3 # This file is part of Koha.
5 # Koha is free software; you can redistribute it and/or modify it under the
6 # terms of the GNU General Public License as published by the Free Software
7 # Foundation; either version 2 of the License, or (at your option) any later
10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
14 # You should have received a copy of the GNU General Public License along with
15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
16 # Suite 330, Boston, MA 02111-1307 USA
19 #use warnings; FIXME - Bug 2505
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
25 use C4::Search::PazPar2;
27 use C4::Dates qw(format_date);
28 use C4::Members qw(GetHideLostItemsPreference);
31 use C4::Reserves; # GetReserveStatus
40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
42 # set the version for version checking
44 $VERSION = 3.07.00.049;
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
74 &enabled_staff_search_views
78 # make all your functions, whether exported or not;
82 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
84 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
90 my $dbh = C4::Context->dbh;
91 my $result = TransformMarcToKoha( $dbh, $record, '' );
96 my ( $biblionumber, $title );
98 # search duplicate on ISBN, easy and fast..
100 if ( $result->{isbn} ) {
101 $result->{isbn} =~ s/\(.*$//;
102 $result->{isbn} =~ s/\s+$//;
103 $query = "isbn:$result->{isbn}";
107 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
113 $titleindex = 'title|exact';
114 $authorindex = 'author|exact';
117 $titleindex = 'ti,ext';
118 $authorindex = 'au,ext';
122 $result->{title} =~ s /\\//g;
123 $result->{title} =~ s /\"//g;
124 $result->{title} =~ s /\(//g;
125 $result->{title} =~ s /\)//g;
127 # FIXME: instead of removing operators, could just do
128 # quotes around the value
129 $result->{title} =~ s/(and|or|not)//g;
130 $query = "$titleindex:\"$result->{title}\"";
131 if ( $result->{author} ) {
132 $result->{author} =~ s /\\//g;
133 $result->{author} =~ s /\"//g;
134 $result->{author} =~ s /\(//g;
135 $result->{author} =~ s /\)//g;
137 # remove valid operators
138 $result->{author} =~ s/(and|or|not)//g;
139 $query .= " $op $authorindex:\"$result->{author}\"";
143 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
145 if (!defined $error) {
146 foreach my $possible_duplicate_record (@{$searchresults}) {
148 MARC::Record->new_from_usmarc($possible_duplicate_record);
149 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
151 # FIXME :: why 2 $biblionumber ?
153 push @results, $result->{'biblionumber'};
154 push @results, $result->{'title'};
163 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
165 This function provides a simple search API on the bibliographic catalog
171 * $query can be a simple keyword or a complete CCL query
172 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
173 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
174 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
179 Returns an array consisting of three elements
180 * $error is undefined unless an error is detected
181 * $results is a reference to an array of records.
182 * $total_hits is the number of hits that would have been returned with no limit
184 If an error is returned the two other return elements are undefined. If error itself is undefined
185 the other two elements are always defined
187 =item C<usage in the script:>
191 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
193 if (defined $error) {
194 $template->param(query_error => $error);
195 warn "error: ".$error;
196 output_html_with_http_headers $input, $cookie, $template->output;
200 my $hits = @{$marcresults};
203 for my $r ( @{$marcresults} ) {
204 my $marcrecord = MARC::File::USMARC::decode($r);
205 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
207 #build the iarray of hashs for the template.
209 title => $biblio->{'title'},
210 subtitle => $biblio->{'subtitle'},
211 biblionumber => $biblio->{'biblionumber'},
212 author => $biblio->{'author'},
213 publishercode => $biblio->{'publishercode'},
214 publicationyear => $biblio->{'publicationyear'},
219 $template->param(result=>\@results);
224 my ( $query, $offset, $max_results, $servers ) = @_;
226 if ( C4::Context->preference('NoZebra') ) {
227 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
230 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
231 return ( undef, $search_result, scalar($result->{hits}) );
234 return ( 'No query entered', undef, undef ) unless $query;
235 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
236 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
244 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
246 # Initialize & Search Zebra
247 for ( my $i = 0 ; $i < @servers ; $i++ ) {
249 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
252 $QParser->parse( $query );
253 $query = $QParser->target_syntax($servers[$i]);
254 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
256 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
258 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
262 $zconns[$i]->errmsg() . " ("
263 . $zconns[$i]->errcode() . ") "
264 . $zconns[$i]->addinfo() . " "
265 . $zconns[$i]->diagset();
267 return ( $error, undef, undef ) if $zconns[$i]->errcode();
271 # caught a ZOOM::Exception
275 . $@->addinfo() . " "
277 warn $error." for query: $query";
278 return ( $error, undef, undef );
287 my $first_record = defined($offset) ? $offset + 1 : 1;
288 my $hits = $tmpresults[ $i - 1 ]->size();
289 $total_hits += $hits;
290 my $last_record = $hits;
291 if ( defined $max_results && $offset + $max_results < $hits ) {
292 $last_record = $offset + $max_results;
295 for my $j ( $first_record .. $last_record ) {
297 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
299 push @{$results}, $record;
304 foreach my $zoom_query (@zoom_queries) {
305 $zoom_query->destroy();
308 return ( undef, $results, $total_hits );
314 ( undef, $results_hashref, \@facets_loop ) = getRecords (
316 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
317 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
321 The all singing, all dancing, multi-server, asynchronous, scanning,
322 searching, record nabbing, facet-building
324 See verbse embedded documentation.
330 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
331 $results_per_page, $offset, $expanded_facet, $branches,
332 $itemtypes, $query_type, $scan, $opac
335 my @servers = @$servers_ref;
336 my @sort_by = @$sort_by_ref;
338 # Initialize variables for the ZOOM connection and results object
342 my $results_hashref = ();
344 # Initialize variables for the faceted results objects
345 my $facets_counter = ();
346 my $facets_info = ();
347 my $facets = getFacets();
348 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
350 my @facets_loop; # stores the ref to array of hashes for template facets loop
352 ### LOOP THROUGH THE SERVERS
353 for ( my $i = 0 ; $i < @servers ; $i++ ) {
354 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
356 # perform the search, create the results objects
357 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
358 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
360 #$query_to_use = $simple_query if $scan;
361 warn $simple_query if ( $scan and $DEBUG );
363 # Check if we've got a query_type defined, if so, use it
366 if ($query_type =~ /^ccl/) {
367 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
368 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
369 } elsif ($query_type =~ /^cql/) {
370 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
371 } elsif ($query_type =~ /^pqf/) {
372 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
374 warn "Unknown query_type '$query_type'. Results undetermined.";
377 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
379 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
383 warn "WARNING: query problem with $query_to_use " . $@;
386 # Concatenate the sort_by limits and pass them to the results object
387 # Note: sort will override rank
389 foreach my $sort (@sort_by) {
390 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
391 $sort_by .= "1=1003 <i ";
393 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
394 $sort_by .= "1=1003 >i ";
396 elsif ( $sort eq "popularity_asc" ) {
397 $sort_by .= "1=9003 <i ";
399 elsif ( $sort eq "popularity_dsc" ) {
400 $sort_by .= "1=9003 >i ";
402 elsif ( $sort eq "call_number_asc" ) {
403 $sort_by .= "1=8007 <i ";
405 elsif ( $sort eq "call_number_dsc" ) {
406 $sort_by .= "1=8007 >i ";
408 elsif ( $sort eq "pubdate_asc" ) {
409 $sort_by .= "1=31 <i ";
411 elsif ( $sort eq "pubdate_dsc" ) {
412 $sort_by .= "1=31 >i ";
414 elsif ( $sort eq "acqdate_asc" ) {
415 $sort_by .= "1=32 <i ";
417 elsif ( $sort eq "acqdate_dsc" ) {
418 $sort_by .= "1=32 >i ";
420 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
421 $sort_by .= "1=4 <i ";
423 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
424 $sort_by .= "1=4 >i ";
427 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
430 if ($sort_by && !$scan) {
431 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
432 warn "WARNING sort $sort_by failed";
435 } # finished looping through servers
437 # The big moment: asynchronously retrieve results from all servers
