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; # CheckReserves
38 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
40 # set the version for version checking
42 $VERSION = 3.07.00.049;
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
72 &enabled_staff_search_views
76 # make all your functions, whether exported or not;
80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
88 my $dbh = C4::Context->dbh;
89 my $result = TransformMarcToKoha( $dbh, $record, '' );
94 my ( $biblionumber, $title );
96 # search duplicate on ISBN, easy and fast..
98 if ( $result->{isbn} ) {
99 $result->{isbn} =~ s/\(.*$//;
100 $result->{isbn} =~ s/\s+$//;
101 $query = "isbn=$result->{isbn}";
104 $result->{title} =~ s /\\//g;
105 $result->{title} =~ s /\"//g;
106 $result->{title} =~ s /\(//g;
107 $result->{title} =~ s /\)//g;
109 # FIXME: instead of removing operators, could just do
110 # quotes around the value
111 $result->{title} =~ s/(and|or|not)//g;
112 $query = "ti,ext=$result->{title}";
113 $query .= " and itemtype=$result->{itemtype}"
114 if ( $result->{itemtype} );
115 if ( $result->{author} ) {
116 $result->{author} =~ s /\\//g;
117 $result->{author} =~ s /\"//g;
118 $result->{author} =~ s /\(//g;
119 $result->{author} =~ s /\)//g;
121 # remove valid operators
122 $result->{author} =~ s/(and|or|not)//g;
123 $query .= " and au,ext=$result->{author}";
127 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
129 if (!defined $error) {
130 foreach my $possible_duplicate_record (@{$searchresults}) {
132 MARC::Record->new_from_usmarc($possible_duplicate_record);
133 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
135 # FIXME :: why 2 $biblionumber ?
137 push @results, $result->{'biblionumber'};
138 push @results, $result->{'title'};
147 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
149 This function provides a simple search API on the bibliographic catalog
155 * $query can be a simple keyword or a complete CCL query
156 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
157 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
158 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
163 Returns an array consisting of three elements
164 * $error is undefined unless an error is detected
165 * $results is a reference to an array of records.
166 * $total_hits is the number of hits that would have been returned with no limit
168 If an error is returned the two other return elements are undefined. If error itself is undefined
169 the other two elements are always defined
171 =item C<usage in the script:>
175 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
177 if (defined $error) {
178 $template->param(query_error => $error);
179 warn "error: ".$error;
180 output_html_with_http_headers $input, $cookie, $template->output;
184 my $hits = @{$marcresults};
187 for my $r ( @{$marcresults} ) {
188 my $marcrecord = MARC::File::USMARC::decode($r);
189 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
191 #build the iarray of hashs for the template.
193 title => $biblio->{'title'},
194 subtitle => $biblio->{'subtitle'},
195 biblionumber => $biblio->{'biblionumber'},
196 author => $biblio->{'author'},
197 publishercode => $biblio->{'publishercode'},
198 publicationyear => $biblio->{'publicationyear'},
203 $template->param(result=>\@results);
208 my ( $query, $offset, $max_results, $servers ) = @_;
210 if ( C4::Context->preference('NoZebra') ) {
211 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
214 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
215 return ( undef, $search_result, scalar($result->{hits}) );
218 return ( 'No query entered', undef, undef ) unless $query;
219 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
220 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
227 # Initialize & Search Zebra
228 for ( my $i = 0 ; $i < @servers ; $i++ ) {
230 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
231 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
232 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
236 $zconns[$i]->errmsg() . " ("
237 . $zconns[$i]->errcode() . ") "
238 . $zconns[$i]->addinfo() . " "
239 . $zconns[$i]->diagset();
241 return ( $error, undef, undef ) if $zconns[$i]->errcode();
245 # caught a ZOOM::Exception
249 . $@->addinfo() . " "
252 return ( $error, undef, undef );
255 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
256 my $event = $zconns[ $i - 1 ]->last_event();
257 if ( $event == ZOOM::Event::ZEND ) {
259 my $first_record = defined( $offset ) ? $offset+1 : 1;
260 my $hits = $tmpresults[ $i - 1 ]->size();
261 $total_hits += $hits;
262 my $last_record = $hits;
263 if ( defined $max_results && $offset + $max_results < $hits ) {
264 $last_record = $offset + $max_results;
267 for my $j ( $first_record..$last_record ) {
268 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
269 push @{$results}, $record;
274 foreach my $result (@tmpresults) {
277 foreach my $zoom_query (@zoom_queries) {
278 $zoom_query->destroy();
281 return ( undef, $results, $total_hits );
287 ( undef, $results_hashref, \@facets_loop ) = getRecords (
289 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
290 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
294 The all singing, all dancing, multi-server, asynchronous, scanning,
295 searching, record nabbing, facet-building
297 See verbse embedded documentation.
303 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
304 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
308 my @servers = @$servers_ref;
309 my @sort_by = @$sort_by_ref;
311 # Initialize variables for the ZOOM connection and results object
315 my $results_hashref = ();
317 # Initialize variables for the faceted results objects
318 my $facets_counter = ();
319 my $facets_info = ();
320 my $facets = getFacets();
321 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
323 my @facets_loop; # stores the ref to array of hashes for template facets loop
325 ### LOOP THROUGH THE SERVERS
326 for ( my $i = 0 ; $i < @servers ; $i++ ) {
327 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
329 # perform the search, create the results objects
330 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
331 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
333 #$query_to_use = $simple_query if $scan;
334 warn $simple_query if ( $scan and $DEBUG );
336 # Check if we've got a query_type defined, if so, use it
339 if ($query_type =~ /^ccl/) {
340 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
341 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
342 } elsif ($query_type =~ /^cql/) {
343 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
344 } elsif ($query_type =~ /^pqf/) {
345 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
347 warn "Unknown query_type '$query_type'. Results undetermined.";
350 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
352 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
356 warn "WARNING: query problem with $query_to_use " . $@;
359 # Concatenate the sort_by limits and pass them to the results object
360 # Note: sort will override rank
362 foreach my $sort (@sort_by) {
363 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
364 $sort_by .= "1=1003 <i ";
366 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
367 $sort_by .= "1=1003 >i ";
369 elsif ( $sort eq "popularity_asc" ) {
370 $sort_by .= "1=9003 <i ";
372 elsif ( $sort eq "popularity_dsc" ) {
373 $sort_by .= "1=9003 >i ";
375 elsif ( $sort eq "call_number_asc" ) {
376 $sort_by .= "1=8007 <i ";
378 elsif ( $sort eq "call_number_dsc" ) {
379 $sort_by .= "1=8007 >i ";
381 elsif ( $sort eq "pubdate_asc" ) {
382 $sort_by .= "1=31 <i ";
384 elsif ( $sort eq "pubdate_dsc" ) {
385 $sort_by .= "1=31 >i ";
387 elsif ( $sort eq "acqdate_asc" ) {
388 $sort_by .= "1=32 <i ";
390 elsif ( $sort eq "acqdate_dsc" ) {
