',
1076 'Other-control-number',
1079 'Personal-name-heading',
1080 'Personal-name-see',
1081 'Personal-name-seealso',
1083 'Place-publication',
1090 'Reading-grade-level',
1091 'Record-control-number',
1102 'Subject-heading-thesaurus',
1103 'Subject-name-personal',
1104 'Subject-subdivision',
1113 'Term-genre-form-heading',
1114 'Term-genre-form-see',
1115 'Term-genre-form-seealso',
1121 'Title-uniform-heading',
1122 'Title-uniform-see',
1123 'Title-uniform-seealso',
1133 'classification-source',
1135 'coded-location-qualifier',
1146 'Local-classification',
1149 'materials-specified',
1154 'Number-local-acquisition',
1159 'replacementpricedate',
1177 $simple_query, $query_cgi,
1178 $query_desc, $limit,
1179 $limit_cgi, $limit_desc,
1180 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1182 Build queries and limits in CCL, CGI, Human,
1183 handle truncation, stemming, field weighting, fuzziness, etc.
1185 See verbose embedded documentation.
1191 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1192 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1197 my @operators = $operators ? @$operators : ();
1198 my @indexes = $indexes ? @$indexes : ();
1199 my @operands = $operands ? @$operands : ();
1200 my @limits = $limits ? @$limits : ();
1201 my @sort_by = $sort_by ? @$sort_by : ();
1203 my $stemming = C4::Context->preference("QueryStemming") || 0;
1204 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1205 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1206 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1208 my $query = $operands[0] // "";
1209 my $simple_query = $operands[0];
1211 # initialize the variables we're passing back
1220 my $cclindexes = getIndexes();
1221 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1222 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1224 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1226 $query = "ccl=$query" if $cclq;
1229 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1231 if ( $query =~ /^ccl=/ ) {
1233 # This is needed otherwise ccl= and &limit won't work together, and
1234 # this happens when selecting a subject on the opac-detail page
1235 @limits = grep {!/^$/} @limits;
1236 my $original_q = $q; # without available part
1237 unless ( grep { $_ eq 'available' } @limits ) {
1238 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1242 if ( grep { $_ eq 'available' } @limits ) {
1243 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1244 @limits = grep {!/^available$/} @limits;
1246 $q .= ' and '.join(' and ', @limits) if @limits;
1248 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1250 if ( $query =~ /^cql=/ ) {
1251 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1253 if ( $query =~ /^pqf=/ ) {
1255 $query_cgi = "q=pqf=".uri_escape_utf8($');
1256 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1259 # pass nested queries directly
1260 # FIXME: need better handling of some of these variables in this case
1261 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1262 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1263 # if ( $query =~ /(\(|\))/ ) {
1265 # undef, $query, $simple_query, $query_cgi,
1266 # $query, $limit, $limit_cgi, $limit_desc,
1271 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1272 # query operands and indexes and add stemming, truncation, field weighting, etc.
1273 # Once we do so, we'll end up with a value in $query, just like if we had an
1274 # incoming $query from the user
1277 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1278 my $previous_operand
1279 ; # a flag used to keep track if there was a previous query
1280 # if there was, we can apply the current operator
1282 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1284 # COMBINE OPERANDS, INDEXES AND OPERATORS
1285 if ( ($operands[$i] // '') ne '' ) {
1286 $operands[$i]=~s/^\s+//;
1288 # A flag to determine whether or not to add the index to the query
1291 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1292 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1296 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1298 my $operand = $operands[$i];
1299 my $index = $indexes[$i] || 'kw';
1301 # Add index-specific attributes
1303 #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search
1304 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1305 #which is processed higher up in this sub. Other than that, year searches are typically
1306 #handled as limits which are not processed her either.
1308 # Search ranges: Date of Publication, st-numeric
1309 if ( $index =~ /(yr|st-numeric)/ ) {
1310 #weight_fields/relevance search causes errors with date ranges
1311 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1312 #In the case of YYYY-YYYY, it will return no results
1313 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1316 # Date of Acquisition
1317 elsif ( $index =~ /acqdate/ ) {
1318 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1319 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1320 #top of the results just because they have lots of item records matching that date.