442 my ( $i, $size ) = @_;
445 # loop through the results
446 $results_hash->{'hits'} = $size;
448 if ( $offset + $results_per_page <= $size ) {
449 $times = $offset + $results_per_page;
454 for ( my $j = $offset ; $j < $times ; $j++ ) {
458 ## Check if it's an index scan
460 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
462 # here we create a minimal MARC record and hand it off to the
463 # template just like a normal result ... perhaps not ideal, but
465 my $tmprecord = MARC::Record->new();
466 $tmprecord->encoding('UTF-8');
470 # the minimal record in author/title (depending on MARC flavour)
471 if ( C4::Context->preference("marcflavour") eq
474 $tmptitle = MARC::Field->new(
479 $tmprecord->append_fields($tmptitle);
483 MARC::Field->new( '245', ' ', ' ', a => $term, );
485 MARC::Field->new( '100', ' ', ' ', a => $occ, );
486 $tmprecord->append_fields($tmptitle);
487 $tmprecord->append_fields($tmpauthor);
489 $results_hash->{'RECORDS'}[$j] =
490 $tmprecord->as_usmarc();
495 $record = $results[ $i - 1 ]->record($j)->raw();
497 # warn "RECORD $j:".$record;
498 $results_hash->{'RECORDS'}[$j] = $record;
502 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
504 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
505 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
508 $size > $facets_maxrecs ? $facets_maxrecs : $size;
509 for my $facet (@$facets) {
510 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
512 $results[ $i - 1 ]->record($j)->render();
514 foreach my $tag ( @{ $facet->{tags} } ) {
517 my $tag_num = substr( $tag, 0, 3 );
518 my $letters = substr( $tag, 3 );
520 '\n' . $tag_num . ' ([^z][^\n]+)';
521 $field_pattern = '\n' . $tag_num . ' ([^\n]+)'
522 if ( int($tag_num) < 10 );
524 ( $render_record =~ /$field_pattern/g );
525 foreach my $field_token (@field_tokens) {
526 my @subf = ( $field_token =~
527 /\$([a-zA-Z0-9]) ([^\$]+)/g );
529 for ( my $i = 0 ; $i < @subf ; $i += 2 ) {
530 if ( $letters =~ $subf[$i] ) {
531 my $value = $subf[ $i + 1 ];
534 push @values, $value;
537 my $data = join( $facet->{sep}, @values );
538 unless ( $data ~~ @used_datas ) {
539 $facets_counter->{ $facet->{idx} }
541 push @used_datas, $data;
546 $facets_info->{ $facet->{idx} }->{label_value} =
548 $facets_info->{ $facet->{idx} }->{expanded} =
553 # warn "connection ", $i-1, ": $size hits";
554 # warn $results[$i-1]->record(0)->render() if $size > 0;
557 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
559 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
560 keys %$facets_counter
564 my $number_of_facets;
565 my @this_facets_array;
568 $facets_counter->{$link_value}
569 ->{$b} <=> $facets_counter->{$link_value}
571 } keys %{ $facets_counter->{$link_value} }
575 if ( ( $number_of_facets < 6 )
576 || ( $expanded_facet eq $link_value )
577 || ( $facets_info->{$link_value}->{'expanded'} )
581 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
582 my $facet_link_value = $one_facet;
583 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
585 # fix the length that will display in the label,
586 my $facet_label_value = $one_facet;
587 my $facet_max_length = C4::Context->preference(
588 'FacetLabelTruncationLength')
591 substr( $one_facet, 0, $facet_max_length )
593 if length($facet_label_value) >
596 # if it's a branch, label by the name, not the code,
597 if ( $link_value =~ /branch/ ) {
598 if ( defined $branches
599 && ref($branches) eq "HASH"
600 && defined $branches->{$one_facet}
601 && ref( $branches->{$one_facet} ) eq
605 $branches->{$one_facet}
609 $facet_label_value = "*";
613 # if it's a itemtype, label by the name, not the code,
614 if ( $link_value =~ /itype/ ) {
615 if ( defined $itemtypes
616 && ref($itemtypes) eq "HASH"
617 && defined $itemtypes->{$one_facet}
618 && ref( $itemtypes->{$one_facet} ) eq
622 $itemtypes->{$one_facet}
627 # also, if it's a location code, use the name instead of the code
628 if ( $link_value =~ /location/ ) {
630 GetKohaAuthorisedValueLib( 'LOC',
634 # but we're down with the whole label being in the link's title.
635 push @this_facets_array,
638 $facets_counter->{$link_value}
640 facet_label_value => $facet_label_value,
641 facet_title_value => $one_facet,
642 facet_link_value => $facet_link_value,
643 type_link_value => $link_value,
645 if ($facet_label_value);
649 # handle expanded option
650 unless ( $facets_info->{$link_value}->{'expanded'} ) {
652 if ( ( $number_of_facets > 6 )
653 && ( $expanded_facet ne $link_value ) );
657 type_link_value => $link_value,
658 type_id => $link_value . "_id",
660 . $facets_info->{$link_value}->{'label_value'} =>
662 facets => \@this_facets_array,
663 expandable => $expandable,
664 expand => $link_value,
668 $facets_info->{$link_value}->{'label_value'} =~
671 and ( C4::Context->preference('singleBranchMode') )
677 return ( undef, $results_hashref, \@facets_loop );
682 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
683 $results_per_page, $offset, $expanded_facet, $branches,
687 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
689 $paz->search($simple_query);
690 sleep 1; # FIXME: WHY?
693 my $results_hashref = {};
694 my $stats = XMLin($paz->stat);
695 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
697 # for a grouped search result, the number of hits
698 # is the number of groups returned; 'bib_hits' will have
699 # the total number of bibs.
700 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
701 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
703 HIT: foreach my $hit (@{ $results->{'hit'} }) {
704 my $recid = $hit->{recid}->[0];
706 my $work_title = $hit->{'md-work-title'}->[0];
708 if (exists $hit->{'md-work-author'}) {
709 $work_author = $hit->{'md-work-author'}->[0];
711 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
713 my $result_group = {};
714 $result_group->{'group_label'} = $group_label;
715 $result_group->{'group_merge_key'} = $recid;
718 if (exists $hit->{count}) {
719 $count = $hit->{count}->[0];
721 $result_group->{'group_count'} = $count;
723 for (my $i = 0; $i < $count; $i++) {
724 # FIXME -- may need to worry about diacritics here
725 my $rec = $paz->record($recid, $i);
726 push @{ $result_group->{'RECORDS'} }, $rec;
729 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
732 # pass through facets
733 my $termlist_xml = $paz->termlist('author,subject');
734 my $terms = XMLin($termlist_xml, forcearray => 1);
735 my @facets_loop = ();
736 #die Dumper($results);
737 # foreach my $list (sort keys %{ $terms->{'list'} }) {
739 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
741 # facet_label_value => $facet->{'name'}->[0],
744 # push @facets_loop, ( {
745 # type_label => $list,
746 # facets => \@facets,
750 return ( undef, $results_hashref, \@facets_loop );
754 sub _remove_stopwords {
755 my ( $operand, $index ) = @_;
756 my @stopwords_removed;
758 # phrase and exact-qualified indexes shouldn't have stopwords removed
759 if ( $index !~ m/phr|ext/ ) {
761 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
762 # we use IsAlpha unicode definition, to deal correctly with diacritics.
763 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
764 # is a stopword, we'd get "çon" and wouldn't find anything...
766 foreach ( keys %{ C4::Context->stopwords } ) {
767 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
768 if ( my ($matched) = ($operand =~
769 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
771 $operand =~ s/\Q$matched\E/ /gi;
772 push @stopwords_removed, $_;
776 return ( $operand, \@stopwords_removed );
780 sub _detect_truncation {
781 my ( $operand, $index ) = @_;
782 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
785 my @wordlist = split( /\s/, $operand );
786 foreach my $word (@wordlist) {
787 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
788 push @rightlefttruncated, $word;
790 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
791 push @lefttruncated, $word;
793 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
794 push @righttruncated, $word;
796 elsif ( index( $word, "*" ) < 0 ) {
797 push @nontruncated, $word;
800 push @regexpr, $word;
804 \@nontruncated, \@righttruncated, \@lefttruncated,
805 \@rightlefttruncated, \@regexpr
810 sub _build_stemmed_operand {
811 my ($operand,$lang) = @_;
812 require Lingua::Stem::Snowball ;
813 my $stemmed_operand=q{};
815 # If operand contains a digit, it is almost certainly an identifier, and should
816 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
817 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
818 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
819 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
820 return $operand if $operand =~ /\d/;
822 # FIXME: the locale should be set based on the user's language and/or search choice
824 # Make sure we only use the first two letters from the language code
825 $lang = lc(substr($lang, 0, 2));
826 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
827 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
828 if ($lang eq 'nb' || $lang eq 'nn') {
831 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
832 encoding => "UTF-8" );
834 my @words = split( / /, $operand );
835 my @stems = $stemmer->stem(\@words);
836 for my $stem (@stems) {
837 $stemmed_operand .= "$stem";
838 $stemmed_operand .= "?"