391 $sort_by .= "1=32 >i ";
393 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
394 $sort_by .= "1=4 <i ";
396 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
397 $sort_by .= "1=4 >i ";
400 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
403 if ($sort_by && !$scan) {
404 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
405 warn "WARNING sort $sort_by failed";
408 } # finished looping through servers
410 # The big moment: asynchronously retrieve results from all servers
411 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
412 my $ev = $zconns[ $i - 1 ]->last_event();
413 if ( $ev == ZOOM::Event::ZEND ) {
414 next unless $results[ $i - 1 ];
415 my $size = $results[ $i - 1 ]->size();
419 # loop through the results
420 $results_hash->{'hits'} = $size;
422 if ( $offset + $results_per_page <= $size ) {
423 $times = $offset + $results_per_page;
428 for ( my $j = $offset ; $j < $times ; $j++ ) {
432 ## Check if it's an index scan
434 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
436 # here we create a minimal MARC record and hand it off to the
437 # template just like a normal result ... perhaps not ideal, but
439 my $tmprecord = MARC::Record->new();
440 $tmprecord->encoding('UTF-8');
444 # the minimal record in author/title (depending on MARC flavour)
445 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
446 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
447 $tmprecord->append_fields($tmptitle);
449 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
450 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
451 $tmprecord->append_fields($tmptitle);
452 $tmprecord->append_fields($tmpauthor);
454 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
459 $record = $results[ $i - 1 ]->record($j)->raw();
461 # warn "RECORD $j:".$record;
462 $results_hash->{'RECORDS'}[$j] = $record;
466 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
468 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
469 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
471 my $jmax = $size>$facets_maxrecs? $facets_maxrecs: $size;
472 for my $facet ( @$facets ) {
473 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
474 my $render_record = $results[ $i - 1 ]->record($j)->render();
476 foreach my $tag ( @{$facet->{tags}} ) {
478 my $tag_num = substr($tag, 0, 3);
479 my $letters = substr($tag, 3);
480 my $field_pattern = '\n' . $tag_num . ' ([^\n]+)';
481 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
482 foreach my $field_token (@field_tokens) {
483 my @subf = ( $field_token =~ /\$([a-zA-Z0-9]) ([^\$]+)/g );
485 for (my $i = 0; $i < @subf; $i += 2) {
486 if ( $letters =~ $subf[$i] ) {
487 my $value = $subf[$i+1];
490 push @values, $value;
493 my $data = join($facet->{sep}, @values);
494 unless ( $data ~~ @used_datas ) {
495 $facets_counter->{ $facet->{idx} }->{$data}++;
496 push @used_datas, $data;
501 $facets_info->{ $facet->{idx} }->{label_value} = $facet->{label};
502 $facets_info->{ $facet->{idx} }->{expanded} = $facet->{expanded};
507 # warn "connection ", $i-1, ": $size hits";
508 # warn $results[$i-1]->record(0)->render() if $size > 0;
511 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
513 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
514 keys %$facets_counter )
517 my $number_of_facets;
518 my @this_facets_array;
521 $facets_counter->{$link_value}->{$b}
522 <=> $facets_counter->{$link_value}->{$a}
523 } keys %{ $facets_counter->{$link_value} }
527 if ( ( $number_of_facets < 6 )
528 || ( $expanded_facet eq $link_value )
529 || ( $facets_info->{$link_value}->{'expanded'} ) )
532 # Sanitize the link value ), ( will cause errors with CCL,
533 my $facet_link_value = $one_facet;
534 $facet_link_value =~ s/(\(|\))/ /g;
536 # fix the length that will display in the label,
537 my $facet_label_value = $one_facet;
538 my $facet_max_length =
539 C4::Context->preference('FacetLabelTruncationLength') || 20;
541 substr( $one_facet, 0, $facet_max_length ) . "..."
542 if length($facet_label_value) > $facet_max_length;
544 # if it's a branch, label by the name, not the code,
545 if ( $link_value =~ /branch/ ) {
546 if (defined $branches
547 && ref($branches) eq "HASH"
548 && defined $branches->{$one_facet}
549 && ref ($branches->{$one_facet}) eq "HASH")
552 $branches->{$one_facet}->{'branchname'};
555 $facet_label_value = "*";
558 # if it's a itemtype, label by the name, not the code,
559 if ( $link_value =~ /itype/ ) {
560 if (defined $itemtypes
561 && ref($itemtypes) eq "HASH"
562 && defined $itemtypes->{$one_facet}
563 && ref ($itemtypes->{$one_facet}) eq "HASH")
566 $itemtypes->{$one_facet}->{'description'};
570 # but we're down with the whole label being in the link's title.
571 push @this_facets_array, {
572 facet_count => $facets_counter->{$link_value}->{$one_facet},
573 facet_label_value => $facet_label_value,
574 facet_title_value => $one_facet,
575 facet_link_value => $facet_link_value,
576 type_link_value => $link_value,
581 # handle expanded option
582 unless ( $facets_info->{$link_value}->{'expanded'} ) {
584 if ( ( $number_of_facets > 6 )
585 && ( $expanded_facet ne $link_value ) );
588 type_link_value => $link_value,
589 type_id => $link_value . "_id",
590 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
591 facets => \@this_facets_array,
592 expandable => $expandable,
593 expand => $link_value,
594 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
599 return ( undef, $results_hashref, \@facets_loop );
604 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
605 $results_per_page, $offset, $expanded_facet, $branches,
609 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
611 $paz->search($simple_query);
612 sleep 1; # FIXME: WHY?
615 my $results_hashref = {};
616 my $stats = XMLin($paz->stat);
617 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
619 # for a grouped search result, the number of hits
620 # is the number of groups returned; 'bib_hits' will have
621 # the total number of bibs.
622 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
623 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
625 HIT: foreach my $hit (@{ $results->{'hit'} }) {
626 my $recid = $hit->{recid}->[0];
628 my $work_title = $hit->{'md-work-title'}->[0];
630 if (exists $hit->{'md-work-author'}) {
631 $work_author = $hit->{'md-work-author'}->[0];
633 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
635 my $result_group = {};
636 $result_group->{'group_label'} = $group_label;
637 $result_group->{'group_merge_key'} = $recid;
640 if (exists $hit->{count}) {
641 $count = $hit->{count}->[0];
643 $result_group->{'group_count'} = $count;
645 for (my $i = 0; $i < $count; $i++) {
646 # FIXME -- may need to worry about diacritics here
647 my $rec = $paz->record($recid, $i);
648 push @{ $result_group->{'RECORDS'} }, $rec;
651 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
654 # pass through facets
655 my $termlist_xml = $paz->termlist('author,subject');
656 my $terms = XMLin($termlist_xml, forcearray => 1);
657 my @facets_loop = ();
658 #die Dumper($results);
659 # foreach my $list (sort keys %{ $terms->{'list'} }) {
661 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
663 # facet_label_value => $facet->{'name'}->[0],
666 # push @facets_loop, ( {
667 # type_label => $list,
668 # facets => \@facets,
672 return ( undef, $results_hashref, \@facets_loop );
676 sub _remove_stopwords {
677 my ( $operand, $index ) = @_;
678 my @stopwords_removed;
680 # phrase and exact-qualified indexes shouldn't have stopwords removed
681 if ( $index !~ m/phr|ext/ ) {
683 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
684 # we use IsAlpha unicode definition, to deal correctly with diacritics.
685 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
686 # is a stopword, we'd get "çon" and wouldn't find anything...