1321 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1323 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1325 # ISBN,ISSN,Standard Number, don't need special treatment
1326 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1328 $stemming, $auto_truncation,
1329 $weight_fields, $fuzzy_enabled
1332 if ( $index eq 'nb' ) {
1333 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1334 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1335 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1336 $indexes[$i] = $index = 'kw';
1341 # Set default structure attribute (word list)
1342 my $struct_attr = q{};
1343 unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1344 $struct_attr = ",wrdl";
1347 # Some helpful index variants
1348 my $index_plus = $index . $struct_attr . ':';
1349 my $index_plus_comma = $index . $struct_attr . ',';
1351 if ($auto_truncation){
1352 unless ( $index =~ /,(st-|phr|ext)/ ) {
1353 #FIXME only valid with LTR scripts
1354 $operand=join(" ",map{
1355 (index($_,"*")>0?"$_":"$_*")
1356 }split (/\s+/,$operand));
1357 warn $operand if $DEBUG;
1362 my $truncated_operand = q{};
1363 my( $nontruncated, $righttruncated, $lefttruncated,
1364 $rightlefttruncated, $regexpr
1365 ) = _detect_truncation( $operand, $index );
1367 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1372 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1373 scalar(@$rightlefttruncated) > 0 )
1376 # Don't field weight or add the index to the query, we do it here
1378 undef $weight_fields;
1379 my $previous_truncation_operand;
1380 if (scalar @$nontruncated) {
1381 $truncated_operand .= "$index_plus @$nontruncated ";
1382 $previous_truncation_operand = 1;
1384 if (scalar @$righttruncated) {
1385 $truncated_operand .= "and " if $previous_truncation_operand;
1386 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1387 $previous_truncation_operand = 1;
1389 if (scalar @$lefttruncated) {
1390 $truncated_operand .= "and " if $previous_truncation_operand;
1391 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1392 $previous_truncation_operand = 1;
1394 if (scalar @$rightlefttruncated) {
1395 $truncated_operand .= "and " if $previous_truncation_operand;
1396 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1397 $previous_truncation_operand = 1;
1400 $operand = $truncated_operand if $truncated_operand;
1401 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1404 my $stemmed_operand = q{};
1405 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1408 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1410 # Handle Field Weighting
1411 my $weighted_operand = q{};
1412 if ($weight_fields) {
1413 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1414 $operand = $weighted_operand;
1418 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1420 #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own)
1422 #N.B. Truncation is mutually exclusive with Weighted Queries,
1423 #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus
1424 #the need for this relevance wrapper.
1425 $operand = "(rk=($operand))" unless $weight_fields;
1427 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1429 query_cgi => $query_cgi,
1430 query_desc => $query_desc,
1431 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1432 parsed_operand => $operand,
1433 original_operand => $operands[$i] // '',
1435 index_plus => $index_plus,
1436 indexes_set => $indexes_set,
1437 previous_operand => $previous_operand,
1443 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1446 my %group_OR_limits;
1447 my $availability_limit;
1448 foreach my $this_limit (@limits) {
1449 next unless $this_limit;
1450 if ( $this_limit =~ /available/ ) {
1452 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1454 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1455 $availability_limit .=
1456 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1457 $limit_cgi .= "&limit=available";
1461 # group_OR_limits, prefixed by mc-
1462 # OR every member of the group
1463 elsif ( $this_limit =~ /mc/ ) {
1464 my ($k,$v) = split(/:/, $this_limit,2);
1465 if ( $k !~ /mc-i(tem)?type/ ) {
1466 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1467 $this_limit =~ tr/"//d;
1468 $this_limit = $k.':"'.$v.'"';
1471 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1472 $limit_desc .= " or " if $group_OR_limits{$k};
