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
37 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
39 # set the version for version checking
42 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
47 C4::Search - Functions for searching the Koha catalog.
51 See opac/opac-search.pl or catalogue/search.pl for example of usage
55 This module provides searching functions for Koha's bibliographic databases
73 #FIXME: i had to add BiblioAddAuthorities here because in Biblios.pm it caused circular dependencies (C4::Search uses C4::Biblio, and BiblioAddAuthorities uses SimpleSearch from C4::Search)
75 # make all your functions, whether exported or not;
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $dbh, $record, '' );
93 my ( $biblionumber, $title );
95 # search duplicate on ISBN, easy and fast..
97 if ( $result->{isbn} ) {
98 $result->{isbn} =~ s/\(.*$//;
99 $result->{isbn} =~ s/\s+$//;
100 $query = "isbn=$result->{isbn}";
103 $result->{title} =~ s /\\//g;
104 $result->{title} =~ s /\"//g;
105 $result->{title} =~ s /\(//g;
106 $result->{title} =~ s /\)//g;
108 # FIXME: instead of removing operators, could just do
109 # quotes around the value
110 $result->{title} =~ s/(and|or|not)//g;
111 $query = "ti,ext=$result->{title}";
112 $query .= " and itemtype=$result->{itemtype}"
113 if ( $result->{itemtype} );
114 if ( $result->{author} ) {
115 $result->{author} =~ s /\\//g;
116 $result->{author} =~ s /\"//g;
117 $result->{author} =~ s /\(//g;
118 $result->{author} =~ s /\)//g;
120 # remove valid operators
121 $result->{author} =~ s/(and|or|not)//g;
122 $query .= " and au,ext=$result->{author}";
126 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
128 if (!defined $error) {
129 foreach my $possible_duplicate_record (@{$searchresults}) {
131 MARC::Record->new_from_usmarc($possible_duplicate_record);
132 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
134 # FIXME :: why 2 $biblionumber ?
136 push @results, $result->{'biblionumber'};
137 push @results, $result->{'title'};
146 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
148 This function provides a simple search API on the bibliographic catalog
154 * $query can be a simple keyword or a complete CCL query
155 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
156 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
157 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
162 Returns an array consisting of three elements
163 * $error is undefined unless an error is detected
164 * $results is a reference to an array of records.
165 * $total_hits is the number of hits that would have been returned with no limit
167 If an error is returned the two other return elements are undefined. If error itself is undefined
168 the other two elements are always defined
170 =item C<usage in the script:>
174 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
176 if (defined $error) {
177 $template->param(query_error => $error);
178 warn "error: ".$error;
179 output_html_with_http_headers $input, $cookie, $template->output;
183 my $hits = @{$marcresults};
186 for my $r ( @{$marcresults} ) {
187 my $marcrecord = MARC::File::USMARC::decode($r);
188 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
190 #build the iarray of hashs for the template.
192 title => $biblio->{'title'},
193 subtitle => $biblio->{'subtitle'},
194 biblionumber => $biblio->{'biblionumber'},
195 author => $biblio->{'author'},
196 publishercode => $biblio->{'publishercode'},
197 publicationyear => $biblio->{'publicationyear'},
202 $template->param(result=>\@results);
207 my ( $query, $offset, $max_results, $servers ) = @_;
209 if ( C4::Context->preference('NoZebra') ) {
210 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
213 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
214 return ( undef, $search_result, scalar($result->{hits}) );
217 return ( 'No query entered', undef, undef ) unless $query;
218 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
219 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
226 # Initialize & Search Zebra
227 for ( my $i = 0 ; $i < @servers ; $i++ ) {
229 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
230 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
231 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
235 $zconns[$i]->errmsg() . " ("
236 . $zconns[$i]->errcode() . ") "
237 . $zconns[$i]->addinfo() . " "
238 . $zconns[$i]->diagset();
240 return ( $error, undef, undef ) if $zconns[$i]->errcode();
244 # caught a ZOOM::Exception
248 . $@->addinfo() . " "
251 return ( $error, undef, undef );
254 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
255 my $event = $zconns[ $i - 1 ]->last_event();
256 if ( $event == ZOOM::Event::ZEND ) {
258 my $first_record = defined( $offset ) ? $offset+1 : 1;
259 my $hits = $tmpresults[ $i - 1 ]->size();
260 $total_hits += $hits;
261 my $last_record = $hits;
262 if ( defined $max_results && $offset + $max_results < $hits ) {
263 $last_record = $offset + $max_results;
266 for my $j ( $first_record..$last_record ) {
267 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
268 push @{$results}, $record;
273 foreach my $result (@tmpresults) {
276 foreach my $zoom_query (@zoom_queries) {
277 $zoom_query->destroy();
280 return ( undef, $results, $total_hits );
286 ( undef, $results_hashref, \@facets_loop ) = getRecords (
288 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
289 $results_per_page, $offset, $expanded_facet, $branches,
293 The all singing, all dancing, multi-server, asynchronous, scanning,
294 searching, record nabbing, facet-building
296 See verbse embedded documentation.
302 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
303 $results_per_page, $offset, $expanded_facet, $branches,
307 my @servers = @$servers_ref;
308 my @sort_by = @$sort_by_ref;
310 # Initialize variables for the ZOOM connection and results object
314 my $results_hashref = ();
316 # Initialize variables for the faceted results objects
317 my $facets_counter = ();
318 my $facets_info = ();
319 my $facets = getFacets();
320 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
322 my @facets_loop; # stores the ref to array of hashes for template facets loop
324 ### LOOP THROUGH THE SERVERS
325 for ( my $i = 0 ; $i < @servers ; $i++ ) {
326 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
328 # perform the search, create the results objects
329 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
330 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
332 #$query_to_use = $simple_query if $scan;
333 warn $simple_query if ( $scan and $DEBUG );
335 # Check if we've got a query_type defined, if so, use it
338 if ($query_type =~ /^ccl/) {
339 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
340 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
341 } elsif ($query_type =~ /^cql/) {
342 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
343 } elsif ($query_type =~ /^pqf/) {
344 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
346 warn "Unknown query_type '$query_type'. Results undetermined.";
349 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
351 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
355 warn "WARNING: query problem with $query_to_use " . $@;
358 # Concatenate the sort_by limits and pass them to the results object
359 # Note: sort will override rank
361 foreach my $sort (@sort_by) {
362 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
363 $sort_by .= "1=1003 <i ";
365 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
366 $sort_by .= "1=1003 >i ";
368 elsif ( $sort eq "popularity_asc" ) {
369 $sort_by .= "1=9003 <i ";
371 elsif ( $sort eq "popularity_dsc" ) {
372 $sort_by .= "1=9003 >i ";
374 elsif ( $sort eq "call_number_asc" ) {
375 $sort_by .= "1=8007 <i ";
377 elsif ( $sort eq "call_number_dsc" ) {
378 $sort_by .= "1=8007 >i ";
380 elsif ( $sort eq "pubdate_asc" ) {
381 $sort_by .= "1=31 <i ";
383 elsif ( $sort eq "pubdate_dsc" ) {
384 $sort_by .= "1=31 >i ";
386 elsif ( $sort eq "acqdate_asc" ) {
387 $sort_by .= "1=32 <i ";
389 elsif ( $sort eq "acqdate_dsc" ) {
390 $sort_by .= "1=32 >i ";
392 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
393 $sort_by .= "1=4 <i ";
395 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
396 $sort_by .= "1=4 >i ";
399 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
402 if ($sort_by && !$scan) {
403 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
404 warn "WARNING sort $sort_by failed";
407 } # finished looping through servers
409 # The big moment: asynchronously retrieve results from all servers
410 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
411 my $ev = $zconns[ $i - 1 ]->last_event();
412 if ( $ev == ZOOM::Event::ZEND ) {
413 next unless $results[ $i - 1 ];