688 foreach ( keys %{ C4::Context->stopwords } ) {
689 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
690 if ( my ($matched) = ($operand =~
691 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
693 $operand =~ s/\Q$matched\E/ /gi;
694 push @stopwords_removed, $_;
698 return ( $operand, \@stopwords_removed );
702 sub _detect_truncation {
703 my ( $operand, $index ) = @_;
704 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
707 my @wordlist = split( /\s/, $operand );
708 foreach my $word (@wordlist) {
709 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
710 push @rightlefttruncated, $word;
712 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
713 push @lefttruncated, $word;
715 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
716 push @righttruncated, $word;
718 elsif ( index( $word, "*" ) < 0 ) {
719 push @nontruncated, $word;
722 push @regexpr, $word;
726 \@nontruncated, \@righttruncated, \@lefttruncated,
727 \@rightlefttruncated, \@regexpr
732 sub _build_stemmed_operand {
733 my ($operand,$lang) = @_;
734 require Lingua::Stem::Snowball ;
735 my $stemmed_operand=q{};
737 # If operand contains a digit, it is almost certainly an identifier, and should
738 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
739 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
740 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
741 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
742 return $operand if $operand =~ /\d/;
744 # FIXME: the locale should be set based on the user's language and/or search choice
746 # Make sure we only use the first two letters from the language code
747 $lang = lc(substr($lang, 0, 2));
748 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
749 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
750 if ($lang eq 'nb' || $lang eq 'nn') {
753 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
754 encoding => "UTF-8" );
756 my @words = split( / /, $operand );
757 my @stems = $stemmer->stem(\@words);
758 for my $stem (@stems) {
759 $stemmed_operand .= "$stem";
760 $stemmed_operand .= "?"
761 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
762 $stemmed_operand .= " ";
764 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
765 return $stemmed_operand;
769 sub _build_weighted_query {
771 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
772 # pretty well but could work much better if we had a smarter query parser
773 my ( $operand, $stemmed_operand, $index ) = @_;
774 my $stemming = C4::Context->preference("QueryStemming") || 0;
775 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
776 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
778 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
780 # Keyword, or, no index specified
781 if ( ( $index eq 'kw' ) || ( !$index ) ) {
783 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
784 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
785 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
786 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
787 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
788 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
789 if $fuzzy_enabled; # add fuzzy, word list
790 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
791 if ( $stemming and $stemmed_operand )
792 ; # add stemming, right truncation
793 $weighted_query .= " or wrdl,r9=\"$operand\"";
795 # embedded sorting: 0 a-z; 1 z-a
796 # $weighted_query .= ") or (sort1,aut=1";
799 # Barcode searches should skip this process
800 elsif ( $index eq 'bc' ) {
801 $weighted_query .= "bc=\"$operand\"";
804 # Authority-number searches should skip this process
805 elsif ( $index eq 'an' ) {
806 $weighted_query .= "an=\"$operand\"";
809 # If the index already has more than one qualifier, wrap the operand
810 # in quotes and pass it back (assumption is that the user knows what they
811 # are doing and won't appreciate us mucking up their query
812 elsif ( $index =~ ',' ) {
813 $weighted_query .= " $index=\"$operand\"";
816 #TODO: build better cases based on specific search indexes
818 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
819 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
820 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
822 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
825 $weighted_query .= "))"; # close rank specification
826 return $weighted_query;
831 Return an array with available indexes.
853 'Author-personal-bibliography',
863 'Chronological-subdivision',
873 'Conference-name-heading',
874 'Conference-name-see',
875 'Conference-name-seealso',
880 'Corporate-name-heading',
881 'Corporate-name-see',
882 'Corporate-name-seealso',
884 'date-entered-on-file',
885 'Date-of-acquisition',
886 'Date-of-publication',
887 'Dewey-classification',
894 'Geographic-subdivision',
897 'Heading-use-main-or-added-entry',
898 'Heading-use-series-added-entry ',
899 'Heading-use-subject-added-entry',
917 'Local-classification',
920 'Match-heading-see-from',
928 'Name-geographic-heading',
929 'Name-geographic-see',
930 'Name-geographic-seealso',
938 'Personal-name-heading',
940 'Personal-name-seealso',
947 'Record-control-number',
958 'Subject-heading-thesaurus',
959 'Subject-name-personal',
960 'Subject-subdivision',
970 'Term-genre-form-heading',
971 'Term-genre-form-see',
972 'Term-genre-form-seealso',
979 'Title-uniform-heading',
981 'Title-uniform-seealso',
991 'classification-source',
993 'coded-location-qualifier',
1004 'Local-classification',
1007 'materials-specified',
1016 'replacementpricedate',
1034 $simple_query, $query_cgi,
1035 $query_desc, $limit,
1036 $limit_cgi, $limit_desc,
1037 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1039 Build queries and limits in CCL, CGI, Human,
1040 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1042 See verbose embedded documentation.
1048 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1050 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1053 my @operators = $operators ? @$operators : ();
1054 my @indexes = $indexes ? @$indexes : ();
1055 my @operands = $operands ? @$operands : ();
1056 my @limits = $limits ? @$limits : ();
1057 my @sort_by = $sort_by ? @$sort_by : ();
1059 my $stemming = C4::Context->preference("QueryStemming") || 0;
1060 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1061 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1062 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1063 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1065 # no stemming/weight/fuzzy in NoZebra
1066 if ( C4::Context->preference("NoZebra") ) {
1070 $auto_truncation = 0;
1073 my $query = $operands[0];
1074 my $simple_query = $operands[0];
1076 # initialize the variables we're passing back
1085 my $stopwords_removed; # flag to determine if stopwords have been removed
1088 my $cclindexes = getIndexes();
1089 if ( $query !~ /\s*ccl=/ ) {
1090 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1092 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1094 $query = "ccl=$query" if $cclq;
1097 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1099 if ( $query =~ /^ccl=/ ) {
1101 # This is needed otherwise ccl= and &limit won't work together, and
1102 # this happens when selecting a subject on the opac-detail page
1104 $q .= ' and '.join(' and ', @limits);
1106 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1108 if ( $query =~ /^cql=/ ) {
1109 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1111 if ( $query =~ /^pqf=/ ) {
1112 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1115 # pass nested queries directly
1116 # FIXME: need better handling of some of these variables in this case
1117 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1118 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1119 # if ( $query =~ /(\(|\))/ ) {
1121 # undef, $query, $simple_query, $query_cgi,
1122 # $query, $limit, $limit_cgi, $limit_desc,
1123 # $stopwords_removed, 'ccl'
1127 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1128 # query operands and indexes and add stemming, truncation, field weighting, etc.