1473 $group_OR_limits{$k} .= "$this_limit";
1474 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1475 $limit_desc .= " $this_limit";
1478 # Regular old limits
1480 $limit .= " and " if $limit || $query;
1481 $limit .= "$this_limit";
1482 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1483 if ($this_limit =~ /^branch:(.+)/) {
1484 my $branchcode = $1;
1485 my $library = Koha::Libraries->find( $branchcode );
1486 if (defined $library) {
1487 $limit_desc .= " branch:" . $library->branchname;
1489 $limit_desc .= " $this_limit";
1492 $limit_desc .= " $this_limit";
1496 foreach my $k (keys (%group_OR_limits)) {
1497 $limit .= " and " if ( $query || $limit );
1498 $limit .= "($group_OR_limits{$k})";
1500 if ($availability_limit) {
1501 $limit .= " and " if ( $query || $limit );
1502 $limit .= "($availability_limit)";
1505 # Normalize the query and limit strings
1506 # This is flawed , means we can't search anything with : in it
1507 # if user wants to do ccl or cql, start the query with that
1508 # $query =~ s/:/=/g;
1509 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1510 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1511 $query =~ s/(?<=(wrdl)):/=/g;
1512 $query =~ s/(?<=(trn|phr)):/=/g;
1513 $query =~ s/(?<=(st-numeric)):/=/g;
1514 $query =~ s/(?<=(st-year)):/=/g;
1515 $query =~ s/(?<=(st-date-normalized)):/=/g;
1517 # Removing warnings for later substitutions
1519 $query_desc //= q{};
1522 $limit_desc //= q{};
1524 $simple_query //= q{};
1526 for ( $query, $query_desc, $limit, $limit_desc ) {
1527 s/ +/ /g; # remove extra spaces
1528 s/^ //g; # remove any beginning spaces
1529 s/ $//g; # remove any ending spaces
1530 s/==/=/g; # remove double == from query
1532 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1534 for ($query_cgi,$simple_query) {
1537 # append the limit to the query
1538 $query .= " " . $limit;
1542 warn "QUERY:" . $query;
1543 warn "QUERY CGI:" . $query_cgi;
1544 warn "QUERY DESC:" . $query_desc;
1545 warn "LIMIT:" . $limit;
1546 warn "LIMIT CGI:" . $limit_cgi;
1547 warn "LIMIT DESC:" . $limit_desc;
1548 warn "---------\nLeave buildQuery\n---------";
1552 undef, $query, $simple_query, $query_cgi,
1553 $query_desc, $limit, $limit_cgi, $limit_desc,
1558 =head2 _build_initial_query
1560 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1562 Build a section of the initial query containing indexes, operators, and operands.
1566 sub _build_initial_query {
1570 if ($params->{previous_operand}){
1571 #If there is a previous operand, add a supplied operator or the default 'and'
1572 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1575 #NOTE: indexes_set is typically set when doing truncation or field weighting
1576 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1578 #e.g. "kw,wrdl:test"
1579 #e.g. " and kw,wrdl:test"
1580 $params->{query} .= $operator . $operand;
1582 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1583 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1584 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1586 #e.g. " and kw,wrdl: test"
1587 $params->{query_desc} .= $operator . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} );
1589 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1591 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1594 =head2 searchResults
1596 my @search_results = searchResults($search_context, $searchdesc, $hits,
1597 $results_per_page, $offset, $scan,
1600 Format results in a form suitable for passing to the template
1604 # IMO this subroutine is pretty messy still -- it's responsible for
1605 # building the HTML output for the template
1607 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_;
1608 my $dbh = C4::Context->dbh;
1613 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1614 my ($is_opac, $hidelostitems);
1615 if ($search_context->{'interface'} eq 'opac') {
1616 $hidelostitems = C4::Context->preference('hidelostitems');
1620 my $record_processor = Koha::RecordProcessor->new({
1621 filters => 'ViewPolicy'
1624 #Build branchnames hash
1625 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1627 # FIXME - We build an authorised values hash here, using the default framework
1628 # though it is possible to have different authvals for different fws.