414 my $size = $results[ $i - 1 ]->size();
418 # loop through the results
419 $results_hash->{'hits'} = $size;
421 if ( $offset + $results_per_page <= $size ) {
422 $times = $offset + $results_per_page;
427 for ( my $j = $offset ; $j < $times ; $j++ ) {
431 ## Check if it's an index scan
433 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
435 # here we create a minimal MARC record and hand it off to the
436 # template just like a normal result ... perhaps not ideal, but
438 my $tmprecord = MARC::Record->new();
439 $tmprecord->encoding('UTF-8');
443 # the minimal record in author/title (depending on MARC flavour)
444 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
445 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
446 $tmprecord->append_fields($tmptitle);
448 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
449 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
450 $tmprecord->append_fields($tmptitle);
451 $tmprecord->append_fields($tmpauthor);
453 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
458 $record = $results[ $i - 1 ]->record($j)->raw();
460 # warn "RECORD $j:".$record;
461 $results_hash->{'RECORDS'}[$j] = $record;
465 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
467 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
468 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
470 my $jmax = $size>$facets_maxrecs? $facets_maxrecs: $size;
472 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
473 ($facets->[$k]) or next;
474 my @fcodes = @{$facets->[$k]->{'tags'}};
475 my $sfcode = $facets->[$k]->{'subfield'};
477 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
478 my $render_record = $results[ $i - 1 ]->record($j)->render();
481 foreach my $fcode (@fcodes) {
484 my $field_pattern = '\n'.$fcode.' ([^\n]+)';
485 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
487 foreach my $field_token (@field_tokens) {
488 my $subfield_pattern = '\$'.$sfcode.' ([^\$]+)';
489 my @subfield_values = ( $field_token =~ /$subfield_pattern/g );
491 foreach my $subfield_value (@subfield_values) {
493 my $data = $subfield_value;
494 $data =~ s/^\s+//; # trim left
495 $data =~ s/\s+$//; # trim right
497 unless ( $data ~~ @used_datas ) {
498 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
499 push @used_datas, $data;
506 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} = $facets->[$k]->{'label_value'};
507 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} = $facets->[$k]->{'expanded'};
513 # warn "connection ", $i-1, ": $size hits";
514 # warn $results[$i-1]->record(0)->render() if $size > 0;
517 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
519 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
520 keys %$facets_counter )
523 my $number_of_facets;
524 my @this_facets_array;
527 $facets_counter->{$link_value}->{$b}
528 <=> $facets_counter->{$link_value}->{$a}
529 } keys %{ $facets_counter->{$link_value} }
533 if ( ( $number_of_facets < 6 )
534 || ( $expanded_facet eq $link_value )
535 || ( $facets_info->{$link_value}->{'expanded'} ) )
538 # Sanitize the link value ), ( will cause errors with CCL,
539 my $facet_link_value = $one_facet;
540 $facet_link_value =~ s/(\(|\))/ /g;
542 # fix the length that will display in the label,
543 my $facet_label_value = $one_facet;
544 my $facet_max_length =
545 C4::Context->preference('FacetLabelTruncationLength') || 20;
547 substr( $one_facet, 0, $facet_max_length ) . "..."
548 if length($facet_label_value) > $facet_max_length;
550 # if it's a branch, label by the name, not the code,
551 if ( $link_value =~ /branch/ ) {
552 if (defined $branches
553 && ref($branches) eq "HASH"
554 && defined $branches->{$one_facet}
555 && ref ($branches->{$one_facet}) eq "HASH")
558 $branches->{$one_facet}->{'branchname'};
561 $facet_label_value = "*";
565 # but we're down with the whole label being in the link's title.
566 push @this_facets_array, {
567 facet_count => $facets_counter->{$link_value}->{$one_facet},
568 facet_label_value => $facet_label_value,
569 facet_title_value => $one_facet,
570 facet_link_value => $facet_link_value,
571 type_link_value => $link_value,
576 # handle expanded option
577 unless ( $facets_info->{$link_value}->{'expanded'} ) {
579 if ( ( $number_of_facets > 6 )
580 && ( $expanded_facet ne $link_value ) );
583 type_link_value => $link_value,
584 type_id => $link_value . "_id",
585 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
586 facets => \@this_facets_array,
587 expandable => $expandable,
588 expand => $link_value,
589 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
594 return ( undef, $results_hashref, \@facets_loop );
599 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
600 $results_per_page, $offset, $expanded_facet, $branches,
604 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
606 $paz->search($simple_query);
607 sleep 1; # FIXME: WHY?
610 my $results_hashref = {};
611 my $stats = XMLin($paz->stat);
612 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
614 # for a grouped search result, the number of hits
615 # is the number of groups returned; 'bib_hits' will have
616 # the total number of bibs.
617 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
618 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
620 HIT: foreach my $hit (@{ $results->{'hit'} }) {
621 my $recid = $hit->{recid}->[0];
623 my $work_title = $hit->{'md-work-title'}->[0];
625 if (exists $hit->{'md-work-author'}) {
626 $work_author = $hit->{'md-work-author'}->[0];
628 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
630 my $result_group = {};
631 $result_group->{'group_label'} = $group_label;
632 $result_group->{'group_merge_key'} = $recid;
635 if (exists $hit->{count}) {
636 $count = $hit->{count}->[0];
638 $result_group->{'group_count'} = $count;
640 for (my $i = 0; $i < $count; $i++) {
641 # FIXME -- may need to worry about diacritics here
642 my $rec = $paz->record($recid, $i);
643 push @{ $result_group->{'RECORDS'} }, $rec;
646 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
649 # pass through facets
650 my $termlist_xml = $paz->termlist('author,subject');
651 my $terms = XMLin($termlist_xml, forcearray => 1);
652 my @facets_loop = ();
653 #die Dumper($results);
654 # foreach my $list (sort keys %{ $terms->{'list'} }) {
656 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
658 # facet_label_value => $facet->{'name'}->[0],
661 # push @facets_loop, ( {
662 # type_label => $list,
663 # facets => \@facets,
667 return ( undef, $results_hashref, \@facets_loop );
671 sub _remove_stopwords {
672 my ( $operand, $index ) = @_;
673 my @stopwords_removed;
675 # phrase and exact-qualified indexes shouldn't have stopwords removed
676 if ( $index !~ m/phr|ext/ ) {
678 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
679 # we use IsAlpha unicode definition, to deal correctly with diacritics.
680 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
681 # is a stopword, we'd get "çon" and wouldn't find anything...
683 foreach ( keys %{ C4::Context->stopwords } ) {
684 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
685 if ( my ($matched) = ($operand =~
686 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
688 $operand =~ s/\Q$matched\E/ /gi;
689 push @stopwords_removed, $_;
693 return ( $operand, \@stopwords_removed );
697 sub _detect_truncation {
698 my ( $operand, $index ) = @_;
699 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
702 my @wordlist = split( /\s/, $operand );
703 foreach my $word (@wordlist) {
704 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
705 push @rightlefttruncated, $word;
707 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
708 push @lefttruncated, $word;
710 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
711 push @righttruncated, $word;
713 elsif ( index( $word, "*" ) < 0 ) {
714 push @nontruncated, $word;
717 push @regexpr, $word;
721 \@nontruncated, \@righttruncated, \@lefttruncated,
722 \@rightlefttruncated, \@regexpr
727 sub _build_stemmed_operand {
728 my ($operand,$lang) = @_;
729 require Lingua::Stem::Snowball ;
732 # If operand contains a digit, it is almost certainly an identifier, and should
733 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
734 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
735 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
736 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
737 return $operand if $operand =~ /\d/;
739 # FIXME: the locale should be set based on the user's language and/or search choice
741 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
742 encoding => "UTF-8" );
744 my @words = split( / /, $operand );
745 my @stems = $stemmer->stem(\@words);
746 for my $stem (@stems) {
747 $stemmed_operand .= "$stem";
748 $stemmed_operand .= "?"