1129 # Once we do so, we'll end up with a value in $query, just like if we had an
1130 # incoming $query from the user
1133 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1134 my $previous_operand
1135 ; # a flag used to keep track if there was a previous query
1136 # if there was, we can apply the current operator
1138 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1140 # COMBINE OPERANDS, INDEXES AND OPERATORS
1141 if ( $operands[$i] ) {
1142 $operands[$i]=~s/^\s+//;
1144 # A flag to determine whether or not to add the index to the query
1147 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1148 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1151 $remove_stopwords = 0;
1153 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1155 my $operand = $operands[$i];
1156 my $index = $indexes[$i];
1158 # Add index-specific attributes
1159 # Date of Publication
1160 if ( $index eq 'yr' ) {
1161 $index .= ",st-numeric";
1163 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1166 # Date of Acquisition
1167 elsif ( $index eq 'acqdate' ) {
1168 $index .= ",st-date-normalized";
1170 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1172 # ISBN,ISSN,Standard Number, don't need special treatment
1173 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1175 $stemming, $auto_truncation,
1176 $weight_fields, $fuzzy_enabled,
1178 ) = ( 0, 0, 0, 0, 0 );
1186 # Set default structure attribute (word list)
1187 my $struct_attr = q{};
1188 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1189 $struct_attr = ",wrdl";
1192 # Some helpful index variants
1193 my $index_plus = $index . $struct_attr . ':';
1194 my $index_plus_comma = $index . $struct_attr . ',';
1197 if ($remove_stopwords) {
1198 ( $operand, $stopwords_removed ) =
1199 _remove_stopwords( $operand, $index );
1200 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1201 warn "REMOVED STOPWORDS: @$stopwords_removed"
1202 if ( $stopwords_removed && $DEBUG );
1205 if ($auto_truncation){
1206 unless ( $index =~ /(st-|phr|ext)/ ) {
1207 #FIXME only valid with LTR scripts
1208 $operand=join(" ",map{
1209 (index($_,"*")>0?"$_":"$_*")
1210 }split (/\s+/,$operand));
1211 warn $operand if $DEBUG;
1216 my $truncated_operand;
1217 my( $nontruncated, $righttruncated, $lefttruncated,
1218 $rightlefttruncated, $regexpr
1219 ) = _detect_truncation( $operand, $index );
1221 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1226 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1227 scalar(@$rightlefttruncated) > 0 )
1230 # Don't field weight or add the index to the query, we do it here
1232 undef $weight_fields;
1233 my $previous_truncation_operand;
1234 if (scalar @$nontruncated) {
1235 $truncated_operand .= "$index_plus @$nontruncated ";
1236 $previous_truncation_operand = 1;
1238 if (scalar @$righttruncated) {
1239 $truncated_operand .= "and " if $previous_truncation_operand;
1240 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1241 $previous_truncation_operand = 1;
1243 if (scalar @$lefttruncated) {
1244 $truncated_operand .= "and " if $previous_truncation_operand;
1245 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1246 $previous_truncation_operand = 1;
1248 if (scalar @$rightlefttruncated) {
1249 $truncated_operand .= "and " if $previous_truncation_operand;
1250 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1251 $previous_truncation_operand = 1;
1254 $operand = $truncated_operand if $truncated_operand;
1255 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1258 my $stemmed_operand;
1259 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1262 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1264 # Handle Field Weighting
1265 my $weighted_operand;
1266 if ($weight_fields) {
1267 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1268 $operand = $weighted_operand;
1272 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1274 # If there's a previous operand, we need to add an operator
1275 if ($previous_operand) {
1277 # User-specified operator
1278 if ( $operators[ $i - 1 ] ) {
1279 $query .= " $operators[$i-1] ";
1280 $query .= " $index_plus " unless $indexes_set;
1281 $query .= " $operand";
1282 $query_cgi .= "&op=$operators[$i-1]";
1283 $query_cgi .= "&idx=$index" if $index;
1284 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1286 " $operators[$i-1] $index_plus $operands[$i]";
1289 # Default operator is and
1292 $query .= "$index_plus " unless $indexes_set;
1293 $query .= "$operand";
1294 $query_cgi .= "&op=and&idx=$index" if $index;
1295 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1296 $query_desc .= " and $index_plus $operands[$i]";
1300 # There isn't a pervious operand, don't need an operator
1303 # Field-weighted queries already have indexes set
1304 $query .= " $index_plus " unless $indexes_set;
1306 $query_desc .= " $index_plus $operands[$i]";
1307 $query_cgi .= "&idx=$index" if $index;
1308 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1309 $previous_operand = 1;
1314 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1317 my %group_OR_limits;
1318 my $availability_limit;
1319 foreach my $this_limit (@limits) {
1320 if ( $this_limit =~ /available/ ) {
1322 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1324 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1325 $availability_limit .=
1326 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1327 $limit_cgi .= "&limit=available";
1331 # group_OR_limits, prefixed by mc-
1332 # OR every member of the group
1333 elsif ( $this_limit =~ /mc/ ) {
1334 my ($k,$v) = split(/:/, $this_limit,2);
1335 if ( $k !~ /mc-i(tem)?type/ ) {
1336 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1337 $this_limit =~ tr/"//d;
1338 $this_limit = $k.":\"".$v."\"";
1341 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1342 $limit_desc .= " or " if $group_OR_limits{$k};
1343 $group_OR_limits{$k} .= "$this_limit";
1344 $limit_cgi .= "&limit=$this_limit";
1345 $limit_desc .= " $this_limit";
1348 # Regular old limits
1350 $limit .= " and " if $limit || $query;
1351 $limit .= "$this_limit";
1352 $limit_cgi .= "&limit=$this_limit";
1353 if ($this_limit =~ /^branch:(.+)/) {
1354 my $branchcode = $1;
1355 my $branchname = GetBranchName($branchcode);
1356 if (defined $branchname) {
1357 $limit_desc .= " branch:$branchname";
1359 $limit_desc .= " $this_limit";
1362 $limit_desc .= " $this_limit";
1366 foreach my $k (keys (%group_OR_limits)) {
1367 $limit .= " and " if ( $query || $limit );
1368 $limit .= "($group_OR_limits{$k})";
1370 if ($availability_limit) {
1371 $limit .= " and " if ( $query || $limit );
1372 $limit .= "($availability_limit)";
1375 # Normalize the query and limit strings
1376 # This is flawed , means we can't search anything with : in it
1377 # if user wants to do ccl or cql, start the query with that
1378 # $query =~ s/:/=/g;
1379 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1380 $query =~ s/(?<=(wrdl)):/=/g;
1381 $query =~ s/(?<=(trn|phr)):/=/g;
1383 for ( $query, $query_desc, $limit, $limit_desc ) {
1384 s/ +/ /g; # remove extra spaces
1385 s/^ //g; # remove any beginning spaces
1386 s/ $//g; # remove any ending spaces
1387 s/==/=/g; # remove double == from query
1389 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1391 for ($query_cgi,$simple_query) {
1394 # append the limit to the query
1395 $query .= " " . $limit;
1399 warn "QUERY:" . $query;
1400 warn "QUERY CGI:" . $query_cgi;
1401 warn "QUERY DESC:" . $query_desc;
1402 warn "LIMIT:" . $limit;
1403 warn "LIMIT CGI:" . $limit_cgi;
1404 warn "LIMIT DESC:" . $limit_desc;
1405 warn "---------\nLeave buildQuery\n---------";
1408 undef, $query, $simple_query, $query_cgi,
1409 $query_desc, $limit, $limit_cgi, $limit_desc,
1410 $stopwords_removed, $query_type
1414 =head2 searchResults
1416 my @search_results = searchResults($search_context, $searchdesc, $hits,
1417 $results_per_page, $offset, $scan,
1418 @marcresults, $hidelostitems);
1420 Format results in a form suitable for passing to the template
1424 # IMO this subroutine is pretty messy still -- it's responsible for
1425 # building the HTML output for the template
1427 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1428 my $dbh = C4::Context->dbh;
1433 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1434 my ($is_opac, $hidelostitems);
1435 if ($search_context eq 'opac') {
1436 $hidelostitems = C4::Context->preference('hidelostitems');
1440 #Build branchnames hash
1442 #get branch information.....
1444 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1446 while ( my $bdata = $bsth->fetchrow_hashref ) {
1447 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1449 # FIXME - We build an authorised values hash here, using the default framework
1450 # though it is possible to have different authvals for different fws.