839 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
840 $stemmed_operand .= " ";
842 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
843 return $stemmed_operand;
847 sub _build_weighted_query {
849 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
850 # pretty well but could work much better if we had a smarter query parser
851 my ( $operand, $stemmed_operand, $index ) = @_;
852 my $stemming = C4::Context->preference("QueryStemming") || 0;
853 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
854 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
856 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
858 # Keyword, or, no index specified
859 if ( ( $index eq 'kw' ) || ( !$index ) ) {
861 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
862 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
863 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
864 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
865 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
866 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
867 if $fuzzy_enabled; # add fuzzy, word list
868 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
869 if ( $stemming and $stemmed_operand )
870 ; # add stemming, right truncation
871 $weighted_query .= " or wrdl,r9=\"$operand\"";
873 # embedded sorting: 0 a-z; 1 z-a
874 # $weighted_query .= ") or (sort1,aut=1";
877 # Barcode searches should skip this process
878 elsif ( $index eq 'bc' ) {
879 $weighted_query .= "bc=\"$operand\"";
882 # Authority-number searches should skip this process
883 elsif ( $index eq 'an' ) {
884 $weighted_query .= "an=\"$operand\"";
887 # If the index already has more than one qualifier, wrap the operand
888 # in quotes and pass it back (assumption is that the user knows what they
889 # are doing and won't appreciate us mucking up their query
890 elsif ( $index =~ ',' ) {
891 $weighted_query .= " $index=\"$operand\"";
894 #TODO: build better cases based on specific search indexes
896 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
897 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
898 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
900 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
903 $weighted_query .= "))"; # close rank specification
904 return $weighted_query;
909 Return an array with available indexes.
931 'Author-personal-bibliography',
941 'Chronological-subdivision',
951 'Conference-name-heading',
952 'Conference-name-see',
953 'Conference-name-seealso',
958 'Corporate-name-heading',
959 'Corporate-name-see',
960 'Corporate-name-seealso',
962 'date-entered-on-file',
963 'Date-of-acquisition',
964 'Date-of-publication',
965 'Dewey-classification',
972 'Geographic-subdivision',
975 'Heading-use-main-or-added-entry',
976 'Heading-use-series-added-entry ',
977 'Heading-use-subject-added-entry',
995 'Local-classification',
998 'Match-heading-see-from',
1006 'Name-geographic-heading',
1007 'Name-geographic-see',
1008 'Name-geographic-seealso',
1016 'Personal-name-heading',
1017 'Personal-name-see',
1018 'Personal-name-seealso',
1020 'Place-publication',
1025 'Record-control-number',
1036 'Subject-heading-thesaurus',
1037 'Subject-name-personal',
1038 'Subject-subdivision',
1048 'Term-genre-form-heading',
1049 'Term-genre-form-see',
1050 'Term-genre-form-seealso',
1057 'Title-uniform-heading',
1058 'Title-uniform-see',
1059 'Title-uniform-seealso',
1069 'classification-source',
1071 'coded-location-qualifier',
1082 'Local-classification',
1085 'materials-specified',
1094 'replacementpricedate',
1109 =head2 _handle_exploding_index
1111 my $query = _handle_exploding_index($index, $term)
1113 Callback routine to generate the search for "exploding" indexes (i.e.
1114 those indexes which are turned into multiple or-connected searches based
1119 sub _handle_exploding_index {
1120 my ($QParser, $struct, $filter, $params, $negate, $server) = @_;
1121 my $index = $filter;
1122 my $term = join(' ', @$params);
1124 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1126 my $marcflavour = C4::Context->preference('marcflavour');
1128 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1129 my $wantedcodes = '';
1130 my @subqueries = ( "su:\"$term\"");
1131 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1132 foreach my $auth (@$results) {
1133 my $record = MARC::Record->new_from_usmarc($auth);
1134 my @references = $record->field('5..');
1136 if ($index eq 'su-br') {
1138 } elsif ($index eq 'su-na') {
1140 } elsif ($index eq 'su-rl') {
1143 foreach my $reference (@references) {
1144 my $codes = $reference->subfield($codesubfield);
1145 push @subqueries, 'su:"' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1149 my $query = '(' x scalar(@subqueries) . join(') || ', @subqueries) . ')';
1156 ( $operators, $operands, $indexes, $limits,
1157 $sort_by, $scan, $lang ) =
1158 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1160 Shim function to ease the transition from buildQuery to a new QueryParser.
1161 This function is called at the beginning of buildQuery, and modifies
1162 buildQuery's input. If it can handle the input, it returns a query that
1163 buildQuery will not try to parse.
1167 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1169 my @operators = $operators ? @$operators : ();
1170 my @indexes = $indexes ? @$indexes : ();
1171 my @operands = $operands ? @$operands : ();
1172 my @limits = $limits ? @$limits : ();
1173 my @sort_by = $sort_by ? @$sort_by : ();
1175 my $query = $operands[0];
1181 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1182 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands) );
1183 undef $QParser if (scalar @limits > 0);
1188 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1189 next unless $operands[$ii];
1190 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1193 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1195 foreach my $limit (@limits) {
1197 if (scalar (@sort_by) > 0) {
1198 my $modifier_re = '#(' . join( '|', @{$QParser->modifiers}) . ')';
1199 $query =~ s/$modifier_re//g;
1200 foreach my $modifier (@sort_by) {
1201 $query .= " #$modifier";
1205 $query_desc = $query;
1206 $query_desc =~ s/\s+/ /g;
1207 if ( C4::Context->preference("QueryWeightFields") ) {
1209 $QParser->add_bib1_filter_map( 'biblioserver', 'su-br', { 'callback' => \&_handle_exploding_index });
1210 $QParser->add_bib1_filter_map( 'biblioserver', 'su-na', { 'callback' => \&_handle_exploding_index });
1211 $QParser->add_bib1_filter_map( 'biblioserver', 'su-rl', { 'callback' => \&_handle_exploding_index });
1212 $QParser->parse( $query );
1213 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1215 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1216 s/$modifier_re//g for @operands;
1219 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1225 $simple_query, $query_cgi,
1226 $query_desc, $limit,
1227 $limit_cgi, $limit_desc,
1228 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1230 Build queries and limits in CCL, CGI, Human,
1231 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1233 See verbose embedded documentation.
1239 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1241 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1244 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1247 my @operators = $operators ? @$operators : ();
1248 my @indexes = $indexes ? @$indexes : ();
1249 my @operands = $operands ? @$operands : ();
1250 my @limits = $limits ? @$limits : ();
1251 my @sort_by = $sort_by ? @$sort_by : ();
1253 my $stemming = C4::Context->preference("QueryStemming") || 0;
1254 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1255 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1256 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1257 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1259 # no stemming/weight/fuzzy in NoZebra
1260 if ( C4::Context->preference("NoZebra") ) {
1264 $auto_truncation = 0;
1267 my $query = $operands[0];
1268 my $simple_query = $operands[0];
1270 # initialize the variables we're passing back
1278 my $stopwords_removed; # flag to determine if stopwords have been removed
1281 my $cclindexes = getIndexes();
1282 if ( $query !~ /\s*ccl=/ ) {
1283 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1285 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1287 $query = "ccl=$query" if $cclq;
1290 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1292 if ( $query =~ /^ccl=/ ) {
1294 # This is needed otherwise ccl= and &limit won't work together, and
1295 # this happens when selecting a subject on the opac-detail page
1296 @limits = grep {!/^$/} @limits;
1298 $q .= ' and '.join(' and ', @limits);
1300 return ( undef, $q, $q, "q=ccl=".uri_escape($q), $q, '', '', '', '', 'ccl' );
1302 if ( $query =~ /^cql=/ ) {
1303 return ( undef, $', $', "q=cql=".uri_escape($'), $', '', '', '', '', 'cql' );
1305 if ( $query =~ /^pqf=/ ) {
1307 $query_cgi = "q=".uri_escape($query_desc);
1310 $query_cgi = "q=pqf=".uri_escape($');
1312 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1315 # pass nested queries directly
1316 # FIXME: need better handling of some of these variables in this case
1317 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1318 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1319 # if ( $query =~ /(\(|\))/ ) {
1321 # undef, $query, $simple_query, $query_cgi,
1322 # $query, $limit, $limit_cgi, $limit_desc,
1323 # $stopwords_removed, 'ccl'
1327 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1328 # query operands and indexes and add stemming, truncation, field weighting, etc.