749 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
750 $stemmed_operand .= " ";
752 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
753 return $stemmed_operand;
757 sub _build_weighted_query {
759 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
760 # pretty well but could work much better if we had a smarter query parser
761 my ( $operand, $stemmed_operand, $index ) = @_;
762 my $stemming = C4::Context->preference("QueryStemming") || 0;
763 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
764 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
766 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
768 # Keyword, or, no index specified
769 if ( ( $index eq 'kw' ) || ( !$index ) ) {
771 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
772 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
773 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
774 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
775 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
776 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
777 if $fuzzy_enabled; # add fuzzy, word list
778 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
779 if ( $stemming and $stemmed_operand )
780 ; # add stemming, right truncation
781 $weighted_query .= " or wrdl,r9=\"$operand\"";
783 # embedded sorting: 0 a-z; 1 z-a
784 # $weighted_query .= ") or (sort1,aut=1";
787 # Barcode searches should skip this process
788 elsif ( $index eq 'bc' ) {
789 $weighted_query .= "bc=\"$operand\"";
792 # Authority-number searches should skip this process
793 elsif ( $index eq 'an' ) {
794 $weighted_query .= "an=\"$operand\"";
797 # If the index already has more than one qualifier, wrap the operand
798 # in quotes and pass it back (assumption is that the user knows what they
799 # are doing and won't appreciate us mucking up their query
800 elsif ( $index =~ ',' ) {
801 $weighted_query .= " $index=\"$operand\"";
804 #TODO: build better cases based on specific search indexes
806 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
807 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
808 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
810 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
813 $weighted_query .= "))"; # close rank specification
814 return $weighted_query;
819 Return an array with available indexes.
841 'Author-personal-bibliography',
851 'Chronological-subdivision',
861 'Conference-name-heading',
862 'Conference-name-see',
863 'Conference-name-seealso',
868 'Corporate-name-heading',
869 'Corporate-name-see',
870 'Corporate-name-seealso',
872 'date-entered-on-file',
873 'Date-of-acquisition',
874 'Date-of-publication',
875 'Dewey-classification',
882 'Geographic-subdivision',
885 'Heading-use-main-or-added-entry',
886 'Heading-use-series-added-entry ',
887 'Heading-use-subject-added-entry',
905 'Local-classification',
908 'Match-heading-see-from',
916 'Name-geographic-heading',
917 'Name-geographic-see',
918 'Name-geographic-seealso',
926 'Personal-name-heading',
928 'Personal-name-seealso',
935 'Record-control-number',
946 'Subject-heading-thesaurus',
947 'Subject-name-personal',
948 'Subject-subdivision',
958 'Term-genre-form-heading',
959 'Term-genre-form-see',
960 'Term-genre-form-seealso',
967 'Title-uniform-heading',
969 'Title-uniform-seealso',
979 'classification-source',
981 'coded-location-qualifier',
992 'Local-classification',
995 'materials-specified',
1004 'replacementpricedate',
1022 $simple_query, $query_cgi,
1023 $query_desc, $limit,
1024 $limit_cgi, $limit_desc,
1025 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1027 Build queries and limits in CCL, CGI, Human,
1028 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1030 See verbose embedded documentation.
1036 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1038 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1041 my @operators = $operators ? @$operators : ();
1042 my @indexes = $indexes ? @$indexes : ();
1043 my @operands = $operands ? @$operands : ();
1044 my @limits = $limits ? @$limits : ();
1045 my @sort_by = $sort_by ? @$sort_by : ();
1047 my $stemming = C4::Context->preference("QueryStemming") || 0;
1048 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1049 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1050 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1051 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1053 # no stemming/weight/fuzzy in NoZebra
1054 if ( C4::Context->preference("NoZebra") ) {
1058 $auto_truncation = 0;
1061 my $query = $operands[0];
1062 my $simple_query = $operands[0];
1064 # initialize the variables we're passing back
1073 my $stopwords_removed; # flag to determine if stopwords have been removed
1076 my $cclindexes = getIndexes();
1077 if ( $query !~ /\s*ccl=/ ) {
1078 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1080 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1082 $query = "ccl=$query" if $cclq;
1085 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1087 if ( $query =~ /^ccl=/ ) {
1089 # This is needed otherwise ccl= and &limit won't work together, and
1090 # this happens when selecting a subject on the opac-detail page
1092 $q .= ' and '.join(' and ', @limits);
1094 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1096 if ( $query =~ /^cql=/ ) {
1097 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1099 if ( $query =~ /^pqf=/ ) {
1100 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1103 # pass nested queries directly
1104 # FIXME: need better handling of some of these variables in this case
1105 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1106 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1107 # if ( $query =~ /(\(|\))/ ) {
1109 # undef, $query, $simple_query, $query_cgi,
1110 # $query, $limit, $limit_cgi, $limit_desc,
1111 # $stopwords_removed, 'ccl'
1115 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1116 # query operands and indexes and add stemming, truncation, field weighting, etc.
1117 # Once we do so, we'll end up with a value in $query, just like if we had an
1118 # incoming $query from the user
1121 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1122 my $previous_operand
1123 ; # a flag used to keep track if there was a previous query
1124 # if there was, we can apply the current operator
1126 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1128 # COMBINE OPERANDS, INDEXES AND OPERATORS
1129 if ( $operands[$i] ) {
1130 $operands[$i]=~s/^\s+//;
1132 # A flag to determine whether or not to add the index to the query
1135 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1136 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1139 $remove_stopwords = 0;
1141 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1143 my $operand = $operands[$i];
1144 my $index = $indexes[$i];
1146 # Add index-specific attributes
1147 # Date of Publication
1148 if ( $index eq 'yr' ) {
1149 $index .= ",st-numeric";
1151 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1154 # Date of Acquisition
1155 elsif ( $index eq 'acqdate' ) {
1156 $index .= ",st-date-normalized";
1158 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1160 # ISBN,ISSN,Standard Number, don't need special treatment
1161 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1163 $stemming, $auto_truncation,