1452 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1454 # get notforloan authorised value list (see $shelflocations FIXME)
1455 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1457 #Build itemtype hash
1458 #find itemtype & itemtype image
1462 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1465 while ( my $bdata = $bsth->fetchrow_hashref ) {
1466 foreach (qw(description imageurl summary notforloan)) {
1467 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1471 #search item field code
1472 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1474 ## find column names of items related to MARC
1475 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1477 my %subfieldstosearch;
1478 while ( ( my $column ) = $sth2->fetchrow ) {
1479 my ( $tagfield, $tagsubfield ) =
1480 &GetMarcFromKohaField( "items." . $column, "" );
1481 $subfieldstosearch{$column} = $tagsubfield;
1484 # handle which records to actually retrieve
1486 if ( $hits && $offset + $results_per_page <= $hits ) {
1487 $times = $offset + $results_per_page;
1490 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1493 my $marcflavour = C4::Context->preference("marcflavour");
1494 # We get the biblionumber position in MARC
1495 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1497 # loop through all of the records we've retrieved
1498 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1499 my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1503 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1504 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1505 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1506 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1507 $oldbiblio->{result_number} = $i + 1;
1509 # add imageurl to itemtype if there is one
1510 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1512 $oldbiblio->{'authorised_value_images'} = ($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 ) ) : [];
1513 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1514 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1515 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1516 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1517 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1519 # edition information, if any
1520 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1521 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1522 # Build summary if there is one (the summary is defined in the itemtypes table)
1523 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1524 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1525 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1526 my @fields = $marcrecord->fields();
1529 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1531 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1532 $tag =~ /(.{3})(.)/;
1533 if($marcrecord->field($1)){
1534 my @abc = $marcrecord->field($1)->subfield($2);
1535 $tags->{$tag} = $#abc + 1 ;
1539 # We catch how many times to repeat this line
1541 foreach my $tag (keys(%$tags)){
1542 $max = $tags->{$tag} if($tags->{$tag} > $max);
1545 # we replace, and repeat each line
1546 for (my $i = 0 ; $i < $max ; $i++){
1547 my $newline = $line;
1549 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1550 $tag =~ /(.{3})(.)/;
1552 if($marcrecord->field($1)){
1553 my @repl = $marcrecord->field($1)->subfield($2);
1554 my $subfieldvalue = $repl[$i];
1556 if (! utf8::is_utf8($subfieldvalue)) {
1557 utf8::decode($subfieldvalue);
1560 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1563 $newsummary .= "$newline\n";
1567 $newsummary =~ s/\[(.*?)]//g;
1568 $newsummary =~ s/\n/<br\/>/g;
1569 $oldbiblio->{summary} = $newsummary;
1572 # Pull out the items fields
1573 my @fields = $marcrecord->field($itemtag);
1574 my $marcflavor = C4::Context->preference("marcflavour");
1575 # adding linked items that belong to host records
1576 my $analyticsfield = '773';
1577 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1578 $analyticsfield = '773';
1579 } elsif ($marcflavor eq 'UNIMARC') {
1580 $analyticsfield = '461';
1582 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1583 my $hostbiblionumber = $hostfield->subfield("0");
1584 my $linkeditemnumber = $hostfield->subfield("9");
1585 if(!$hostbiblionumber eq undef){
1586 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1587 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1588 if(!$hostbiblio eq undef){
1589 my @hostitems = $hostbiblio->field($itemfield);
1590 foreach my $hostitem (@hostitems){
1591 if ($hostitem->subfield("9") eq $linkeditemnumber){
1592 my $linkeditem =$hostitem;
1593 # append linked items if they exist
1594 if (!$linkeditem eq undef){
1595 push (@fields, $linkeditem);}
1602 # Setting item statuses for display
1603 my @available_items_loop;
1604 my @onloan_items_loop;
1605 my @other_items_loop;
1607 my $available_items;
1611 my $ordered_count = 0;
1612 my $available_count = 0;
1613 my $onloan_count = 0;
1614 my $longoverdue_count = 0;
1615 my $other_count = 0;
1616 my $wthdrawn_count = 0;
1617 my $itemlost_count = 0;
1618 my $hideatopac_count = 0;
1619 my $itembinding_count = 0;
1620 my $itemdamaged_count = 0;
1621 my $item_in_transit_count = 0;
1622 my $can_place_holds = 0;
1623 my $item_onhold_count = 0;
1624 my $items_count = scalar(@fields);
1625 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1626 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1628 # loop through every item
1630 foreach my $field (@fields) {
1633 # populate the items hash
1634 foreach my $code ( keys %subfieldstosearch ) {
1635 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1637 $item->{description} = $itemtypes{ $item->{itype} }{description};
1641 my @hi = C4::Items::GetHiddenItemnumbers($item);
1642 $item->{'hideatopac'} = @hi;
1643 push @hiddenitems, @hi;
1646 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1647 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1649 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1650 if ($item->{$hbranch}) {
1651 $item->{'branchname'} = $branches{$item->{$hbranch}};
1653 elsif ($item->{$otherbranch}) { # Last resort
1654 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1657 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1658 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1659 my $userenv = C4::Context->userenv;
1660 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1662 my $key = $prefix . $item->{onloan} . $item->{barcode};
1663 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1664 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1665 $onloan_items->{$key}->{branchname} = $item->{branchname};
1666 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1667 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1668 $onloan_items->{$key}->{description} = $item->{description};
1669 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1670 # if something's checked out and lost, mark it as 'long overdue'
1671 if ( $item->{itemlost} ) {
1672 $onloan_items->{$prefix}->{longoverdue}++;
1673 $longoverdue_count++;
1674 } else { # can place holds as long as item isn't lost
1675 $can_place_holds = 1;
1679 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1683 if ( $item->{notforloan} == -1 ) {
1687 # is item in transit?
1688 my $transfertwhen = '';
1689 my ($transfertfrom, $transfertto);
1691 # is item on the reserve shelf?
1692 my $reservestatus = '';
1695 unless ($item->{wthdrawn}
1696 || $item->{itemlost}
1698 || $item->{notforloan}
1699 || $items_count > 20) {
1701 # A couple heuristics to limit how many times
1702 # we query the database for item transfer information, sacrificing
1703 # accuracy in some cases for speed;
1705 # 1. don't query if item has one of the other statuses
1706 # 2. don't check transit status if the bib has
1707 # more than 20 items
1709 # FIXME: to avoid having the query the database like this, and to make
1710 # the in transit status count as unavailable for search limiting,
1711 # should map transit status to record indexed in Zebra.
1713 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1714 ($reservestatus, $reserveitem, undef) = C4::Reserves::CheckReserves($item->{itemnumber});
1717 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1718 if ( $item->{wthdrawn}
1719 || $item->{itemlost}
1721 || $item->{notforloan} > 0
1722 || $item->{hideatopac}
1723 || $reservestatus eq 'Waiting'
1724 || ($transfertwhen ne ''))
1726 $wthdrawn_count++ if $item->{wthdrawn};
1727 $itemlost_count++ if $item->{itemlost};
1728 $itemdamaged_count++ if $item->{damaged};
1729 $hideatopac_count++ if $item->{hideatopac};
1730 $item_in_transit_count++ if $transfertwhen ne '';
1731 $item_onhold_count++ if $reservestatus eq 'Waiting';
1732 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1734 # can place hold on item ?