1630 my $shelflocations =
1631 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1633 # get notforloan authorised value list (see $shelflocations FIXME)
1634 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1635 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1638 my $itemtypes = Koha::ItemTypes->search_with_localization;
1639 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1641 #search item field code
1642 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" );
1644 ## find column names of items related to MARC
1645 my %subfieldstosearch;
1646 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1647 for my $column ( @columns ) {
1648 my ( $tagfield, $tagsubfield ) =
1649 &GetMarcFromKohaField( "items." . $column );
1650 if ( defined $tagsubfield ) {
1651 $subfieldstosearch{$column} = $tagsubfield;
1655 # handle which records to actually retrieve
1656 my $times; # Times is which record to process up to
1657 if ( $hits && $offset + $results_per_page <= $hits ) {
1658 $times = $offset + $results_per_page;
1661 $times = $hits; # If less hits than results_per_page+offset we go to the end
1664 my $marcflavour = C4::Context->preference("marcflavour");
1665 # We get the biblionumber position in MARC
1666 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' );
1668 # set stuff for XSLT processing here once, not later again for every record we retrieved
1672 $xslsyspref = "OPACXSLTResultsDisplay";
1673 $xslfile = C4::Context->preference( $xslsyspref );
1675 $xslsyspref = "XSLTResultsDisplay";
1676 $xslfile = C4::Context->preference( $xslsyspref ) || "default";
1678 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1679 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1681 my $userenv = C4::Context->userenv;
1683 = ( defined $userenv and $userenv->{number} )
1684 ? Koha::Patrons->find( $userenv->{number} )
1686 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1688 # loop through all of the records we've retrieved
1689 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1693 # For Scan searches we built USMARC data
1694 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1696 # Normal search, render from Zebra's output
1697 $marcrecord = new_record_from_zebra(
1702 if ( ! defined $marcrecord ) {
1703 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1711 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1712 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1714 SetUTF8Flag($marcrecord);
1715 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw, 'no_items' );
1716 $oldbiblio->{result_number} = $i + 1;
1718 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1719 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1720 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1721 $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one
1722 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1724 # edition information, if any
1725 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1727 my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef;
1728 # add imageurl to itemtype if there is one
1729 $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{};
1730 # Build summary if there is one (the summary is defined in the itemtypes table)
1731 $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{};
1733 # FIXME: this is only used in the deprecated non-XLST opac results
1734 if ( !$xslfile && $is_opac && $itemtype && $itemtype->{summary} ) {
1735 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1736 my @fields = $marcrecord->fields();
1739 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1741 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1742 $tag =~ /(.{3})(.)/;
1743 if($marcrecord->field($1)){
1744 my @abc = $marcrecord->field($1)->subfield($2);
1745 $tags->{$tag} = $#abc + 1 ;
1749 # We catch how many times to repeat this line
1751 foreach my $tag (keys(%$tags)){
1752 $max = $tags->{$tag} if($tags->{$tag} > $max);
1755 # we replace, and repeat each line
1756 for (my $i = 0 ; $i < $max ; $i++){
1757 my $newline = $line;
1759 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1760 $tag =~ /(.{3})(.)/;
1762 if($marcrecord->field($1)){
1763 my @repl = $marcrecord->field($1)->subfield($2);
1764 my $subfieldvalue = $repl[$i];
1765 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1768 $newsummary .= "$newline\n";
1772 $newsummary =~ s/\[(.*?)]//g;
1773 $newsummary =~ s/\n/<br\/>/g;
1774 $oldbiblio->{summary} = $newsummary;
1777 # Pull out the items fields
1778 my @fields = $marcrecord->field($itemtag);
1779 $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT');
1780 my $marcflavor = C4::Context->preference("marcflavour");
1782 # adding linked items that belong to host records