1164 $weight_fields, $fuzzy_enabled,
1166 ) = ( 0, 0, 0, 0, 0 );
1174 # Set default structure attribute (word list)
1175 my $struct_attr = q{};
1176 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1177 $struct_attr = ",wrdl";
1180 # Some helpful index variants
1181 my $index_plus = $index . $struct_attr . ':';
1182 my $index_plus_comma = $index . $struct_attr . ',';
1185 if ($remove_stopwords) {
1186 ( $operand, $stopwords_removed ) =
1187 _remove_stopwords( $operand, $index );
1188 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1189 warn "REMOVED STOPWORDS: @$stopwords_removed"
1190 if ( $stopwords_removed && $DEBUG );
1193 if ($auto_truncation){
1194 unless ( $index =~ /(st-|phr|ext)/ ) {
1195 #FIXME only valid with LTR scripts
1196 $operand=join(" ",map{
1197 (index($_,"*")>0?"$_":"$_*")
1198 }split (/\s+/,$operand));
1199 warn $operand if $DEBUG;
1204 my $truncated_operand;
1205 my( $nontruncated, $righttruncated, $lefttruncated,
1206 $rightlefttruncated, $regexpr
1207 ) = _detect_truncation( $operand, $index );
1209 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1214 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1215 scalar(@$rightlefttruncated) > 0 )
1218 # Don't field weight or add the index to the query, we do it here
1220 undef $weight_fields;
1221 my $previous_truncation_operand;
1222 if (scalar @$nontruncated) {
1223 $truncated_operand .= "$index_plus @$nontruncated ";
1224 $previous_truncation_operand = 1;
1226 if (scalar @$righttruncated) {
1227 $truncated_operand .= "and " if $previous_truncation_operand;
1228 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1229 $previous_truncation_operand = 1;
1231 if (scalar @$lefttruncated) {
1232 $truncated_operand .= "and " if $previous_truncation_operand;
1233 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1234 $previous_truncation_operand = 1;
1236 if (scalar @$rightlefttruncated) {
1237 $truncated_operand .= "and " if $previous_truncation_operand;
1238 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1239 $previous_truncation_operand = 1;
1242 $operand = $truncated_operand if $truncated_operand;
1243 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1246 my $stemmed_operand;
1247 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1250 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1252 # Handle Field Weighting
1253 my $weighted_operand;
1254 if ($weight_fields) {
1255 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1256 $operand = $weighted_operand;
1260 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1262 # If there's a previous operand, we need to add an operator
1263 if ($previous_operand) {
1265 # User-specified operator
1266 if ( $operators[ $i - 1 ] ) {
1267 $query .= " $operators[$i-1] ";
1268 $query .= " $index_plus " unless $indexes_set;
1269 $query .= " $operand";
1270 $query_cgi .= "&op=$operators[$i-1]";
1271 $query_cgi .= "&idx=$index" if $index;
1272 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1274 " $operators[$i-1] $index_plus $operands[$i]";
1277 # Default operator is and
1280 $query .= "$index_plus " unless $indexes_set;
1281 $query .= "$operand";
1282 $query_cgi .= "&op=and&idx=$index" if $index;
1283 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1284 $query_desc .= " and $index_plus $operands[$i]";
1288 # There isn't a pervious operand, don't need an operator
1291 # Field-weighted queries already have indexes set
1292 $query .= " $index_plus " unless $indexes_set;
1294 $query_desc .= " $index_plus $operands[$i]";
1295 $query_cgi .= "&idx=$index" if $index;
1296 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1297 $previous_operand = 1;
1302 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1305 my $group_OR_limits;
1306 my $availability_limit;
1307 foreach my $this_limit (@limits) {
1308 if ( $this_limit =~ /available/ ) {
1310 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1312 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1313 $availability_limit .=
1314 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1315 $limit_cgi .= "&limit=available";
1319 # group_OR_limits, prefixed by mc-
1320 # OR every member of the group
1321 elsif ( $this_limit =~ /mc/ ) {
1323 if ( $this_limit =~ /mc-ccode:/ ) {
1324 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1325 $this_limit =~ tr/"//d;
1326 my ($k,$v) = split(/:/, $this_limit,2);
1327 $this_limit = $k.":\"".$v."\"";
1330 $group_OR_limits .= " or " if $group_OR_limits;
1331 $limit_desc .= " or " if $group_OR_limits;
1332 $group_OR_limits .= "$this_limit";
1333 $limit_cgi .= "&limit=$this_limit";
1334 $limit_desc .= " $this_limit";
1337 # Regular old limits
1339 $limit .= " and " if $limit || $query;
1340 $limit .= "$this_limit";
1341 $limit_cgi .= "&limit=$this_limit";
1342 if ($this_limit =~ /^branch:(.+)/) {
1343 my $branchcode = $1;
1344 my $branchname = GetBranchName($branchcode);
1345 if (defined $branchname) {
1346 $limit_desc .= " branch:$branchname";
1348 $limit_desc .= " $this_limit";
1351 $limit_desc .= " $this_limit";
1355 if ($group_OR_limits) {
1356 $limit .= " and " if ( $query || $limit );
1357 $limit .= "($group_OR_limits)";
1359 if ($availability_limit) {
1360 $limit .= " and " if ( $query || $limit );
1361 $limit .= "($availability_limit)";
1364 # Normalize the query and limit strings
1365 # This is flawed , means we can't search anything with : in it
1366 # if user wants to do ccl or cql, start the query with that
1367 # $query =~ s/:/=/g;
1368 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1369 $query =~ s/(?<=(wrdl)):/=/g;
1370 $query =~ s/(?<=(trn|phr)):/=/g;
1372 for ( $query, $query_desc, $limit, $limit_desc ) {
1373 s/ +/ /g; # remove extra spaces
1374 s/^ //g; # remove any beginning spaces
1375 s/ $//g; # remove any ending spaces
1376 s/==/=/g; # remove double == from query
1378 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1380 for ($query_cgi,$simple_query) {
1383 # append the limit to the query
1384 $query .= " " . $limit;
1388 warn "QUERY:" . $query;
1389 warn "QUERY CGI:" . $query_cgi;
1390 warn "QUERY DESC:" . $query_desc;
1391 warn "LIMIT:" . $limit;
1392 warn "LIMIT CGI:" . $limit_cgi;
1393 warn "LIMIT DESC:" . $limit_desc;
1394 warn "---------\nLeave buildQuery\n---------";
1397 undef, $query, $simple_query, $query_cgi,
1398 $query_desc, $limit, $limit_cgi, $limit_desc,
1399 $stopwords_removed, $query_type
1403 =head2 searchResults
1405 my @search_results = searchResults($search_context, $searchdesc, $hits,
1406 $results_per_page, $offset, $scan,
1407 @marcresults, $hidelostitems);
1409 Format results in a form suitable for passing to the template
1413 # IMO this subroutine is pretty messy still -- it's responsible for
1414 # building the HTML output for the template
1416 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1417 my $dbh = C4::Context->dbh;
1420 $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1422 #Build branchnames hash
1424 #get branch information.....
1426 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1428 while ( my $bdata = $bsth->fetchrow_hashref ) {
1429 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1431 # FIXME - We build an authorised values hash here, using the default framework
1432 # though it is possible to have different authvals for different fws.