1735 if ((!$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems'))
1736 && !$item->{itemlost}
1737 && !$item->{withdrawn}
1739 $can_place_holds = 1;
1744 my $key = $prefix . $item->{status};
1745 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber hideatopac)) {
1746 $other_items->{$key}->{$_} = $item->{$_};
1748 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1749 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1750 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1751 $other_items->{$key}->{count}++ if $item->{$hbranch};
1752 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1753 $other_items->{$key}->{description} = $item->{description};
1754 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1758 $can_place_holds = 1;
1760 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1761 foreach (qw(branchname itemcallnumber hideatopac description)) {
1762 $available_items->{$prefix}->{$_} = $item->{$_};
1764 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1765 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1768 } # notforloan, item level and biblioitem level
1769 if ($items_count > 0) {
1770 next if $is_opac && $hideatopac_count >= $items_count;
1771 next if $hidelostitems && $itemlost_count >= $items_count;
1773 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1774 for my $key ( sort keys %$onloan_items ) {
1775 (++$onloanitemscount > $maxitems) and last;
1776 push @onloan_items_loop, $onloan_items->{$key};
1778 for my $key ( sort keys %$other_items ) {
1779 (++$otheritemscount > $maxitems) and last;
1780 push @other_items_loop, $other_items->{$key};
1782 for my $key ( sort keys %$available_items ) {
1783 (++$availableitemscount > $maxitems) and last;
1784 push @available_items_loop, $available_items->{$key}
1787 # XSLT processing of some stuff
1789 SetUTF8Flag($marcrecord);
1790 warn $marcrecord->as_formatted if $DEBUG;
1791 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1792 if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
1793 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
1794 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1797 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1798 if (!C4::Context->preference("item-level_itypes")) {
1799 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1800 $can_place_holds = 0;
1803 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1804 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1805 $oldbiblio->{items_count} = $items_count;
1806 $oldbiblio->{available_items_loop} = \@available_items_loop;
1807 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1808 $oldbiblio->{other_items_loop} = \@other_items_loop;
1809 $oldbiblio->{availablecount} = $available_count;
1810 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1811 $oldbiblio->{onloancount} = $onloan_count;
1812 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1813 $oldbiblio->{othercount} = $other_count;
1814 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1815 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1816 $oldbiblio->{itemlostcount} = $itemlost_count;
1817 $oldbiblio->{damagedcount} = $itemdamaged_count;
1818 $oldbiblio->{intransitcount} = $item_in_transit_count;
1819 $oldbiblio->{onholdcount} = $item_onhold_count;
1820 $oldbiblio->{orderedcount} = $ordered_count;
1822 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1823 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1824 my $subfields = substr $fieldspec, 3;
1825 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1826 my @alternateholdingsinfo = ();
1827 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1828 my $alternateholdingscount = 0;
1830 for my $field (@holdingsfields) {
1831 my %holding = ( holding => '' );
1832 my $havesubfield = 0;
1833 for my $subfield ($field->subfields()) {
1834 if ((index $subfields, $$subfield[0]) >= 0) {
1835 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1836 $holding{'holding'} .= $$subfield[1];
1840 if ($havesubfield) {
1841 push(@alternateholdingsinfo, \%holding);
1842 $alternateholdingscount++;
1846 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1847 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1850 push( @newresults, $oldbiblio );
1856 =head2 SearchAcquisitions
1857 Search for acquisitions
1860 sub SearchAcquisitions{
1861 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1863 my $dbh=C4::Context->dbh;
1864 # Variable initialization
1868 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1869 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1870 WHERE dateaccessioned BETWEEN ? AND ?
1873 my (@params,@loopcriteria);
1875 push @params, $datebegin->output("iso");
1876 push @params, $dateend->output("iso");
1878 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1879 if(C4::Context->preference("item-level_itypes")){
1880 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1882 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1884 push @params, @$itemtypes;
1887 if ($criteria =~/itemtype/){
1888 if(C4::Context->preference("item-level_itypes")){
1889 $str .= "AND items.itype=? ";
1891 $str .= "AND biblioitems.itemtype=? ";
1894 if(scalar(@$itemtypes) == 0){
1895 my $itypes = GetItemTypes();
1896 for my $key (keys %$itypes){
1897 push @$itemtypes, $key;
1901 @loopcriteria= @$itemtypes;
1902 }elsif ($criteria=~/itemcallnumber/){
1903 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1904 OR items.itemcallnumber is NULL
1905 OR items.itemcallnumber = '')";
1907 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1909 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1910 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1913 if ($orderby =~ /date_desc/){
1914 $str.=" ORDER BY dateaccessioned DESC";
1916 $str.=" ORDER BY title";
1919 my $qdataacquisitions=$dbh->prepare($str);
1921 my @loopacquisitions;
1922 foreach my $value(@loopcriteria){
1923 push @params,$value;
1925 $cell{"title"}=$value;
1926 $cell{"titlecode"}=$value;
1928 eval{$qdataacquisitions->execute(@params);};
1930 if ($@){ warn "recentacquisitions Error :$@";}
1933 while (my $data=$qdataacquisitions->fetchrow_hashref){
1934 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1936 $cell{"loopdata"}=\@loopdata;
1938 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1941 $qdataacquisitions->finish;
1942 return \@loopacquisitions;
1944 #----------------------------------------------------------------------
1946 # Non-Zebra GetRecords#
1947 #----------------------------------------------------------------------
1951 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1957 $query, $simple_query, $sort_by_ref, $servers_ref,
1958 $results_per_page, $offset, $expanded_facet, $branches,
1961 warn "query =$query" if $DEBUG;
1962 my $result = NZanalyse($query);
1963 warn "results =$result" if $DEBUG;
1965 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1971 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1972 the list is built from an inverted index in the nozebra SQL table
1973 note that title is here only for convenience : the sorting will be very fast when requested on title
1974 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1979 my ( $string, $server ) = @_;
1980 # warn "---------" if $DEBUG;
1981 warn " NZanalyse" if $DEBUG;
1982 # warn "---------" if $DEBUG;
1984 # $server contains biblioserver or authorities, depending on what we search on.
1985 #warn "querying : $string on $server";
1986 $server = 'biblioserver' unless $server;
1988 # if we have a ", replace the content to discard temporarily any and/or/not inside
1990 if ( $string =~ /"/ ) {
1991 $string =~ s/"(.*?)"/__X__/;
1993 warn "commacontent : $commacontent" if $DEBUG;
1996 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1997 # then, call again NZanalyse with $left and $right
1998 # (recursive until we find a leaf (=> something without and/or/not)
1999 # delete repeated operator... Would then go in infinite loop
2000 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
2003 #process parenthesis before.