1329 # Once we do so, we'll end up with a value in $query, just like if we had an
1330 # incoming $query from the user
1333 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1334 my $previous_operand
1335 ; # a flag used to keep track if there was a previous query
1336 # if there was, we can apply the current operator
1338 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1340 # COMBINE OPERANDS, INDEXES AND OPERATORS
1341 if ( $operands[$i] ) {
1342 $operands[$i]=~s/^\s+//;
1344 # A flag to determine whether or not to add the index to the query
1347 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1348 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1351 $remove_stopwords = 0;
1353 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1355 my $operand = $operands[$i];
1356 my $index = $indexes[$i];
1358 # Add index-specific attributes
1359 # Date of Publication
1360 if ( $index eq 'yr' ) {
1361 $index .= ",st-numeric";
1363 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1366 # Date of Acquisition
1367 elsif ( $index eq 'acqdate' ) {
1368 $index .= ",st-date-normalized";
1370 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1372 # ISBN,ISSN,Standard Number, don't need special treatment
1373 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1375 $stemming, $auto_truncation,
1376 $weight_fields, $fuzzy_enabled,
1378 ) = ( 0, 0, 0, 0, 0 );
1386 # Set default structure attribute (word list)
1387 my $struct_attr = q{};
1388 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1389 $struct_attr = ",wrdl";
1392 # Some helpful index variants
1393 my $index_plus = $index . $struct_attr . ':';
1394 my $index_plus_comma = $index . $struct_attr . ',';
1397 if ($remove_stopwords) {
1398 ( $operand, $stopwords_removed ) =
1399 _remove_stopwords( $operand, $index );
1400 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1401 warn "REMOVED STOPWORDS: @$stopwords_removed"
1402 if ( $stopwords_removed && $DEBUG );
1405 if ($auto_truncation){
1406 unless ( $index =~ /(st-|phr|ext)/ ) {
1407 #FIXME only valid with LTR scripts
1408 $operand=join(" ",map{
1409 (index($_,"*")>0?"$_":"$_*")
1410 }split (/\s+/,$operand));
1411 warn $operand if $DEBUG;
1416 my $truncated_operand;
1417 my( $nontruncated, $righttruncated, $lefttruncated,
1418 $rightlefttruncated, $regexpr
1419 ) = _detect_truncation( $operand, $index );
1421 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1426 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1427 scalar(@$rightlefttruncated) > 0 )
1430 # Don't field weight or add the index to the query, we do it here
1432 undef $weight_fields;
1433 my $previous_truncation_operand;
1434 if (scalar @$nontruncated) {
1435 $truncated_operand .= "$index_plus @$nontruncated ";
1436 $previous_truncation_operand = 1;
1438 if (scalar @$righttruncated) {
1439 $truncated_operand .= "and " if $previous_truncation_operand;
1440 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1441 $previous_truncation_operand = 1;
1443 if (scalar @$lefttruncated) {
1444 $truncated_operand .= "and " if $previous_truncation_operand;
1445 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1446 $previous_truncation_operand = 1;
1448 if (scalar @$rightlefttruncated) {
1449 $truncated_operand .= "and " if $previous_truncation_operand;
1450 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1451 $previous_truncation_operand = 1;
1454 $operand = $truncated_operand if $truncated_operand;
1455 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1458 my $stemmed_operand;
1459 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1462 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1464 # Handle Field Weighting
1465 my $weighted_operand;
1466 if ($weight_fields) {
1467 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1468 $operand = $weighted_operand;
1472 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1474 # If there's a previous operand, we need to add an operator
1475 if ($previous_operand) {
1477 # User-specified operator
1478 if ( $operators[ $i - 1 ] ) {
1479 $query .= " $operators[$i-1] ";
1480 $query .= " $index_plus " unless $indexes_set;
1481 $query .= " $operand";
1482 $query_cgi .= "&op=".uri_escape($operators[$i-1]);
1483 $query_cgi .= "&idx=".uri_escape($index) if $index;
1484 $query_cgi .= "&q=".uri_escape($operands[$i]) if $operands[$i];
1486 " $operators[$i-1] $index_plus $operands[$i]";
1489 # Default operator is and
1492 $query .= "$index_plus " unless $indexes_set;
1493 $query .= "$operand";
1494 $query_cgi .= "&op=and&idx=".uri_escape($index) if $index;
1495 $query_cgi .= "&q=".uri_escape($operands[$i]) if $operands[$i];
1496 $query_desc .= " and $index_plus $operands[$i]";
1500 # There isn't a pervious operand, don't need an operator
1503 # Field-weighted queries already have indexes set
1504 $query .= " $index_plus " unless $indexes_set;
1506 $query_desc .= " $index_plus $operands[$i]";
1507 $query_cgi .= "&idx=".uri_escape($index) if $index;
1508 $query_cgi .= "&q=".uri_escape($operands[$i]) if $operands[$i];
1509 $previous_operand = 1;
1514 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1517 my %group_OR_limits;
1518 my $availability_limit;
1519 foreach my $this_limit (@limits) {
1520 next unless $this_limit;
1521 if ( $this_limit =~ /available/ ) {
1523 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1525 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1526 $availability_limit .=
1527 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1528 $limit_cgi .= "&limit=available";
1532 # group_OR_limits, prefixed by mc-
1533 # OR every member of the group
1534 elsif ( $this_limit =~ /mc/ ) {
1535 my ($k,$v) = split(/:/, $this_limit,2);
1536 if ( $k !~ /mc-i(tem)?type/ ) {
1537 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1538 $this_limit =~ tr/"//d;
1539 $this_limit = $k.":\"".$v."\"";
1542 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1543 $limit_desc .= " or " if $group_OR_limits{$k};
1544 $group_OR_limits{$k} .= "$this_limit";
1545 $limit_cgi .= "&limit=$this_limit";
1546 $limit_desc .= " $this_limit";
1549 # Regular old limits
1551 $limit .= " and " if $limit || $query;
1552 $limit .= "$this_limit";
1553 $limit_cgi .= "&limit=$this_limit";
1554 if ($this_limit =~ /^branch:(.+)/) {
1555 my $branchcode = $1;
1556 my $branchname = GetBranchName($branchcode);
1557 if (defined $branchname) {
1558 $limit_desc .= " branch:$branchname";
1560 $limit_desc .= " $this_limit";
1563 $limit_desc .= " $this_limit";
1567 foreach my $k (keys (%group_OR_limits)) {
1568 $limit .= " and " if ( $query || $limit );
1569 $limit .= "($group_OR_limits{$k})";
1571 if ($availability_limit) {
1572 $limit .= " and " if ( $query || $limit );
1573 $limit .= "($availability_limit)";
1576 # Normalize the query and limit strings
1577 # This is flawed , means we can't search anything with : in it
1578 # if user wants to do ccl or cql, start the query with that
1579 # $query =~ s/:/=/g;
1580 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1581 $query =~ s/(?<=(wrdl)):/=/g;
1582 $query =~ s/(?<=(trn|phr)):/=/g;
1584 for ( $query, $query_desc, $limit, $limit_desc ) {
1585 s/ +/ /g; # remove extra spaces
1586 s/^ //g; # remove any beginning spaces
1587 s/ $//g; # remove any ending spaces
1588 s/==/=/g; # remove double == from query
1590 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1592 for ($query_cgi,$simple_query) {
1595 # append the limit to the query
1596 $query .= " " . $limit;
1600 warn "QUERY:" . $query;
1601 warn "QUERY CGI:" . $query_cgi;
1602 warn "QUERY DESC:" . $query_desc;
1603 warn "LIMIT:" . $limit;
1604 warn "LIMIT CGI:" . $limit_cgi;
1605 warn "LIMIT DESC:" . $limit_desc;
1606 warn "---------\nLeave buildQuery\n---------";
1609 undef, $query, $simple_query, $query_cgi,
1610 $query_desc, $limit, $limit_cgi, $limit_desc,
1611 $stopwords_removed, $query_type
1615 =head2 searchResults
1617 my @search_results = searchResults($search_context, $searchdesc, $hits,
1618 $results_per_page, $offset, $scan,
1621 Format results in a form suitable for passing to the template
1625 # IMO this subroutine is pretty messy still -- it's responsible for
1626 # building the HTML output for the template
1628 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1629 my $dbh = C4::Context->dbh;
1634 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1635 my ($is_opac, $hidelostitems);
1636 if ($search_context eq 'opac') {
1637 $hidelostitems = C4::Context->preference('hidelostitems');
1641 #Build branchnames hash
1643 #get branch information.....