1434 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1436 # get notforloan authorised value list (see $shelflocations FIXME)
1437 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1439 #Build itemtype hash
1440 #find itemtype & itemtype image
1444 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1447 while ( my $bdata = $bsth->fetchrow_hashref ) {
1448 foreach (qw(description imageurl summary notforloan)) {
1449 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1453 #search item field code
1456 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1459 my ($itemtag) = $sth->fetchrow;
1461 ## find column names of items related to MARC
1462 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1464 my %subfieldstosearch;
1465 while ( ( my $column ) = $sth2->fetchrow ) {
1466 my ( $tagfield, $tagsubfield ) =
1467 &GetMarcFromKohaField( "items." . $column, "" );
1468 $subfieldstosearch{$column} = $tagsubfield;
1471 # handle which records to actually retrieve
1473 if ( $hits && $offset + $results_per_page <= $hits ) {
1474 $times = $offset + $results_per_page;
1477 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1480 my $marcflavour = C4::Context->preference("marcflavour");
1481 # We get the biblionumber position in MARC
1482 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1485 # loop through all of the records we've retrieved
1486 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1487 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1491 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1492 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1493 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1494 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1495 $oldbiblio->{result_number} = $i + 1;
1497 # add imageurl to itemtype if there is one
1498 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1500 $oldbiblio->{'authorised_value_images'} = ($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 ) ) : [];
1501 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1502 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1503 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1504 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1505 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1507 # edition information, if any
1508 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1509 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1510 # Build summary if there is one (the summary is defined in the itemtypes table)
1511 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1512 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1513 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1514 my @fields = $marcrecord->fields();
1517 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1519 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1520 $tag =~ /(.{3})(.)/;
1521 if($marcrecord->field($1)){
1522 my @abc = $marcrecord->field($1)->subfield($2);
1523 $tags->{$tag} = $#abc + 1 ;
1527 # We catch how many times to repeat this line
1529 foreach my $tag (keys(%$tags)){
1530 $max = $tags->{$tag} if($tags->{$tag} > $max);
1533 # we replace, and repeat each line
1534 for (my $i = 0 ; $i < $max ; $i++){
1535 my $newline = $line;
1537 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1538 $tag =~ /(.{3})(.)/;
1540 if($marcrecord->field($1)){
1541 my @repl = $marcrecord->field($1)->subfield($2);
1542 my $subfieldvalue = $repl[$i];
1544 if (! utf8::is_utf8($subfieldvalue)) {
1545 utf8::decode($subfieldvalue);
1548 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1551 $newsummary .= "$newline\n";
1555 $newsummary =~ s/\[(.*?)]//g;
1556 $newsummary =~ s/\n/<br\/>/g;
1557 $oldbiblio->{summary} = $newsummary;
1560 # Pull out the items fields
1561 my @fields = $marcrecord->field($itemtag);
1562 my $marcflavor = C4::Context->preference("marcflavour");
1563 # adding linked items that belong to host records
1564 my $analyticsfield = '773';
1565 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1566 $analyticsfield = '773';
1567 } elsif ($marcflavor eq 'UNIMARC') {
1568 $analyticsfield = '461';
1570 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1571 my $hostbiblionumber = $hostfield->subfield("0");
1572 my $linkeditemnumber = $hostfield->subfield("9");
1573 if(!$hostbiblionumber eq undef){
1574 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1575 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1576 if(!$hostbiblio eq undef){
1577 my @hostitems = $hostbiblio->field($itemfield);
1578 foreach my $hostitem (@hostitems){
1579 if ($hostitem->subfield("9") eq $linkeditemnumber){
1580 my $linkeditem =$hostitem;
1581 # append linked items if they exist
1582 if (!$linkeditem eq undef){
1583 push (@fields, $linkeditem);}
1590 # Setting item statuses for display
1591 my @available_items_loop;
1592 my @onloan_items_loop;
1593 my @other_items_loop;
1595 my $available_items;
1599 my $ordered_count = 0;
1600 my $available_count = 0;
1601 my $onloan_count = 0;
1602 my $longoverdue_count = 0;
1603 my $other_count = 0;
1604 my $wthdrawn_count = 0;
1605 my $itemlost_count = 0;
1606 my $itembinding_count = 0;
1607 my $itemdamaged_count = 0;
1608 my $item_in_transit_count = 0;
1609 my $can_place_holds = 0;
1610 my $item_onhold_count = 0;
1611 my $items_count = scalar(@fields);
1613 ( C4::Context->preference('maxItemsinSearchResults') )
1614 ? C4::Context->preference('maxItemsinSearchResults') - 1
1617 # loop through every item
1618 foreach my $field (@fields) {
1621 # populate the items hash
1622 foreach my $code ( keys %subfieldstosearch ) {
1623 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1627 my @items = ($item);
1628 my (@hiddenitems) = GetHiddenItemnumbers(@items);
1629 $item->{'hideatopac'} = 1 if (@hiddenitems);
1631 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1632 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1634 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1635 if ($item->{$hbranch}) {
1636 $item->{'branchname'} = $branches{$item->{$hbranch}};
1638 elsif ($item->{$otherbranch}) { # Last resort
1639 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1642 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1643 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1644 my $userenv = C4::Context->userenv;
1645 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1647 my $key = $prefix . $item->{onloan} . $item->{barcode};
1648 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1649 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1650 $onloan_items->{$key}->{branchname} = $item->{branchname};
1651 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1652 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1653 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1654 # if something's checked out and lost, mark it as 'long overdue'
1655 if ( $item->{itemlost} ) {
1656 $onloan_items->{$prefix}->{longoverdue}++;
1657 $longoverdue_count++;
1658 } else { # can place holds as long as item isn't lost
1659 $can_place_holds = 1;
1663 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1667 if ( $item->{notforloan} == -1 ) {
1671 # is item in transit?
1672 my $transfertwhen = '';
1673 my ($transfertfrom, $transfertto);
1675 # is item on the reserve shelf?
1676 my $reservestatus = 0;
1679 unless ($item->{wthdrawn}
1680 || $item->{itemlost}
1682 || $item->{notforloan}
1683 || $items_count > 20) {
1685 # A couple heuristics to limit how many times
1686 # we query the database for item transfer information, sacrificing
1687 # accuracy in some cases for speed;
1689 # 1. don't query if item has one of the other statuses
1690 # 2. don't check transit status if the bib has
1691 # more than 20 items
1693 # FIXME: to avoid having the query the database like this, and to make
1694 # the in transit status count as unavailable for search limiting,
1695 # should map transit status to record indexed in Zebra.
1697 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1698 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1701 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1702 if ( $item->{wthdrawn}
1703 || $item->{itemlost}
1705 || $item->{notforloan} > 0
1706 || $item->{hideatopac}
1707 || $reservestatus eq 'Waiting'
1708 || ($transfertwhen ne ''))
1710 $wthdrawn_count++ if $item->{wthdrawn};
1711 $itemlost_count++ if $item->{itemlost};
1712 $itemdamaged_count++ if $item->{damaged};
1713 $item_in_transit_count++ if $transfertwhen ne '';
1714 $item_onhold_count++ if $reservestatus eq 'Waiting';
1715 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1717 # can place hold on item ?
1718 if ((!$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems'))
1719 && !$item->{itemlost}
1720 && !$item->{withdrawn}
1722 $can_place_holds = 1;
1727 my $key = $prefix . $item->{status};
1728 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber hideatopac)) {
1729 $other_items->{$key}->{$_} = $item->{$_};
1731 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1732 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1733 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1734 $other_items->{$key}->{count}++ if $item->{$hbranch};
1735 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1736 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1740 $can_place_holds = 1;
1742 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1743 foreach (qw(branchname itemcallnumber hideatopac)) {
1744 $available_items->{$prefix}->{$_} = $item->{$_};
1746 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1747 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1750 } # notforloan, item level and biblioitem level
1751 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1753 ( C4::Context->preference('maxItemsinSearchResults') )
1754 ? C4::Context->preference('maxItemsinSearchResults') - 1
1756 for my $key ( sort keys %$onloan_items ) {
1757 (++$onloanitemscount > $maxitems) and last;
1758 push @onloan_items_loop, $onloan_items->{$key};
1760 for my $key ( sort keys %$other_items ) {
1761 (++$otheritemscount > $maxitems) and last;
1762 push @other_items_loop, $other_items->{$key};
1764 for my $key ( sort keys %$available_items ) {
1765 (++$availableitemscount > $maxitems) and last;
1766 push @available_items_loop, $available_items->{$key}
1769 # XSLT processing of some stuff
1771 SetUTF8Flag($marcrecord);
1772 $debug && warn $marcrecord->as_formatted;
1773 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1774 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1775 # is not implemented yet
1776 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1777 $search_context, 1);
1778 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1782 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1783 if (!C4::Context->preference("item-level_itypes")) {
1784 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1785 $can_place_holds = 0;
1788 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1789 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1790 $oldbiblio->{items_count} = $items_count;
1791 $oldbiblio->{available_items_loop} = \@available_items_loop;
1792 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1793 $oldbiblio->{other_items_loop} = \@other_items_loop;
1794 $oldbiblio->{availablecount} = $available_count;
1795 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1796 $oldbiblio->{onloancount} = $onloan_count;
1797 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1798 $oldbiblio->{othercount} = $other_count;
1799 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1800 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1801 $oldbiblio->{itemlostcount} = $itemlost_count;
1802 $oldbiblio->{damagedcount} = $itemdamaged_count;
1803 $oldbiblio->{intransitcount} = $item_in_transit_count;
1804 $oldbiblio->{onholdcount} = $item_onhold_count;
1805 $oldbiblio->{orderedcount} = $ordered_count;
1806 $oldbiblio->{isbn} =~
1807 s/-//g; # deleting - in isbn to enable amazon content
1809 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1810 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1811 my $subfields = substr $fieldspec, 3;
1812 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1813 my @alternateholdingsinfo = ();
1814 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1815 my $alternateholdingscount = 0;
1817 for my $field (@holdingsfields) {
1818 my %holding = ( holding => '' );
1819 my $havesubfield = 0;
1820 for my $subfield ($field->subfields()) {
1821 if ((index $subfields, $$subfield[0]) >= 0) {
1822 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1823 $holding{'holding'} .= $$subfield[1];
1827 if ($havesubfield) {
1828 push(@alternateholdingsinfo, \%holding);
1829 $alternateholdingscount++;
1833 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1834 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1837 push( @newresults, $oldbiblio )
1838 if(not $hidelostitems
1839 or (($items_count > $itemlost_count )
1840 && $hidelostitems));
1846 =head2 SearchAcquisitions
1847 Search for acquisitions
1850 sub SearchAcquisitions{
1851 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1853 my $dbh=C4::Context->dbh;
1854 # Variable initialization
1858 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1859 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1860 WHERE dateaccessioned BETWEEN ? AND ?