2004 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
2007 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
2009 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2011 my $leftresult = NZanalyse( $left, $server );
2013 my $rightresult = NZanalyse( $right, $server );
2015 # OK, we have the results for right and left part of the query
2016 # depending of operand, intersect, union or exclude both lists
2017 # to get a result list
2018 if ( $operator eq ' and ' ) {
2019 return NZoperatorAND($leftresult,$rightresult);
2021 elsif ( $operator eq ' or ' ) {
2023 # just merge the 2 strings
2024 return $leftresult . $rightresult;
2026 elsif ( $operator eq ' not ' ) {
2027 return NZoperatorNOT($leftresult,$rightresult);
2031 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2035 warn "string :" . $string if $DEBUG;
2039 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2042 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2044 warn "no parenthesis. left : $left operator: $operator right: $right"
2047 # it's not a leaf, we have a and/or/not
2050 # reintroduce comma content if needed
2051 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2052 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2053 warn "node : $left / $operator / $right\n" if $DEBUG;
2054 my $leftresult = NZanalyse( $left, $server );
2055 my $rightresult = NZanalyse( $right, $server );
2056 warn " leftresult : $leftresult" if $DEBUG;
2057 warn " rightresult : $rightresult" if $DEBUG;
2058 # OK, we have the results for right and left part of the query
2059 # depending of operand, intersect, union or exclude both lists
2060 # to get a result list
2061 if ( $operator eq ' and ' ) {
2062 return NZoperatorAND($leftresult,$rightresult);
2064 elsif ( $operator eq ' or ' ) {
2066 # just merge the 2 strings
2067 return $leftresult . $rightresult;
2069 elsif ( $operator eq ' not ' ) {
2070 return NZoperatorNOT($leftresult,$rightresult);
2074 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2075 die "error : operand unknown : $operator for $string";
2078 # it's a leaf, do the real SQL query and return the result
2081 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2082 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2083 #remove trailing blank at the beginning
2085 warn "leaf:$string" if $DEBUG;
2087 # parse the string in in operator/operand/value again
2091 if ($string =~ /(.*)(>=|<=)(.*)/) {
2098 # warn "handling leaf... left:$left operator:$operator right:$right"
2100 unless ($operator) {
2101 if ($string =~ /(.*)(>|<|=)(.*)/) {
2106 "handling unless (operator)... left:$left operator:$operator right:$right"
2114 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2117 # automatic replace for short operators
2118 $left = 'title' if $left =~ '^ti$';
2119 $left = 'author' if $left =~ '^au$';
2120 $left = 'publisher' if $left =~ '^pb$';
2121 $left = 'subject' if $left =~ '^su$';
2122 $left = 'koha-Auth-Number' if $left =~ '^an$';
2123 $left = 'keyword' if $left =~ '^kw$';
2124 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2125 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2126 my $dbh = C4::Context->dbh;
2127 if ( $operator && $left ne 'keyword' ) {
2128 #do a specific search
2129 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2130 my $sth = $dbh->prepare(
2131 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2133 warn "$left / $operator / $right\n" if $DEBUG;
2135 # split each word, query the DB and build the biblionumbers result
2136 #sanitizing leftpart
2137 $left =~ s/^\s+|\s+$//;
2138 foreach ( split / /, $right ) {
2140 $_ =~ s/^\s+|\s+$//;
2142 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2143 $sth->execute( $server, $left, $_ )
2144 or warn "execute failed: $!";
2145 while ( my ( $line, $value ) = $sth->fetchrow ) {
2147 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2148 # otherwise, fill the result
2149 $biblionumbers .= $line
2150 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2151 warn "result : $value "
2152 . ( $right =~ /\d/ ) . "=="
2153 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2156 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2158 warn "NZAND" if $DEBUG;
2159 $results = NZoperatorAND($biblionumbers,$results);
2161 $results = $biblionumbers;
2166 #do a complete search (all indexes), if index='kw' do complete search too.
2167 my $sth = $dbh->prepare(
2168 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2171 # split each word, query the DB and build the biblionumbers result
2172 foreach ( split / /, $string ) {
2173 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2174 warn "search on all indexes on $_" if $DEBUG;
2177 $sth->execute( $server, $_ );
2178 while ( my $line = $sth->fetchrow ) {
2179 $biblionumbers .= $line;
2182 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2184 $results = NZoperatorAND($biblionumbers,$results);
2187 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2188 $results = $biblionumbers;
2192 warn "return : $results for LEAF : $string" if $DEBUG;
2195 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2199 my ($rightresult, $leftresult)=@_;
2201 my @leftresult = split /;/, $leftresult;
2202 warn " @leftresult / $rightresult \n" if $DEBUG;
2204 # my @rightresult = split /;/,$leftresult;
2207 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2208 # the result is stored twice, to have the same weight for AND than OR.
2209 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2210 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2211 foreach (@leftresult) {
2214 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2215 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2216 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2218 "$value-$countvalue;$value-$countvalue;";
2221 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2222 return $finalresult;
2226 my ($rightresult, $leftresult)=@_;
2227 return $rightresult.$leftresult;
2231 my ($leftresult, $rightresult)=@_;
2233 my @leftresult = split /;/, $leftresult;
2235 # my @rightresult = split /;/,$leftresult;
2237 foreach (@leftresult) {
2239 $value=$1 if $value=~m/(.*)-\d+$/;
2240 unless ($rightresult =~ "$value-") {
2241 $finalresult .= "$_;";
2244 return $finalresult;
2249 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2256 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2257 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2259 # order title asc by default
2260 # $ordering = '1=36 <i' unless $ordering;
2261 $results_per_page = 20 unless $results_per_page;
2262 $offset = 0 unless $offset;
2263 my $dbh = C4::Context->dbh;
2266 # order by POPULARITY
2268 if ( $ordering =~ /popularity/ ) {
2272 # popularity is not in MARC record, it's builded from a specific query
2274 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2275 foreach ( split /;/, $biblionumbers ) {
2276 my ( $biblionumber, $title ) = split /,/, $_;
2277 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2278 $sth->execute($biblionumber);
2279 my $popularity = $sth->fetchrow || 0;
2281 # hint : the key is popularity.title because we can have
2282 # many results with the same popularity. In this case, sub-ordering is done by title
2283 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2284 # (un-frequent, I agree, but we won't forget anything that way ;-)
2285 $popularity{ sprintf( "%10d", $popularity ) . $title
2286 . $biblionumber } = $biblionumber;
2289 # sort the hash and return the same structure as GetRecords (Zebra querying)
2292 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2293 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2294 $result_hash->{'RECORDS'}[ $numbers++ ] =
2295 $result{ $popularity{$key} }->as_usmarc();
2298 else { # sort popularity ASC
2299 foreach my $key ( sort ( keys %popularity ) ) {
2300 $result_hash->{'RECORDS'}[ $numbers++ ] =
2301 $result{ $popularity{$key} }->as_usmarc();
2304 my $finalresult = ();
2305 $result_hash->{'hits'} = $numbers;
2306 $finalresult->{'biblioserver'} = $result_hash;
2307 return $finalresult;
2313 elsif ( $ordering =~ /author/ ) {
2315 foreach ( split /;/, $biblionumbers ) {
2316 my ( $biblionumber, $title ) = split /,/, $_;
2317 my $record = GetMarcBiblio($biblionumber);
2319 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2320 $author = $record->subfield( '200', 'f' );
2321 $author = $record->subfield( '700', 'a' ) unless $author;
2324 $author = $record->subfield( '100', 'a' );
2327 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2328 # and we don't want to get only 1 result for each of them !!!
2329 $result{ $author . $biblionumber } = $record;
2332 # sort the hash and return the same structure as GetRecords (Zebra querying)
2335 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2336 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2337 $result_hash->{'RECORDS'}[ $numbers++ ] =
2338 $result{$key}->as_usmarc();
2341 else { # sort by author ASC
2342 foreach my $key ( sort ( keys %result ) ) {
2343 $result_hash->{'RECORDS'}[ $numbers++ ] =
2344 $result{$key}->as_usmarc();
2347 my $finalresult = ();
2348 $result_hash->{'hits'} = $numbers;
2349 $finalresult->{'biblioserver'} = $result_hash;
2350 return $finalresult;
2353 # ORDER BY callnumber
2356 elsif ( $ordering =~ /callnumber/ ) {
2358 foreach ( split /;/, $biblionumbers ) {
2359 my ( $biblionumber, $title ) = split /,/, $_;
2360 my $record = GetMarcBiblio($biblionumber);
2362 my $frameworkcode = GetFrameworkCode($biblionumber);
2363 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2364 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2365 unless $callnumber_tag;
2366 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2367 $callnumber = $record->subfield( '200', 'f' );
2369 $callnumber = $record->subfield( '100', 'a' );
2372 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2373 # and we don't want to get only 1 result for each of them !!!