1645 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1647 while ( my $bdata = $bsth->fetchrow_hashref ) {
1648 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1650 # FIXME - We build an authorised values hash here, using the default framework
1651 # though it is possible to have different authvals for different fws.
1653 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1655 # get notforloan authorised value list (see $shelflocations FIXME)
1656 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1658 #Build itemtype hash
1659 #find itemtype & itemtype image
1663 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1666 while ( my $bdata = $bsth->fetchrow_hashref ) {
1667 foreach (qw(description imageurl summary notforloan)) {
1668 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1672 #search item field code
1673 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1675 ## find column names of items related to MARC
1676 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1678 my %subfieldstosearch;
1679 while ( ( my $column ) = $sth2->fetchrow ) {
1680 my ( $tagfield, $tagsubfield ) =
1681 &GetMarcFromKohaField( "items." . $column, "" );
1682 $subfieldstosearch{$column} = $tagsubfield;
1685 # handle which records to actually retrieve
1687 if ( $hits && $offset + $results_per_page <= $hits ) {
1688 $times = $offset + $results_per_page;
1691 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1694 my $marcflavour = C4::Context->preference("marcflavour");
1695 # We get the biblionumber position in MARC
1696 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1698 # loop through all of the records we've retrieved
1699 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1700 my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1704 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1705 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1706 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1707 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1708 $oldbiblio->{result_number} = $i + 1;
1710 # add imageurl to itemtype if there is one
1711 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1713 $oldbiblio->{'authorised_value_images'} = ($search_context eq 'opac' && C4::Context->preference('AuthorisedValueImages')) || ($search_context eq 'intranet' && C4::Context->preference('StaffAuthorisedValueImages')) ? C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) ) : [];
1714 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1715 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1716 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1717 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1718 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1720 # edition information, if any
1721 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1722 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1723 # Build summary if there is one (the summary is defined in the itemtypes table)
1724 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1725 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1726 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1727 my @fields = $marcrecord->fields();
1730 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1732 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1733 $tag =~ /(.{3})(.)/;
1734 if($marcrecord->field($1)){
1735 my @abc = $marcrecord->field($1)->subfield($2);
1736 $tags->{$tag} = $#abc + 1 ;
1740 # We catch how many times to repeat this line
1742 foreach my $tag (keys(%$tags)){
1743 $max = $tags->{$tag} if($tags->{$tag} > $max);
1746 # we replace, and repeat each line
1747 for (my $i = 0 ; $i < $max ; $i++){
1748 my $newline = $line;
1750 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1751 $tag =~ /(.{3})(.)/;
1753 if($marcrecord->field($1)){
1754 my @repl = $marcrecord->field($1)->subfield($2);
1755 my $subfieldvalue = $repl[$i];
1757 if (! utf8::is_utf8($subfieldvalue)) {
1758 utf8::decode($subfieldvalue);
1761 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1764 $newsummary .= "$newline\n";
1768 $newsummary =~ s/\[(.*?)]//g;
1769 $newsummary =~ s/\n/<br\/>/g;
1770 $oldbiblio->{summary} = $newsummary;
1773 # Pull out the items fields
1774 my @fields = $marcrecord->field($itemtag);
1775 my $marcflavor = C4::Context->preference("marcflavour");
1776 # adding linked items that belong to host records
1777 my $analyticsfield = '773';
1778 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1779 $analyticsfield = '773';
1780 } elsif ($marcflavor eq 'UNIMARC') {
1781 $analyticsfield = '461';
1783 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1784 my $hostbiblionumber = $hostfield->subfield("0");
1785 my $linkeditemnumber = $hostfield->subfield("9");
1786 if(!$hostbiblionumber eq undef){
1787 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1788 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1789 if(!$hostbiblio eq undef){
1790 my @hostitems = $hostbiblio->field($itemfield);
1791 foreach my $hostitem (@hostitems){
1792 if ($hostitem->subfield("9") eq $linkeditemnumber){
1793 my $linkeditem =$hostitem;
1794 # append linked items if they exist
1795 if (!$linkeditem eq undef){
1796 push (@fields, $linkeditem);}
1803 # Setting item statuses for display
1804 my @available_items_loop;
1805 my @onloan_items_loop;
1806 my @other_items_loop;
1808 my $available_items;
1812 my $ordered_count = 0;
1813 my $available_count = 0;
1814 my $onloan_count = 0;
1815 my $longoverdue_count = 0;
1816 my $other_count = 0;
1817 my $wthdrawn_count = 0;
1818 my $itemlost_count = 0;
1819 my $hideatopac_count = 0;
1820 my $itembinding_count = 0;
1821 my $itemdamaged_count = 0;
1822 my $item_in_transit_count = 0;
1823 my $can_place_holds = 0;
1824 my $item_onhold_count = 0;
1825 my $items_count = scalar(@fields);
1826 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1827 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1828 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1830 # loop through every item
1831 foreach my $field (@fields) {
1834 # populate the items hash
1835 foreach my $code ( keys %subfieldstosearch ) {
1836 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1838 $item->{description} = $itemtypes{ $item->{itype} }{description};
1842 # hidden because lost
1843 if ($hidelostitems && $item->{itemlost}) {
1844 $hideatopac_count++;
1847 # hidden based on OpacHiddenItems syspref
1848 my @hi = C4::Items::GetHiddenItemnumbers($item);
1850 push @hiddenitems, @hi;
1851 $hideatopac_count++;
1856 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1857 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1859 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1860 if ($item->{$hbranch}) {
1861 $item->{'branchname'} = $branches{$item->{$hbranch}};
1863 elsif ($item->{$otherbranch}) { # Last resort
1864 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1867 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1868 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1869 my $userenv = C4::Context->userenv;
1870 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1872 my $key = $prefix . $item->{onloan} . $item->{barcode};
1873 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1874 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1875 $onloan_items->{$key}->{branchname} = $item->{branchname};
1876 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1877 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1878 $onloan_items->{$key}->{description} = $item->{description};
1879 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1880 # if something's checked out and lost, mark it as 'long overdue'
1881 if ( $item->{itemlost} ) {
1882 $onloan_items->{$prefix}->{longoverdue}++;
1883 $longoverdue_count++;
1884 } else { # can place holds as long as item isn't lost
1885 $can_place_holds = 1;
1889 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1893 if ( $item->{notforloan} < 0 ) {
1897 # is item in transit?
1898 my $transfertwhen = '';
1899 my ($transfertfrom, $transfertto);
1901 # is item on the reserve shelf?
1902 my $reservestatus = '';
1904 unless ($item->{wthdrawn}
1905 || $item->{itemlost}
1907 || $item->{notforloan}
1908 || $items_count > 20) {
1910 # A couple heuristics to limit how many times
1911 # we query the database for item transfer information, sacrificing
1912 # accuracy in some cases for speed;
1914 # 1. don't query if item has one of the other statuses
1915 # 2. don't check transit status if the bib has
1916 # more than 20 items
1918 # FIXME: to avoid having the query the database like this, and to make
1919 # the in transit status count as unavailable for search limiting,
1920 # should map transit status to record indexed in Zebra.