1863 my (@params,@loopcriteria);
1865 push @params, $datebegin->output("iso");
1866 push @params, $dateend->output("iso");
1868 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1869 if(C4::Context->preference("item-level_itypes")){
1870 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1872 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1874 push @params, @$itemtypes;
1877 if ($criteria =~/itemtype/){
1878 if(C4::Context->preference("item-level_itypes")){
1879 $str .= "AND items.itype=? ";
1881 $str .= "AND biblioitems.itemtype=? ";
1884 if(scalar(@$itemtypes) == 0){
1885 my $itypes = GetItemTypes();
1886 for my $key (keys %$itypes){
1887 push @$itemtypes, $key;
1891 @loopcriteria= @$itemtypes;
1892 }elsif ($criteria=~/itemcallnumber/){
1893 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1894 OR items.itemcallnumber is NULL
1895 OR items.itemcallnumber = '')";
1897 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1899 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1900 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1903 if ($orderby =~ /date_desc/){
1904 $str.=" ORDER BY dateaccessioned DESC";
1906 $str.=" ORDER BY title";
1909 my $qdataacquisitions=$dbh->prepare($str);
1911 my @loopacquisitions;
1912 foreach my $value(@loopcriteria){
1913 push @params,$value;
1915 $cell{"title"}=$value;
1916 $cell{"titlecode"}=$value;
1918 eval{$qdataacquisitions->execute(@params);};
1920 if ($@){ warn "recentacquisitions Error :$@";}
1923 while (my $data=$qdataacquisitions->fetchrow_hashref){
1924 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1926 $cell{"loopdata"}=\@loopdata;
1928 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1931 $qdataacquisitions->finish;
1932 return \@loopacquisitions;
1934 #----------------------------------------------------------------------
1936 # Non-Zebra GetRecords#
1937 #----------------------------------------------------------------------
1941 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1947 $query, $simple_query, $sort_by_ref, $servers_ref,
1948 $results_per_page, $offset, $expanded_facet, $branches,
1951 warn "query =$query" if $DEBUG;
1952 my $result = NZanalyse($query);
1953 warn "results =$result" if $DEBUG;
1955 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1961 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1962 the list is built from an inverted index in the nozebra SQL table
1963 note that title is here only for convenience : the sorting will be very fast when requested on title
1964 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1969 my ( $string, $server ) = @_;
1970 # warn "---------" if $DEBUG;
1971 warn " NZanalyse" if $DEBUG;
1972 # warn "---------" if $DEBUG;
1974 # $server contains biblioserver or authorities, depending on what we search on.
1975 #warn "querying : $string on $server";
1976 $server = 'biblioserver' unless $server;
1978 # if we have a ", replace the content to discard temporarily any and/or/not inside
1980 if ( $string =~ /"/ ) {
1981 $string =~ s/"(.*?)"/__X__/;
1983 warn "commacontent : $commacontent" if $DEBUG;
1986 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1987 # then, call again NZanalyse with $left and $right
1988 # (recursive until we find a leaf (=> something without and/or/not)
1989 # delete repeated operator... Would then go in infinite loop
1990 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1993 #process parenthesis before.
1994 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1997 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1999 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2001 my $leftresult = NZanalyse( $left, $server );
2003 my $rightresult = NZanalyse( $right, $server );
2005 # OK, we have the results for right and left part of the query
2006 # depending of operand, intersect, union or exclude both lists
2007 # to get a result list
2008 if ( $operator eq ' and ' ) {
2009 return NZoperatorAND($leftresult,$rightresult);
2011 elsif ( $operator eq ' or ' ) {
2013 # just merge the 2 strings
2014 return $leftresult . $rightresult;
2016 elsif ( $operator eq ' not ' ) {
2017 return NZoperatorNOT($leftresult,$rightresult);
2021 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2025 warn "string :" . $string if $DEBUG;
2029 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2032 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2034 warn "no parenthesis. left : $left operator: $operator right: $right"
2037 # it's not a leaf, we have a and/or/not
2040 # reintroduce comma content if needed
2041 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2042 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2043 warn "node : $left / $operator / $right\n" if $DEBUG;
2044 my $leftresult = NZanalyse( $left, $server );
2045 my $rightresult = NZanalyse( $right, $server );
2046 warn " leftresult : $leftresult" if $DEBUG;
2047 warn " rightresult : $rightresult" if $DEBUG;
2048 # OK, we have the results for right and left part of the query
2049 # depending of operand, intersect, union or exclude both lists
2050 # to get a result list
2051 if ( $operator eq ' and ' ) {
2052 return NZoperatorAND($leftresult,$rightresult);
2054 elsif ( $operator eq ' or ' ) {
2056 # just merge the 2 strings
2057 return $leftresult . $rightresult;
2059 elsif ( $operator eq ' not ' ) {
2060 return NZoperatorNOT($leftresult,$rightresult);
2064 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2065 die "error : operand unknown : $operator for $string";
2068 # it's a leaf, do the real SQL query and return the result
2071 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2072 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2073 #remove trailing blank at the beginning
2075 warn "leaf:$string" if $DEBUG;
2077 # parse the string in in operator/operand/value again
2081 if ($string =~ /(.*)(>=|<=)(.*)/) {
2088 # warn "handling leaf... left:$left operator:$operator right:$right"
2090 unless ($operator) {
2091 if ($string =~ /(.*)(>|<|=)(.*)/) {
2096 "handling unless (operator)... left:$left operator:$operator right:$right"
2104 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2107 # automatic replace for short operators
2108 $left = 'title' if $left =~ '^ti$';
2109 $left = 'author' if $left =~ '^au$';
2110 $left = 'publisher' if $left =~ '^pb$';
2111 $left = 'subject' if $left =~ '^su$';
2112 $left = 'koha-Auth-Number' if $left =~ '^an$';
2113 $left = 'keyword' if $left =~ '^kw$';
2114 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2115 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2116 my $dbh = C4::Context->dbh;
2117 if ( $operator && $left ne 'keyword' ) {
2118 #do a specific search
2119 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2120 my $sth = $dbh->prepare(
2121 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2123 warn "$left / $operator / $right\n" if $DEBUG;
2125 # split each word, query the DB and build the biblionumbers result
2126 #sanitizing leftpart
2127 $left =~ s/^\s+|\s+$//;
2128 foreach ( split / /, $right ) {
2130 $_ =~ s/^\s+|\s+$//;
2132 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2133 $sth->execute( $server, $left, $_ )
2134 or warn "execute failed: $!";
2135 while ( my ( $line, $value ) = $sth->fetchrow ) {
2137 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2138 # otherwise, fill the result
2139 $biblionumbers .= $line
2140 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2141 warn "result : $value "
2142 . ( $right =~ /\d/ ) . "=="
2143 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2146 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2148 warn "NZAND" if $DEBUG;
2149 $results = NZoperatorAND($biblionumbers,$results);
2151 $results = $biblionumbers;
2156 #do a complete search (all indexes), if index='kw' do complete search too.