1783 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
1784 my $analyticsfield = '773';
1785 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1786 $analyticsfield = '773';
1787 } elsif ($marcflavor eq 'UNIMARC') {
1788 $analyticsfield = '461';
1790 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1791 my $hostbiblionumber = $hostfield->subfield("0");
1792 my $linkeditemnumber = $hostfield->subfield("9");
1793 if( $hostbiblionumber ) {
1794 my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber );
1795 if ($linkeditemmarc) {
1796 my $linkeditemfield = $linkeditemmarc->field($itemtag);
1797 if ($linkeditemfield) {
1798 push( @fields, $linkeditemfield );
1805 # Setting item statuses for display
1806 my @available_items_loop;
1807 my @onloan_items_loop;
1808 my @other_items_loop;
1810 my $available_items;
1814 my $ordered_count = 0;
1815 my $available_count = 0;
1816 my $onloan_count = 0;
1817 my $longoverdue_count = 0;
1818 my $other_count = 0;
1819 my $withdrawn_count = 0;
1820 my $itemlost_count = 0;
1821 my $hideatopac_count = 0;
1822 my $itembinding_count = 0;
1823 my $itemdamaged_count = 0;
1824 my $item_in_transit_count = 0;
1825 my $can_place_holds = 0;
1826 my $item_onhold_count = 0;
1827 my $notforloan_count = 0;
1828 my $items_count = scalar(@fields);
1829 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1830 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1831 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1833 # loop through every item
1834 foreach my $field (@fields) {
1837 # populate the items hash
1838 foreach my $code ( keys %subfieldstosearch ) {
1839 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1841 $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype};
1845 # hidden because lost
1846 if ($hidelostitems && $item->{itemlost}) {
1847 $hideatopac_count++;
1850 # hidden based on OpacHiddenItems syspref
1851 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
1853 push @hiddenitems, @hi;
1854 $hideatopac_count++;
1859 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
1860 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1862 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1863 if ($item->{$hbranch}) {
1864 $item->{'branchname'} = $branches{$item->{$hbranch}};
1866 elsif ($item->{$otherbranch}) { # Last resort
1867 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1871 ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} )
1872 . ( $item->{location} ? $item->{location} : q{} )
1873 . ( $item->{itype} ? $item->{itype} : q{} )
1874 . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} );
1875 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1876 if ( $item->{onloan}
1878 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
1881 my $key = $prefix . $item->{onloan} . $item->{barcode};
1882 $onloan_items->{$key}->{due_date} = $item->{onloan};
1883 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1884 $onloan_items->{$key}->{branchname} = $item->{branchname};
1885 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1886 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1887 $onloan_items->{$key}->{description} = $item->{description};
1888 $onloan_items->{$key}->{imageurl} =
1889 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
1891 # if something's checked out and lost, mark it as 'long overdue'
1892 if ( $item->{itemlost} ) {
1893 $onloan_items->{$key}->{longoverdue}++;
1894 $longoverdue_count++;
1896 else { # can place holds as long as item isn't lost
1897 $can_place_holds = 1;
1901 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1904 my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype};
1905 $item->{notforloan} = 1 if !$item->{notforloan} &&
1906 $itemtype && $itemtypes{ $itemtype }->{notforloan};
1909 if ( $item->{notforloan} < 0 ) {
1911 } elsif ( $item->{notforloan} > 0 ) {
1912 $notforloan_count++;
1915 # is item in transit?
1916 my $transfertwhen = '';
1917 my ($transfertfrom, $transfertto);
1919 # is item on the reserve shelf?
1920 my $reservestatus = '';
1922 unless ($item->{withdrawn}
1923 || $item->{itemlost}
1925 || $item->{notforloan}
1926 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
1927 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
1929 # A couple heuristics to limit how many times
1930 # we query the database for item transfer information, sacrificing
1931 # accuracy in some cases for speed;
1933 # 1. don't query if item has one of the other statuses
1934 # 2. don't check transit status if the bib has
1935 # more than 20 items
1937 # FIXME: to avoid having the query the database like this, and to make
1938 # the in transit status count as unavailable for search limiting,
1939 # should map transit status to record indexed in Zebra.