1922 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1923 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber}, $oldbiblio->{biblionumber} );
1926 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1927 if ( $item->{wthdrawn}
1928 || $item->{itemlost}
1930 || $item->{notforloan}
1931 || $reservestatus eq 'Waiting'
1932 || ($transfertwhen ne ''))
1934 $wthdrawn_count++ if $item->{wthdrawn};
1935 $itemlost_count++ if $item->{itemlost};
1936 $itemdamaged_count++ if $item->{damaged};
1937 $item_in_transit_count++ if $transfertwhen ne '';
1938 $item_onhold_count++ if $reservestatus eq 'Waiting';
1939 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1941 # can place hold on item ?
1942 if ( !$item->{itemlost} ) {
1943 if ( !$item->{wthdrawn} ){
1944 if ( $item->{damaged} ){
1945 if ( C4::Context->preference('AllowHoldsOnDamagedItems') ){
1946 # can place a hold on a damaged item if AllowHoldsOnDamagedItems is true
1947 if ( ( !$item->{notforloan} || $item->{notforloan} < 0 ) ){
1948 # item is either for loan or has notforloan < 0
1949 $can_place_holds = 1;
1952 } elsif ( $item->{notforloan} < 0 ) {
1953 # item is not damaged and notforloan is < 0
1954 $can_place_holds = 1;
1961 my $key = $prefix . $item->{status};
1962 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1963 $other_items->{$key}->{$_} = $item->{$_};
1965 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1966 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1967 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1968 $other_items->{$key}->{count}++ if $item->{$hbranch};
1969 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1970 $other_items->{$key}->{description} = $item->{description};
1971 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1975 $can_place_holds = 1;
1977 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1978 foreach (qw(branchname itemcallnumber description)) {
1979 $available_items->{$prefix}->{$_} = $item->{$_};
1981 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1982 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1985 } # notforloan, item level and biblioitem level
1987 # if all items are hidden, do not show the record
1988 if ($items_count > 0 && $hideatopac_count == $items_count) {
1992 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1993 for my $key ( sort keys %$onloan_items ) {
1994 (++$onloanitemscount > $maxitems) and last;
1995 push @onloan_items_loop, $onloan_items->{$key};
1997 for my $key ( sort keys %$other_items ) {
1998 (++$otheritemscount > $maxitems) and last;
1999 push @other_items_loop, $other_items->{$key};
2001 for my $key ( sort keys %$available_items ) {
2002 (++$availableitemscount > $maxitems) and last;
2003 push @available_items_loop, $available_items->{$key}
2006 # XSLT processing of some stuff
2008 SetUTF8Flag($marcrecord);
2009 warn $marcrecord->as_formatted if $DEBUG;
2010 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2011 if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2012 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2013 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2016 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2017 if (!C4::Context->preference("item-level_itypes")) {
2018 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2019 $can_place_holds = 0;
2022 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2023 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2024 $oldbiblio->{items_count} = $items_count;
2025 $oldbiblio->{available_items_loop} = \@available_items_loop;
2026 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2027 $oldbiblio->{other_items_loop} = \@other_items_loop;
2028 $oldbiblio->{availablecount} = $available_count;
2029 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2030 $oldbiblio->{onloancount} = $onloan_count;
2031 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2032 $oldbiblio->{othercount} = $other_count;
2033 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2034 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
2035 $oldbiblio->{itemlostcount} = $itemlost_count;
2036 $oldbiblio->{damagedcount} = $itemdamaged_count;
2037 $oldbiblio->{intransitcount} = $item_in_transit_count;
2038 $oldbiblio->{onholdcount} = $item_onhold_count;
2039 $oldbiblio->{orderedcount} = $ordered_count;
2041 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2042 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2043 my $subfields = substr $fieldspec, 3;
2044 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2045 my @alternateholdingsinfo = ();
2046 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2047 my $alternateholdingscount = 0;
2049 for my $field (@holdingsfields) {
2050 my %holding = ( holding => '' );
2051 my $havesubfield = 0;
2052 for my $subfield ($field->subfields()) {
2053 if ((index $subfields, $$subfield[0]) >= 0) {
2054 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2055 $holding{'holding'} .= $$subfield[1];
2059 if ($havesubfield) {
2060 push(@alternateholdingsinfo, \%holding);
2061 $alternateholdingscount++;
2065 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2066 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2069 push( @newresults, $oldbiblio );
2075 =head2 SearchAcquisitions
2076 Search for acquisitions
2079 sub SearchAcquisitions{
2080 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2082 my $dbh=C4::Context->dbh;
2083 # Variable initialization
2087 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2088 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2089 WHERE dateaccessioned BETWEEN ? AND ?
2092 my (@params,@loopcriteria);
2094 push @params, $datebegin->output("iso");
2095 push @params, $dateend->output("iso");
2097 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2098 if(C4::Context->preference("item-level_itypes")){
2099 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2101 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2103 push @params, @$itemtypes;
2106 if ($criteria =~/itemtype/){
2107 if(C4::Context->preference("item-level_itypes")){
2108 $str .= "AND items.itype=? ";
2110 $str .= "AND biblioitems.itemtype=? ";
2113 if(scalar(@$itemtypes) == 0){
2114 my $itypes = GetItemTypes();
2115 for my $key (keys %$itypes){
2116 push @$itemtypes, $key;
2120 @loopcriteria= @$itemtypes;
2121 }elsif ($criteria=~/itemcallnumber/){
2122 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2123 OR items.itemcallnumber is NULL
2124 OR items.itemcallnumber = '')";
2126 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2128 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2129 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2132 if ($orderby =~ /date_desc/){
2133 $str.=" ORDER BY dateaccessioned DESC";
2135 $str.=" ORDER BY title";
2138 my $qdataacquisitions=$dbh->prepare($str);
2140 my @loopacquisitions;
2141 foreach my $value(@loopcriteria){
2142 push @params,$value;
2144 $cell{"title"}=$value;
2145 $cell{"titlecode"}=$value;
2147 eval{$qdataacquisitions->execute(@params);};
2149 if ($@){ warn "recentacquisitions Error :$@";}
2152 while (my $data=$qdataacquisitions->fetchrow_hashref){
2153 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2155 $cell{"loopdata"}=\@loopdata;
2157 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2160 $qdataacquisitions->finish;
2161 return \@loopacquisitions;
2163 #----------------------------------------------------------------------
2165 # Non-Zebra GetRecords#
2166 #----------------------------------------------------------------------
2170 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
2176 $query, $simple_query, $sort_by_ref, $servers_ref,
2177 $results_per_page, $offset, $expanded_facet, $branches,
2180 warn "query =$query" if $DEBUG;
2181 my $result = NZanalyse($query);
2182 warn "results =$result" if $DEBUG;
2184 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
2190 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
2191 the list is built from an inverted index in the nozebra SQL table
2192 note that title is here only for convenience : the sorting will be very fast when requested on title
2193 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
2198 my ( $string, $server ) = @_;
2199 # warn "---------" if $DEBUG;
2200 warn " NZanalyse" if $DEBUG;
2201 # warn "---------" if $DEBUG;
2203 # $server contains biblioserver or authorities, depending on what we search on.
2204 #warn "querying : $string on $server";
2205 $server = 'biblioserver' unless $server;
2207 # if we have a ", replace the content to discard temporarily any and/or/not inside
2209 if ( $string =~ /"/ ) {
2210 $string =~ s/"(.*?)"/__X__/;
2212 warn "commacontent : $commacontent" if $DEBUG;
2215 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
2216 # then, call again NZanalyse with $left and $right
2217 # (recursive until we find a leaf (=> something without and/or/not)
2218 # delete repeated operator... Would then go in infinite loop
2219 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
2222 #process parenthesis before.