2157 my $sth = $dbh->prepare(
2158 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2161 # split each word, query the DB and build the biblionumbers result
2162 foreach ( split / /, $string ) {
2163 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2164 warn "search on all indexes on $_" if $DEBUG;
2167 $sth->execute( $server, $_ );
2168 while ( my $line = $sth->fetchrow ) {
2169 $biblionumbers .= $line;
2172 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2174 $results = NZoperatorAND($biblionumbers,$results);
2177 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2178 $results = $biblionumbers;
2182 warn "return : $results for LEAF : $string" if $DEBUG;
2185 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2189 my ($rightresult, $leftresult)=@_;
2191 my @leftresult = split /;/, $leftresult;
2192 warn " @leftresult / $rightresult \n" if $DEBUG;
2194 # my @rightresult = split /;/,$leftresult;
2197 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2198 # the result is stored twice, to have the same weight for AND than OR.
2199 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2200 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2201 foreach (@leftresult) {
2204 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2205 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2206 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2208 "$value-$countvalue;$value-$countvalue;";
2211 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2212 return $finalresult;
2216 my ($rightresult, $leftresult)=@_;
2217 return $rightresult.$leftresult;
2221 my ($leftresult, $rightresult)=@_;
2223 my @leftresult = split /;/, $leftresult;
2225 # my @rightresult = split /;/,$leftresult;
2227 foreach (@leftresult) {
2229 $value=$1 if $value=~m/(.*)-\d+$/;
2230 unless ($rightresult =~ "$value-") {
2231 $finalresult .= "$_;";
2234 return $finalresult;
2239 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2246 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2247 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2249 # order title asc by default
2250 # $ordering = '1=36 <i' unless $ordering;
2251 $results_per_page = 20 unless $results_per_page;
2252 $offset = 0 unless $offset;
2253 my $dbh = C4::Context->dbh;
2256 # order by POPULARITY
2258 if ( $ordering =~ /popularity/ ) {
2262 # popularity is not in MARC record, it's builded from a specific query
2264 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2265 foreach ( split /;/, $biblionumbers ) {
2266 my ( $biblionumber, $title ) = split /,/, $_;
2267 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2268 $sth->execute($biblionumber);
2269 my $popularity = $sth->fetchrow || 0;
2271 # hint : the key is popularity.title because we can have
2272 # many results with the same popularity. In this case, sub-ordering is done by title
2273 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2274 # (un-frequent, I agree, but we won't forget anything that way ;-)
2275 $popularity{ sprintf( "%10d", $popularity ) . $title
2276 . $biblionumber } = $biblionumber;
2279 # sort the hash and return the same structure as GetRecords (Zebra querying)
2282 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2283 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2284 $result_hash->{'RECORDS'}[ $numbers++ ] =
2285 $result{ $popularity{$key} }->as_usmarc();
2288 else { # sort popularity ASC
2289 foreach my $key ( sort ( keys %popularity ) ) {
2290 $result_hash->{'RECORDS'}[ $numbers++ ] =
2291 $result{ $popularity{$key} }->as_usmarc();
2294 my $finalresult = ();
2295 $result_hash->{'hits'} = $numbers;
2296 $finalresult->{'biblioserver'} = $result_hash;
2297 return $finalresult;
2303 elsif ( $ordering =~ /author/ ) {
2305 foreach ( split /;/, $biblionumbers ) {
2306 my ( $biblionumber, $title ) = split /,/, $_;
2307 my $record = GetMarcBiblio($biblionumber);
2309 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2310 $author = $record->subfield( '200', 'f' );
2311 $author = $record->subfield( '700', 'a' ) unless $author;
2314 $author = $record->subfield( '100', 'a' );
2317 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2318 # and we don't want to get only 1 result for each of them !!!
2319 $result{ $author . $biblionumber } = $record;
2322 # sort the hash and return the same structure as GetRecords (Zebra querying)
2325 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2326 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2327 $result_hash->{'RECORDS'}[ $numbers++ ] =
2328 $result{$key}->as_usmarc();
2331 else { # sort by author ASC
2332 foreach my $key ( sort ( keys %result ) ) {
2333 $result_hash->{'RECORDS'}[ $numbers++ ] =
2334 $result{$key}->as_usmarc();
2337 my $finalresult = ();
2338 $result_hash->{'hits'} = $numbers;
2339 $finalresult->{'biblioserver'} = $result_hash;
2340 return $finalresult;
2343 # ORDER BY callnumber
2346 elsif ( $ordering =~ /callnumber/ ) {
2348 foreach ( split /;/, $biblionumbers ) {
2349 my ( $biblionumber, $title ) = split /,/, $_;
2350 my $record = GetMarcBiblio($biblionumber);
2352 my $frameworkcode = GetFrameworkCode($biblionumber);
2353 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2354 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2355 unless $callnumber_tag;
2356 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2357 $callnumber = $record->subfield( '200', 'f' );
2359 $callnumber = $record->subfield( '100', 'a' );
2362 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2363 # and we don't want to get only 1 result for each of them !!!
2364 $result{ $callnumber . $biblionumber } = $record;
2367 # sort the hash and return the same structure as GetRecords (Zebra querying)
2370 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2371 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2372 $result_hash->{'RECORDS'}[ $numbers++ ] =
2373 $result{$key}->as_usmarc();
2376 else { # sort by title ASC
2377 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2378 $result_hash->{'RECORDS'}[ $numbers++ ] =
2379 $result{$key}->as_usmarc();
2382 my $finalresult = ();
2383 $result_hash->{'hits'} = $numbers;
2384 $finalresult->{'biblioserver'} = $result_hash;
2385 return $finalresult;
2387 elsif ( $ordering =~ /pubdate/ ) { #pub year
2389 foreach ( split /;/, $biblionumbers ) {
2390 my ( $biblionumber, $title ) = split /,/, $_;
2391 my $record = GetMarcBiblio($biblionumber);
2392 my ( $publicationyear_tag, $publicationyear_subfield ) =
2393 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2394 my $publicationyear =
2395 $record->subfield( $publicationyear_tag,
2396 $publicationyear_subfield );
2398 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2399 # and we don't want to get only 1 result for each of them !!!
2400 $result{ $publicationyear . $biblionumber } = $record;
2403 # sort the hash and return the same structure as GetRecords (Zebra querying)
2406 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2407 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2408 $result_hash->{'RECORDS'}[ $numbers++ ] =
2409 $result{$key}->as_usmarc();
2412 else { # sort by pub year ASC
2413 foreach my $key ( sort ( keys %result ) ) {
2414 $result_hash->{'RECORDS'}[ $numbers++ ] =
2415 $result{$key}->as_usmarc();
2418 my $finalresult = ();
2419 $result_hash->{'hits'} = $numbers;
2420 $finalresult->{'biblioserver'} = $result_hash;
2421 return $finalresult;
2427 elsif ( $ordering =~ /title/ ) {
2429 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2431 foreach ( split /;/, $biblionumbers ) {
2432 my ( $biblionumber, $title ) = split /,/, $_;
2434 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2435 # and we don't want to get only 1 result for each of them !!!