1941 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1942 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
1945 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1946 if ( $item->{withdrawn}
1947 || $item->{itemlost}
1949 || $item->{notforloan}
1950 || $reservestatus eq 'Waiting'
1951 || ($transfertwhen && $transfertwhen ne ''))
1953 $withdrawn_count++ if $item->{withdrawn};
1954 $itemlost_count++ if $item->{itemlost};
1955 $itemdamaged_count++ if $item->{damaged};
1956 $item_in_transit_count++ if $transfertwhen && $transfertwhen ne '';
1957 $item_onhold_count++ if $reservestatus eq 'Waiting';
1958 $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{});
1960 # can place a hold on a item if
1961 # not lost nor withdrawn
1962 # not damaged unless AllowHoldsOnDamagedItems is true
1963 # item is either for loan or on order (notforloan < 0)
1964 $can_place_holds = 1
1967 && !$item->{withdrawn}
1968 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
1969 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
1974 my $key = $prefix . $item->{status};
1975 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
1976 $other_items->{$key}->{$_} = $item->{$_};
1978 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1979 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1980 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1981 $other_items->{$key}->{count}++ if $item->{$hbranch};
1982 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1983 $other_items->{$key}->{description} = $item->{description};
1984 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1988 $can_place_holds = 1;
1990 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1991 foreach (qw(branchname itemcallnumber description)) {
1992 $available_items->{$prefix}->{$_} = $item->{$_};
1994 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location};
1995 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} );
1998 } # notforloan, item level and biblioitem level
2000 # if all items are hidden, do not show the record
2001 if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) {
2005 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2006 for my $key ( sort keys %$onloan_items ) {
2007 (++$onloanitemscount > $maxitems) and last;
2008 push @onloan_items_loop, $onloan_items->{$key};
2010 for my $key ( sort keys %$other_items ) {
2011 (++$otheritemscount > $maxitems) and last;
2012 push @other_items_loop, $other_items->{$key};
2014 for my $key ( sort keys %$available_items ) {
2015 (++$availableitemscount > $maxitems) and last;
2016 push @available_items_loop, $available_items->{$key}
2019 # XSLT processing of some stuff
2020 # we fetched the sysprefs already before the loop through all retrieved record!
2021 if (!$scan && $xslfile) {
2022 $record_processor->options({
2023 frameworkcode => $fw,
2024 interface => $search_context->{'interface'}
2027 $record_processor->process($marcrecord);
2028 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang, $xslt_variables);
2031 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2032 if (!C4::Context->preference("item-level_itypes")) {
2033 if ($itemtype && $itemtype->{notforloan}) {
2034 $can_place_holds = 0;
2037 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2038 $oldbiblio->{items_count} = $items_count;
2039 $oldbiblio->{available_items_loop} = \@available_items_loop;
2040 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2041 $oldbiblio->{other_items_loop} = \@other_items_loop;
2042 $oldbiblio->{availablecount} = $available_count;
2043 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2044 $oldbiblio->{onloancount} = $onloan_count;
2045 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2046 $oldbiblio->{othercount} = $other_count;
2047 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2048 $oldbiblio->{withdrawncount} = $withdrawn_count;
2049 $oldbiblio->{itemlostcount} = $itemlost_count;
2050 $oldbiblio->{damagedcount} = $itemdamaged_count;
2051 $oldbiblio->{intransitcount} = $item_in_transit_count;
2052 $oldbiblio->{onholdcount} = $item_onhold_count;
2053 $oldbiblio->{orderedcount} = $ordered_count;
2054 $oldbiblio->{notforloancount} = $notforloan_count;
2056 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2057 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2058 my $subfields = substr $fieldspec, 3;
2059 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2060 my @alternateholdingsinfo = ();
2061 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2062 my $alternateholdingscount = 0;
2064 for my $field (@holdingsfields) {
2065 my %holding = ( holding => '' );
2066 my $havesubfield = 0;
2067 for my $subfield ($field->subfields()) {
2068 if ((index $subfields, $$subfield[0]) >= 0) {
2069 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2070 $holding{'holding'} .= $$subfield[1];
2074 if ($havesubfield) {
2075 push(@alternateholdingsinfo, \%holding);
2076 $alternateholdingscount++;
2080 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2081 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2084 $oldbiblio->{biblio_object} = Koha::Biblios->find( $oldbiblio->{biblionumber} );
2086 push( @newresults, $oldbiblio );
2092 =head2 enabled_staff_search_views
2094 %hash = enabled_staff_search_views()
2096 This function returns a hash that contains three flags obtained from the system
2097 preferences, used to determine whether a particular staff search results view
2102 =item C<Output arg:>
2104 * $hash{can_view_MARC} is true only if the MARC view is enabled
2105 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2106 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2108 =item C<usage in the script:>
2112 $template->param ( C4::Search::enabled_staff_search_views );
2116 sub enabled_staff_search_views
2119 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2120 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2121 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2125 =head2 z3950_search_args
2127 $arrayref = z3950_search_args($matchpoints)
2129 This function returns an array reference that contains the search parameters to be
2130 passed to the Z39.50 search script (z3950_search.pl). The array elements
2131 are hash refs whose keys are name and value, and whose values are the
2132 name of a search parameter, the value of that search parameter and the URL encoded
2133 value of that parameter.