2223 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
2226 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
2228 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2230 my $leftresult = NZanalyse( $left, $server );
2232 my $rightresult = NZanalyse( $right, $server );
2234 # OK, we have the results for right and left part of the query
2235 # depending of operand, intersect, union or exclude both lists
2236 # to get a result list
2237 if ( $operator eq ' and ' ) {
2238 return NZoperatorAND($leftresult,$rightresult);
2240 elsif ( $operator eq ' or ' ) {
2242 # just merge the 2 strings
2243 return $leftresult . $rightresult;
2245 elsif ( $operator eq ' not ' ) {
2246 return NZoperatorNOT($leftresult,$rightresult);
2250 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2254 warn "string :" . $string if $DEBUG;
2258 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2261 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2263 warn "no parenthesis. left : $left operator: $operator right: $right"
2266 # it's not a leaf, we have a and/or/not
2269 # reintroduce comma content if needed
2270 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2271 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2272 warn "node : $left / $operator / $right\n" if $DEBUG;
2273 my $leftresult = NZanalyse( $left, $server );
2274 my $rightresult = NZanalyse( $right, $server );
2275 warn " leftresult : $leftresult" if $DEBUG;
2276 warn " rightresult : $rightresult" if $DEBUG;
2277 # OK, we have the results for right and left part of the query
2278 # depending of operand, intersect, union or exclude both lists
2279 # to get a result list
2280 if ( $operator eq ' and ' ) {
2281 return NZoperatorAND($leftresult,$rightresult);
2283 elsif ( $operator eq ' or ' ) {
2285 # just merge the 2 strings
2286 return $leftresult . $rightresult;
2288 elsif ( $operator eq ' not ' ) {
2289 return NZoperatorNOT($leftresult,$rightresult);
2293 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2294 die "error : operand unknown : $operator for $string";
2297 # it's a leaf, do the real SQL query and return the result
2300 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2301 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2302 #remove trailing blank at the beginning
2304 warn "leaf:$string" if $DEBUG;
2306 # parse the string in in operator/operand/value again
2310 if ($string =~ /(.*)(>=|<=)(.*)/) {
2317 # warn "handling leaf... left:$left operator:$operator right:$right"
2319 unless ($operator) {
2320 if ($string =~ /(.*)(>|<|=)(.*)/) {
2325 "handling unless (operator)... left:$left operator:$operator right:$right"
2333 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2336 # automatic replace for short operators
2337 $left = 'title' if $left =~ '^ti$';
2338 $left = 'author' if $left =~ '^au$';
2339 $left = 'publisher' if $left =~ '^pb$';
2340 $left = 'subject' if $left =~ '^su$';
2341 $left = 'koha-Auth-Number' if $left =~ '^an$';
2342 $left = 'keyword' if $left =~ '^kw$';
2343 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2344 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2345 my $dbh = C4::Context->dbh;
2346 if ( $operator && $left ne 'keyword' ) {
2347 #do a specific search
2348 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2349 my $sth = $dbh->prepare(
2350 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2352 warn "$left / $operator / $right\n" if $DEBUG;
2354 # split each word, query the DB and build the biblionumbers result
2355 #sanitizing leftpart
2356 $left =~ s/^\s+|\s+$//;
2357 foreach ( split / /, $right ) {
2359 $_ =~ s/^\s+|\s+$//;
2361 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2362 $sth->execute( $server, $left, $_ )
2363 or warn "execute failed: $!";
2364 while ( my ( $line, $value ) = $sth->fetchrow ) {
2366 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2367 # otherwise, fill the result
2368 $biblionumbers .= $line
2369 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2370 warn "result : $value "
2371 . ( $right =~ /\d/ ) . "=="
2372 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2375 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2377 warn "NZAND" if $DEBUG;
2378 $results = NZoperatorAND($biblionumbers,$results);
2380 $results = $biblionumbers;
2385 #do a complete search (all indexes), if index='kw' do complete search too.
2386 my $sth = $dbh->prepare(
2387 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2390 # split each word, query the DB and build the biblionumbers result
2391 foreach ( split / /, $string ) {
2392 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2393 warn "search on all indexes on $_" if $DEBUG;
2396 $sth->execute( $server, $_ );
2397 while ( my $line = $sth->fetchrow ) {
2398 $biblionumbers .= $line;
2401 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2403 $results = NZoperatorAND($biblionumbers,$results);
2406 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2407 $results = $biblionumbers;
2411 warn "return : $results for LEAF : $string" if $DEBUG;
2414 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2418 my ($rightresult, $leftresult)=@_;
2420 my @leftresult = split /;/, $leftresult;
2421 warn " @leftresult / $rightresult \n" if $DEBUG;
2423 # my @rightresult = split /;/,$leftresult;
2426 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2427 # the result is stored twice, to have the same weight for AND than OR.
2428 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2429 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2430 foreach (@leftresult) {
2433 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2434 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2435 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2437 "$value-$countvalue;$value-$countvalue;";
2440 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2441 return $finalresult;
2445 my ($rightresult, $leftresult)=@_;
2446 return $rightresult.$leftresult;
2450 my ($leftresult, $rightresult)=@_;
2452 my @leftresult = split /;/, $leftresult;
2454 # my @rightresult = split /;/,$leftresult;
2456 foreach (@leftresult) {
2458 $value=$1 if $value=~m/(.*)-\d+$/;
2459 unless ($rightresult =~ "$value-") {
2460 $finalresult .= "$_;";
2463 return $finalresult;
2468 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2475 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2476 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2478 # order title asc by default
2479 # $ordering = '1=36 <i' unless $ordering;
2480 $results_per_page = 20 unless $results_per_page;
2481 $offset = 0 unless $offset;
2482 my $dbh = C4::Context->dbh;
2485 # order by POPULARITY
2487 if ( $ordering =~ /popularity/ ) {
2491 # popularity is not in MARC record, it's builded from a specific query
2493 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2494 foreach ( split /;/, $biblionumbers ) {
2495 my ( $biblionumber, $title ) = split /,/, $_;
2496 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2497 $sth->execute($biblionumber);
2498 my $popularity = $sth->fetchrow || 0;
2500 # hint : the key is popularity.title because we can have
2501 # many results with the same popularity. In this case, sub-ordering is done by title
2502 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2503 # (un-frequent, I agree, but we won't forget anything that way ;-)
2504 $popularity{ sprintf( "%10d", $popularity ) . $title
2505 . $biblionumber } = $biblionumber;
2508 # sort the hash and return the same structure as GetRecords (Zebra querying)
2511 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2512 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2513 $result_hash->{'RECORDS'}[ $numbers++ ] =
2514 $result{ $popularity{$key} }->as_usmarc();
2517 else { # sort popularity ASC
2518 foreach my $key ( sort ( keys %popularity ) ) {
2519 $result_hash->{'RECORDS'}[ $numbers++ ] =
2520 $result{ $popularity{$key} }->as_usmarc();
2523 my $finalresult = ();
2524 $result_hash->{'hits'} = $numbers;
2525 $finalresult->{'biblioserver'} = $result_hash;
2526 return $finalresult;
2532 elsif ( $ordering =~ /author/ ) {
2534 foreach ( split /;/, $biblionumbers ) {
2535 my ( $biblionumber, $title ) = split /,/, $_;
2536 my $record = GetMarcBiblio($biblionumber);
2538 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2539 $author = $record->subfield( '200', 'f' );
2540 $author = $record->subfield( '700', 'a' ) unless $author;
2543 $author = $record->subfield( '100', 'a' );
2546 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2547 # and we don't want to get only 1 result for each of them !!!
2548 $result{ $author . $biblionumber } = $record;
2551 # sort the hash and return the same structure as GetRecords (Zebra querying)
2554 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2555 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2556 $result_hash->{'RECORDS'}[ $numbers++ ] =
2557 $result{$key}->as_usmarc();
2560 else { # sort by author ASC
2561 foreach my $key ( sort ( keys %result ) ) {
2562 $result_hash->{'RECORDS'}[ $numbers++ ] =
2563 $result{$key}->as_usmarc();
2566 my $finalresult = ();
2567 $result_hash->{'hits'} = $numbers;
2568 $finalresult->{'biblioserver'} = $result_hash;
2569 return $finalresult;
2572 # ORDER BY callnumber
2575 elsif ( $ordering =~ /callnumber/ ) {
2577 foreach ( split /;/, $biblionumbers ) {
2578 my ( $biblionumber, $title ) = split /,/, $_;
2579 my $record = GetMarcBiblio($biblionumber);
2581 my $frameworkcode = GetFrameworkCode($biblionumber);
2582 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2583 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2584 unless $callnumber_tag;
2585 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2586 $callnumber = $record->subfield( '200', 'f' );
2588 $callnumber = $record->subfield( '100', 'a' );
2591 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2592 # and we don't want to get only 1 result for each of them !!!