2436 # hint & speed improvement : we can order without reading the record
2437 # so order, and read records only for the requested page !
2438 $result{ $title . $biblionumber } = $biblionumber;
2441 # sort the hash and return the same structure as GetRecords (Zebra querying)
2444 if ( $ordering eq 'title_az' ) { # sort by title desc
2445 foreach my $key ( sort ( keys %result ) ) {
2446 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2449 else { # sort by title ASC
2450 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2451 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2455 # limit the $results_per_page to result size if it's more
2456 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2458 # for the requested page, replace biblionumber by the complete record
2459 # speed improvement : avoid reading too much things
2461 my $counter = $offset ;
2462 $counter <= $offset + $results_per_page ;
2466 $result_hash->{'RECORDS'}[$counter] =
2467 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2469 my $finalresult = ();
2470 $result_hash->{'hits'} = $numbers;
2471 $finalresult->{'biblioserver'} = $result_hash;
2472 return $finalresult;
2479 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2482 foreach ( split /;/, $biblionumbers ) {
2483 my ( $biblionumber, $title ) = split /,/, $_;
2484 $title =~ /(.*)-(\d)/;
2489 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2490 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2491 # biblio N has ranking = 6
2492 $count_ranking{$biblionumber} += $ranking;
2495 # build the result by "inverting" the count_ranking hash
2496 # 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
2498 foreach ( keys %count_ranking ) {
2499 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2502 # sort the hash and return the same structure as GetRecords (Zebra querying)
2505 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2506 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2509 # limit the $results_per_page to result size if it's more
2510 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2512 # for the requested page, replace biblionumber by the complete record
2513 # speed improvement : avoid reading too much things
2515 my $counter = $offset ;
2516 $counter <= $offset + $results_per_page ;
2520 $result_hash->{'RECORDS'}[$counter] =
2521 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2522 if $result_hash->{'RECORDS'}[$counter];
2524 my $finalresult = ();
2525 $result_hash->{'hits'} = $numbers;
2526 $finalresult->{'biblioserver'} = $result_hash;
2527 return $finalresult;
2531 =head2 enabled_staff_search_views
2533 %hash = enabled_staff_search_views()
2535 This function returns a hash that contains three flags obtained from the system
2536 preferences, used to determine whether a particular staff search results view
2541 =item C<Output arg:>
2543 * $hash{can_view_MARC} is true only if the MARC view is enabled
2544 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2545 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2547 =item C<usage in the script:>
2551 $template->param ( C4::Search::enabled_staff_search_views );
2555 sub enabled_staff_search_views
2558 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2559 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2560 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2564 sub AddSearchHistory{
2565 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2566 my $dbh = C4::Context->dbh;
2568 # Add the request the user just made
2569 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2570 my $sth = $dbh->prepare($sql);
2571 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2572 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2575 sub GetSearchHistory{
2576 my ($borrowernumber,$session)=@_;
2577 my $dbh = C4::Context->dbh;
2579 # Add the request the user just made
2580 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2581 my $sth = $dbh->prepare($query);
2582 $sth->execute($borrowernumber, $session);
2583 return $sth->fetchall_hashref({});
2586 =head2 z3950_search_args
2588 $arrayref = z3950_search_args($matchpoints)
2590 This function returns an array reference that contains the search parameters to be
2591 passed to the Z39.50 search script (z3950_search.pl). The array elements
2592 are hash refs whose keys are name, value and encvalue, and whose values are the
2593 name of a search parameter, the value of that search parameter and the URL encoded
2594 value of that parameter.
2596 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2598 The search parameter values are obtained from the bibliographic record whose
2599 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2601 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2602 a general purpose search argument. In this case, the returned array contains only
2603 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2605 If a search parameter value is undefined or empty, it is not included in the returned
2608 The returned array reference may be passed directly to the template parameters.
2612 =item C<Output arg:>
2614 * $array containing hash refs as described above
2616 =item C<usage in the script:>
2620 $data = Biblio::GetBiblioData($bibno);
2621 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2625 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2629 sub z3950_search_args {
2631 $bibrec = { title => $bibrec } if !ref $bibrec;
2633 for my $field (qw/ lccn isbn issn title author dewey subject /)
2635 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2636 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2641 =head2 BiblioAddAuthorities
2643 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2645 this function finds the authorities linked to the biblio
2646 * search in the authority DB for the same authid (in $9 of the biblio)
2647 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2648 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2649 OR adds a new authority record
2655 * $record is the MARC record in question (marc blob)
2656 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2658 =item C<Output arg:>
2660 * $countlinked is the number of authorities records that are linked to this authority
2664 * I had to add this to Search.pm (instead of the logical Biblio.pm) because of a circular dependency (this sub uses SimpleSearch, and Search.pm uses Biblio.pm)
2671 sub BiblioAddAuthorities{
2672 my ( $record, $frameworkcode ) = @_;
2673 my $dbh=C4::Context->dbh;
2674 my $query=$dbh->prepare(qq|
2675 SELECT authtypecode,tagfield
2676 FROM marc_subfield_structure
2677 WHERE frameworkcode=?
2678 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2679 # SELECT authtypecode,tagfield
2680 # FROM marc_subfield_structure
2681 # WHERE frameworkcode=?
2682 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2683 $query->execute($frameworkcode);
2684 my ($countcreated,$countlinked);
2685 while (my $data=$query->fetchrow_hashref){
2686 foreach my $field ($record->field($data->{tagfield})){
2687 next if ($field->subfield('3')||$field->subfield('9'));
2688 # No authorities id in the tag.
2689 # Search if there is any authorities to link to.
2690 my $query='at='.$data->{authtypecode}.' ';
2691 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2692 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2693 # there is only 1 result
2695 warn "BIBLIOADDSAUTHORITIES: $error";
2698 if ( @{$results} == 1 ) {
2699 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2700 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2702 } elsif ( @{$results} > 1 ) {
2703 #More than One result
2704 #This can comes out of a lack of a subfield.
2705 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2706 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2709 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2710 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2711 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2712 my $authtypedata=C4::AuthoritiesMarc::GetAuthType($data->{authtypecode});
2713 next unless $authtypedata;
2714 my $marcrecordauth=MARC::Record->new();
2715 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2716 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2717 $marcrecordauth->insert_fields_ordered($authfield);
2719 # bug 2317: ensure new authority knows it's using UTF-8; currently
2720 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2721 # automatically for UNIMARC (by not transcoding)
2722 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2723 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2724 # of change to a core API just before the 3.0 release.
2725 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2726 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2729 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2731 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2733 $field->add_subfields('9'=>$authid);
2737 return ($countlinked,$countcreated);
2740 =head2 GetDistinctValues($field);
2742 C<$field> is a reference to the fields array
2746 sub GetDistinctValues {
2747 my ($fieldname,$string)=@_;
2748 # returns a reference to a hash of references to branches...
2749 if ($fieldname=~/\./){
2750 my ($table,$column)=split /\./, $fieldname;
2751 my $dbh = C4::Context->dbh;
2752 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2753 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 ");
2755 my $elements=$sth->fetchall_arrayref({});
2760 my @servers=qw<biblioserver authorityserver>;
2761 my (@zconns,@results);
2762 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2763 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2766 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2769 # The big moment: asynchronously retrieve results from all servers
2771 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2772 my $ev = $zconns[ $i - 1 ]->last_event();
2773 if ( $ev == ZOOM::Event::ZEND ) {
2774 next unless $results[ $i - 1 ];
2775 my $size = $results[ $i - 1 ]->size();
2777 for (my $j=0;$j<$size;$j++){
2779 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2780 push @elements, \%hashscan;
2790 END { } # module clean-up code here (global destructor)
2797 Koha Development Team <http://koha-community.org/>