2135 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2137 The search parameter values are obtained from the bibliographic record whose
2138 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2140 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2141 a general purpose search argument. In this case, the returned array contains only
2142 entry: the key is 'title' and the value is derived from $matchpoints.
2144 If a search parameter value is undefined or empty, it is not included in the returned
2147 The returned array reference may be passed directly to the template parameters.
2151 =item C<Output arg:>
2153 * $array containing hash refs as described above
2155 =item C<usage in the script:>
2159 $data = Biblio::GetBiblioData($bibno);
2160 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2164 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2168 sub z3950_search_args {
2171 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2172 my $isbn = Business::ISBN->new( $isbn_string );
2174 if (defined $isbn && $isbn->is_valid)
2176 if ( ref($bibrec) ) {
2177 $bibrec->{isbn} = $isbn_string;
2178 $bibrec->{title} = undef;
2180 $bibrec = { isbn => $isbn_string };
2184 $bibrec = { title => $bibrec } if !ref $bibrec;
2187 for my $field (qw/ lccn isbn issn title author dewey subject /)
2189 push @$array, { name => $field, value => $bibrec->{$field} }
2190 if defined $bibrec->{$field};
2195 =head2 GetDistinctValues($field);
2197 C<$field> is a reference to the fields array
2201 sub GetDistinctValues {
2202 my ($fieldname,$string)=@_;
2203 # returns a reference to a hash of references to branches...
2204 if ($fieldname=~/\./){
2205 my ($table,$column)=split /\./, $fieldname;
2206 my $dbh = C4::Context->dbh;
2207 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2208 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 ");
2210 my $elements=$sth->fetchall_arrayref({});
2215 my @servers=qw<biblioserver authorityserver>;
2216 my (@zconns,@results);
2217 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2218 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2221 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2224 # The big moment: asynchronously retrieve results from all servers
2230 my ( $i, $size ) = @_;
2231 for ( my $j = 0 ; $j < $size ; $j++ ) {
2233 @hashscan{qw(value cnt)} =
2234 $results[ $i - 1 ]->display_term($j);
2235 push @elements, \%hashscan;
2243 =head2 _ZOOM_event_loop
2245 _ZOOM_event_loop(\@zconns, \@results, sub {
2246 my ( $i, $size ) = @_;
2250 Processes a ZOOM event loop and passes control to a closure for
2251 processing the results, and destroying the resultsets.
2255 sub _ZOOM_event_loop {
2256 my ($zconns, $results, $callback) = @_;
2257 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2258 my $ev = $zconns->[ $i - 1 ]->last_event();
2259 if ( $ev == ZOOM::Event::ZEND ) {
2260 next unless $results->[ $i - 1 ];
2261 my $size = $results->[ $i - 1 ]->size();
2263 $callback->($i, $size);
2268 foreach my $result (@$results) {
2273 =head2 new_record_from_zebra
2275 Given raw data from a searchengine result set, return a MARC::Record object
2277 This helper function is needed to take into account all the involved
2278 system preferences and configuration variables to properly create the
2279 MARC::Record object.
2281 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2282 data. If we are using DOM, then it has to be MARCXML.
2284 If we are using elasticsearch, it'll already be a MARC::Record and this
2285 function needs a new name.
2289 sub new_record_from_zebra {
2292 my $raw_data = shift;
2293 # Set the default indexing modes
2294 my $search_engine = C4::Context->preference("SearchEngine");
2295 if ($search_engine eq 'Elasticsearch') {
2296 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2298 my $index_mode = ( $server eq 'biblioserver' )
2299 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2300 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2302 my $marc_record = eval {
2303 if ( $index_mode eq 'dom' ) {
2304 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2306 MARC::Record->new_from_usmarc( $raw_data );
2313 return $marc_record;
2318 END { } # module clean-up code here (global destructor)
2325 Koha Development Team <http://koha-community.org/>