2593 $result{ $callnumber . $biblionumber } = $record;
2596 # sort the hash and return the same structure as GetRecords (Zebra querying)
2599 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2600 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2601 $result_hash->{'RECORDS'}[ $numbers++ ] =
2602 $result{$key}->as_usmarc();
2605 else { # sort by title ASC
2606 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2607 $result_hash->{'RECORDS'}[ $numbers++ ] =
2608 $result{$key}->as_usmarc();
2611 my $finalresult = ();
2612 $result_hash->{'hits'} = $numbers;
2613 $finalresult->{'biblioserver'} = $result_hash;
2614 return $finalresult;
2616 elsif ( $ordering =~ /pubdate/ ) { #pub year
2618 foreach ( split /;/, $biblionumbers ) {
2619 my ( $biblionumber, $title ) = split /,/, $_;
2620 my $record = GetMarcBiblio($biblionumber);
2621 my ( $publicationyear_tag, $publicationyear_subfield ) =
2622 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2623 my $publicationyear =
2624 $record->subfield( $publicationyear_tag,
2625 $publicationyear_subfield );
2627 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2628 # and we don't want to get only 1 result for each of them !!!
2629 $result{ $publicationyear . $biblionumber } = $record;
2632 # sort the hash and return the same structure as GetRecords (Zebra querying)
2635 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2636 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2637 $result_hash->{'RECORDS'}[ $numbers++ ] =
2638 $result{$key}->as_usmarc();
2641 else { # sort by pub year ASC
2642 foreach my $key ( sort ( keys %result ) ) {
2643 $result_hash->{'RECORDS'}[ $numbers++ ] =
2644 $result{$key}->as_usmarc();
2647 my $finalresult = ();
2648 $result_hash->{'hits'} = $numbers;
2649 $finalresult->{'biblioserver'} = $result_hash;
2650 return $finalresult;
2656 elsif ( $ordering =~ /title/ ) {
2658 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2660 foreach ( split /;/, $biblionumbers ) {
2661 my ( $biblionumber, $title ) = split /,/, $_;
2663 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2664 # and we don't want to get only 1 result for each of them !!!
2665 # hint & speed improvement : we can order without reading the record
2666 # so order, and read records only for the requested page !
2667 $result{ $title . $biblionumber } = $biblionumber;
2670 # sort the hash and return the same structure as GetRecords (Zebra querying)
2673 if ( $ordering eq 'title_az' ) { # sort by title desc
2674 foreach my $key ( sort ( keys %result ) ) {
2675 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2678 else { # sort by title ASC
2679 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2680 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2684 # limit the $results_per_page to result size if it's more
2685 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2687 # for the requested page, replace biblionumber by the complete record
2688 # speed improvement : avoid reading too much things
2690 my $counter = $offset ;
2691 $counter <= $offset + $results_per_page ;
2695 $result_hash->{'RECORDS'}[$counter] =
2696 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2698 my $finalresult = ();
2699 $result_hash->{'hits'} = $numbers;
2700 $finalresult->{'biblioserver'} = $result_hash;
2701 return $finalresult;
2708 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2711 foreach ( split /;/, $biblionumbers ) {
2712 my ( $biblionumber, $title ) = split /,/, $_;
2713 $title =~ /(.*)-(\d)/;
2718 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2719 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2720 # biblio N has ranking = 6
2721 $count_ranking{$biblionumber} += $ranking;
2724 # build the result by "inverting" the count_ranking hash
2725 # hing : as usual, we don't order by ranking only, to avoid having only 1 result for each rank. We build an hash on concat(ranking,biblionumber) instead
2727 foreach ( keys %count_ranking ) {
2728 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2731 # sort the hash and return the same structure as GetRecords (Zebra querying)
2734 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2735 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2738 # limit the $results_per_page to result size if it's more
2739 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2741 # for the requested page, replace biblionumber by the complete record
2742 # speed improvement : avoid reading too much things
2744 my $counter = $offset ;
2745 $counter <= $offset + $results_per_page ;
2749 $result_hash->{'RECORDS'}[$counter] =
2750 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2751 if $result_hash->{'RECORDS'}[$counter];
2753 my $finalresult = ();
2754 $result_hash->{'hits'} = $numbers;
2755 $finalresult->{'biblioserver'} = $result_hash;
2756 return $finalresult;
2760 =head2 enabled_staff_search_views
2762 %hash = enabled_staff_search_views()
2764 This function returns a hash that contains three flags obtained from the system
2765 preferences, used to determine whether a particular staff search results view
2770 =item C<Output arg:>
2772 * $hash{can_view_MARC} is true only if the MARC view is enabled
2773 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2774 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2776 =item C<usage in the script:>
2780 $template->param ( C4::Search::enabled_staff_search_views );
2784 sub enabled_staff_search_views
2787 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2788 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2789 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2793 sub AddSearchHistory{
2794 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2795 my $dbh = C4::Context->dbh;
2797 # Add the request the user just made
2798 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2799 my $sth = $dbh->prepare($sql);
2800 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2801 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2804 sub GetSearchHistory{
2805 my ($borrowernumber,$session)=@_;
2806 my $dbh = C4::Context->dbh;
2808 # Add the request the user just made
2809 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2810 my $sth = $dbh->prepare($query);
2811 $sth->execute($borrowernumber, $session);
2812 return $sth->fetchall_hashref({});
2815 =head2 z3950_search_args
2817 $arrayref = z3950_search_args($matchpoints)
2819 This function returns an array reference that contains the search parameters to be
2820 passed to the Z39.50 search script (z3950_search.pl). The array elements
2821 are hash refs whose keys are name, value and encvalue, and whose values are the
2822 name of a search parameter, the value of that search parameter and the URL encoded
2823 value of that parameter.
2825 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2827 The search parameter values are obtained from the bibliographic record whose
2828 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2830 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2831 a general purpose search argument. In this case, the returned array contains only
2832 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2834 If a search parameter value is undefined or empty, it is not included in the returned
2837 The returned array reference may be passed directly to the template parameters.
2841 =item C<Output arg:>
2843 * $array containing hash refs as described above
2845 =item C<usage in the script:>
2849 $data = Biblio::GetBiblioData($bibno);
2850 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2854 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2858 sub z3950_search_args {
2860 my $isbn = Business::ISBN->new($bibrec);
2862 if (defined $isbn && $isbn->is_valid)
2864 $bibrec = { isbn => $bibrec } if !ref $bibrec;
2867 $bibrec = { title => $bibrec } if !ref $bibrec;
2870 for my $field (qw/ lccn isbn issn title author dewey subject /)
2872 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2873 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2878 =head2 GetDistinctValues($field);
2880 C<$field> is a reference to the fields array
2884 sub GetDistinctValues {
2885 my ($fieldname,$string)=@_;
2886 # returns a reference to a hash of references to branches...
2887 if ($fieldname=~/\./){
2888 my ($table,$column)=split /\./, $fieldname;
2889 my $dbh = C4::Context->dbh;
2890 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2891 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 ");
2893 my $elements=$sth->fetchall_arrayref({});
2898 my @servers=qw<biblioserver authorityserver>;
2899 my (@zconns,@results);
2900 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2901 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2904 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2907 # The big moment: asynchronously retrieve results from all servers
2913 my ( $i, $size ) = @_;
2914 for ( my $j = 0 ; $j < $size ; $j++ ) {
2916 @hashscan{qw(value cnt)} =
2917 $results[ $i - 1 ]->display_term($j);
2918 push @elements, \%hashscan;
2926 =head2 _ZOOM_event_loop
2928 _ZOOM_event_loop(\@zconns, \@results, sub {
2929 my ( $i, $size ) = @_;
2933 Processes a ZOOM event loop and passes control to a closure for
2934 processing the results, and destroying the resultsets.
2938 sub _ZOOM_event_loop {
2939 my ($zconns, $results, $callback) = @_;
2940 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2941 my $ev = $zconns->[ $i - 1 ]->last_event();
2942 if ( $ev == ZOOM::Event::ZEND ) {
2943 next unless $results->[ $i - 1 ];
2944 my $size = $results->[ $i - 1 ]->size();
2946 $callback->($i, $size);
2951 foreach my $result (@$results) {
2957 END { } # module clean-up code here (global destructor)
2964 Koha Development Team <http://koha-community.org/>