2374 $result{ $callnumber . $biblionumber } = $record;
2377 # sort the hash and return the same structure as GetRecords (Zebra querying)
2380 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2381 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2382 $result_hash->{'RECORDS'}[ $numbers++ ] =
2383 $result{$key}->as_usmarc();
2386 else { # sort by title ASC
2387 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2388 $result_hash->{'RECORDS'}[ $numbers++ ] =
2389 $result{$key}->as_usmarc();
2392 my $finalresult = ();
2393 $result_hash->{'hits'} = $numbers;
2394 $finalresult->{'biblioserver'} = $result_hash;
2395 return $finalresult;
2397 elsif ( $ordering =~ /pubdate/ ) { #pub year
2399 foreach ( split /;/, $biblionumbers ) {
2400 my ( $biblionumber, $title ) = split /,/, $_;
2401 my $record = GetMarcBiblio($biblionumber);
2402 my ( $publicationyear_tag, $publicationyear_subfield ) =
2403 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2404 my $publicationyear =
2405 $record->subfield( $publicationyear_tag,
2406 $publicationyear_subfield );
2408 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2409 # and we don't want to get only 1 result for each of them !!!
2410 $result{ $publicationyear . $biblionumber } = $record;
2413 # sort the hash and return the same structure as GetRecords (Zebra querying)
2416 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2417 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2418 $result_hash->{'RECORDS'}[ $numbers++ ] =
2419 $result{$key}->as_usmarc();
2422 else { # sort by pub year ASC
2423 foreach my $key ( sort ( keys %result ) ) {
2424 $result_hash->{'RECORDS'}[ $numbers++ ] =
2425 $result{$key}->as_usmarc();
2428 my $finalresult = ();
2429 $result_hash->{'hits'} = $numbers;
2430 $finalresult->{'biblioserver'} = $result_hash;
2431 return $finalresult;
2437 elsif ( $ordering =~ /title/ ) {
2439 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2441 foreach ( split /;/, $biblionumbers ) {
2442 my ( $biblionumber, $title ) = split /,/, $_;
2444 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2445 # and we don't want to get only 1 result for each of them !!!
2446 # hint & speed improvement : we can order without reading the record
2447 # so order, and read records only for the requested page !
2448 $result{ $title . $biblionumber } = $biblionumber;
2451 # sort the hash and return the same structure as GetRecords (Zebra querying)
2454 if ( $ordering eq 'title_az' ) { # sort by title desc
2455 foreach my $key ( sort ( keys %result ) ) {
2456 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2459 else { # sort by title ASC
2460 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2461 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2465 # limit the $results_per_page to result size if it's more
2466 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2468 # for the requested page, replace biblionumber by the complete record
2469 # speed improvement : avoid reading too much things
2471 my $counter = $offset ;
2472 $counter <= $offset + $results_per_page ;
2476 $result_hash->{'RECORDS'}[$counter] =
2477 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2479 my $finalresult = ();
2480 $result_hash->{'hits'} = $numbers;
2481 $finalresult->{'biblioserver'} = $result_hash;
2482 return $finalresult;
2489 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2492 foreach ( split /;/, $biblionumbers ) {
2493 my ( $biblionumber, $title ) = split /,/, $_;
2494 $title =~ /(.*)-(\d)/;
2499 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2500 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2501 # biblio N has ranking = 6
2502 $count_ranking{$biblionumber} += $ranking;
2505 # build the result by "inverting" the count_ranking hash
2506 # 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
2508 foreach ( keys %count_ranking ) {
2509 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2512 # sort the hash and return the same structure as GetRecords (Zebra querying)
2515 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2516 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2519 # limit the $results_per_page to result size if it's more
2520 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2522 # for the requested page, replace biblionumber by the complete record
2523 # speed improvement : avoid reading too much things
2525 my $counter = $offset ;
2526 $counter <= $offset + $results_per_page ;
2530 $result_hash->{'RECORDS'}[$counter] =
2531 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2532 if $result_hash->{'RECORDS'}[$counter];
2534 my $finalresult = ();
2535 $result_hash->{'hits'} = $numbers;
2536 $finalresult->{'biblioserver'} = $result_hash;
2537 return $finalresult;
2541 =head2 enabled_staff_search_views
2543 %hash = enabled_staff_search_views()
2545 This function returns a hash that contains three flags obtained from the system
2546 preferences, used to determine whether a particular staff search results view
2551 =item C<Output arg:>
2553 * $hash{can_view_MARC} is true only if the MARC view is enabled
2554 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2555 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2557 =item C<usage in the script:>
2561 $template->param ( C4::Search::enabled_staff_search_views );
2565 sub enabled_staff_search_views
2568 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2569 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2570 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2574 sub AddSearchHistory{
2575 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2576 my $dbh = C4::Context->dbh;
2578 # Add the request the user just made
2579 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2580 my $sth = $dbh->prepare($sql);
2581 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2582 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2585 sub GetSearchHistory{
2586 my ($borrowernumber,$session)=@_;
2587 my $dbh = C4::Context->dbh;
2589 # Add the request the user just made
2590 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2591 my $sth = $dbh->prepare($query);
2592 $sth->execute($borrowernumber, $session);
2593 return $sth->fetchall_hashref({});
2596 =head2 z3950_search_args
2598 $arrayref = z3950_search_args($matchpoints)
2600 This function returns an array reference that contains the search parameters to be
2601 passed to the Z39.50 search script (z3950_search.pl). The array elements
2602 are hash refs whose keys are name, value and encvalue, and whose values are the
2603 name of a search parameter, the value of that search parameter and the URL encoded
2604 value of that parameter.
2606 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2608 The search parameter values are obtained from the bibliographic record whose
2609 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2611 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2612 a general purpose search argument. In this case, the returned array contains only
2613 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2615 If a search parameter value is undefined or empty, it is not included in the returned
2618 The returned array reference may be passed directly to the template parameters.
2622 =item C<Output arg:>
2624 * $array containing hash refs as described above
2626 =item C<usage in the script:>
2630 $data = Biblio::GetBiblioData($bibno);
2631 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2635 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2639 sub z3950_search_args {
2641 my $isbn = Business::ISBN->new($bibrec);
2643 if (defined $isbn && $isbn->is_valid)
2645 $bibrec = { isbn => $bibrec } if !ref $bibrec;
2648 $bibrec = { title => $bibrec } if !ref $bibrec;
2651 for my $field (qw/ lccn isbn issn title author dewey subject /)
2653 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2654 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2659 =head2 GetDistinctValues($field);
2661 C<$field> is a reference to the fields array
2665 sub GetDistinctValues {
2666 my ($fieldname,$string)=@_;
2667 # returns a reference to a hash of references to branches...
2668 if ($fieldname=~/\./){
2669 my ($table,$column)=split /\./, $fieldname;
2670 my $dbh = C4::Context->dbh;
2671 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2672 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 ");
2674 my $elements=$sth->fetchall_arrayref({});
2679 my @servers=qw<biblioserver authorityserver>;
2680 my (@zconns,@results);
2681 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2682 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2685 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2688 # The big moment: asynchronously retrieve results from all servers
2690 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2691 my $ev = $zconns[ $i - 1 ]->last_event();
2692 if ( $ev == ZOOM::Event::ZEND ) {
2693 next unless $results[ $i - 1 ];
2694 my $size = $results[ $i - 1 ]->size();
2696 for (my $j=0;$j<$size;$j++){
2698 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2699 push @elements, \%hashscan;
2709 END { } # module clean-up code here (global destructor)
2716 Koha Development Team <http://koha-community.org/>