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
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
74 #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)
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,
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, $branches,
308 my @servers = @$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;
473 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
474 ($facets->[$k]) or next;
475 my @fcodes = @{$facets->[$k]->{'tags'}};
476 my $sfcode = $facets->[$k]->{'subfield'};
478 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
479 my $render_record = $results[ $i - 1 ]->record($j)->render();
482 foreach my $fcode (@fcodes) {
485 my $field_pattern = '\n'.$fcode.' ([^\n]+)';
486 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
488 foreach my $field_token (@field_tokens) {
489 my $subfield_pattern = '\$'.$sfcode.' ([^\$]+)';
490 my @subfield_values = ( $field_token =~ /$subfield_pattern/g );
492 foreach my $subfield_value (@subfield_values) {
494 my $data = $subfield_value;
495 $data =~ s/^\s+//; # trim left
496 $data =~ s/\s+$//; # trim right
498 unless ( $data ~~ @used_datas ) {
499 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
500 push @used_datas, $data;
507 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} = $facets->[$k]->{'label_value'};
508 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} = $facets->[$k]->{'expanded'};
514 # warn "connection ", $i-1, ": $size hits";
515 # warn $results[$i-1]->record(0)->render() if $size > 0;
518 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
520 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
521 keys %$facets_counter )
524 my $number_of_facets;
525 my @this_facets_array;
528 $facets_counter->{$link_value}->{$b}
529 <=> $facets_counter->{$link_value}->{$a}
530 } keys %{ $facets_counter->{$link_value} }
534 if ( ( $number_of_facets < 6 )
535 || ( $expanded_facet eq $link_value )
536 || ( $facets_info->{$link_value}->{'expanded'} ) )
539 # Sanitize the link value ), ( will cause errors with CCL,
540 my $facet_link_value = $one_facet;
541 $facet_link_value =~ s/(\(|\))/ /g;
543 # fix the length that will display in the label,
544 my $facet_label_value = $one_facet;
545 my $facet_max_length =
546 C4::Context->preference('FacetLabelTruncationLength') || 20;
548 substr( $one_facet, 0, $facet_max_length ) . "..."
549 if length($facet_label_value) > $facet_max_length;
551 # if it's a branch, label by the name, not the code,
552 if ( $link_value =~ /branch/ ) {
553 if (defined $branches
554 && ref($branches) eq "HASH"
555 && defined $branches->{$one_facet}
556 && ref ($branches->{$one_facet}) eq "HASH")
559 $branches->{$one_facet}->{'branchname'};
562 $facet_label_value = "*";
566 # but we're down with the whole label being in the link's title.
567 push @this_facets_array, {
568 facet_count => $facets_counter->{$link_value}->{$one_facet},
569 facet_label_value => $facet_label_value,
570 facet_title_value => $one_facet,
571 facet_link_value => $facet_link_value,
572 type_link_value => $link_value,
577 # handle expanded option
578 unless ( $facets_info->{$link_value}->{'expanded'} ) {
580 if ( ( $number_of_facets > 6 )
581 && ( $expanded_facet ne $link_value ) );
584 type_link_value => $link_value,
585 type_id => $link_value . "_id",
586 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
587 facets => \@this_facets_array,
588 expandable => $expandable,
589 expand => $link_value,
590 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
595 return ( undef, $results_hashref, \@facets_loop );
600 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
601 $results_per_page, $offset, $expanded_facet, $branches,
605 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
607 $paz->search($simple_query);
608 sleep 1; # FIXME: WHY?
611 my $results_hashref = {};
612 my $stats = XMLin($paz->stat);
613 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
615 # for a grouped search result, the number of hits
616 # is the number of groups returned; 'bib_hits' will have
617 # the total number of bibs.
618 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
619 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
621 HIT: foreach my $hit (@{ $results->{'hit'} }) {
622 my $recid = $hit->{recid}->[0];
624 my $work_title = $hit->{'md-work-title'}->[0];
626 if (exists $hit->{'md-work-author'}) {
627 $work_author = $hit->{'md-work-author'}->[0];
629 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
631 my $result_group = {};
632 $result_group->{'group_label'} = $group_label;
633 $result_group->{'group_merge_key'} = $recid;
636 if (exists $hit->{count}) {
637 $count = $hit->{count}->[0];
639 $result_group->{'group_count'} = $count;
641 for (my $i = 0; $i < $count; $i++) {
642 # FIXME -- may need to worry about diacritics here
643 my $rec = $paz->record($recid, $i);
644 push @{ $result_group->{'RECORDS'} }, $rec;
647 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
650 # pass through facets
651 my $termlist_xml = $paz->termlist('author,subject');
652 my $terms = XMLin($termlist_xml, forcearray => 1);
653 my @facets_loop = ();
654 #die Dumper($results);
655 # foreach my $list (sort keys %{ $terms->{'list'} }) {
657 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
659 # facet_label_value => $facet->{'name'}->[0],
662 # push @facets_loop, ( {
663 # type_label => $list,
664 # facets => \@facets,
668 return ( undef, $results_hashref, \@facets_loop );
672 sub _remove_stopwords {
673 my ( $operand, $index ) = @_;
674 my @stopwords_removed;
676 # phrase and exact-qualified indexes shouldn't have stopwords removed
677 if ( $index !~ m/phr|ext/ ) {
679 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
680 # we use IsAlpha unicode definition, to deal correctly with diacritics.
681 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
682 # is a stopword, we'd get "çon" and wouldn't find anything...
684 foreach ( keys %{ C4::Context->stopwords } ) {
685 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
686 if ( my ($matched) = ($operand =~
687 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
689 $operand =~ s/\Q$matched\E/ /gi;
690 push @stopwords_removed, $_;
694 return ( $operand, \@stopwords_removed );
698 sub _detect_truncation {
699 my ( $operand, $index ) = @_;
700 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
703 my @wordlist = split( /\s/, $operand );
704 foreach my $word (@wordlist) {
705 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
706 push @rightlefttruncated, $word;
708 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
709 push @lefttruncated, $word;
711 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
712 push @righttruncated, $word;
714 elsif ( index( $word, "*" ) < 0 ) {
715 push @nontruncated, $word;
718 push @regexpr, $word;
722 \@nontruncated, \@righttruncated, \@lefttruncated,
723 \@rightlefttruncated, \@regexpr
728 sub _build_stemmed_operand {
729 my ($operand,$lang) = @_;
730 require Lingua::Stem::Snowball ;
733 # If operand contains a digit, it is almost certainly an identifier, and should
734 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
735 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
736 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
737 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
738 return $operand if $operand =~ /\d/;
740 # FIXME: the locale should be set based on the user's language and/or search choice
742 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
743 encoding => "UTF-8" );
745 my @words = split( / /, $operand );
746 my @stems = $stemmer->stem(\@words);
747 for my $stem (@stems) {
748 $stemmed_operand .= "$stem";
749 $stemmed_operand .= "?"
750 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
751 $stemmed_operand .= " ";
753 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
754 return $stemmed_operand;
758 sub _build_weighted_query {
760 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
761 # pretty well but could work much better if we had a smarter query parser
762 my ( $operand, $stemmed_operand, $index ) = @_;
763 my $stemming = C4::Context->preference("QueryStemming") || 0;
764 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
765 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
767 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
769 # Keyword, or, no index specified
770 if ( ( $index eq 'kw' ) || ( !$index ) ) {
772 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
773 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
774 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
775 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
776 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
777 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
778 if $fuzzy_enabled; # add fuzzy, word list
779 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
780 if ( $stemming and $stemmed_operand )
781 ; # add stemming, right truncation
782 $weighted_query .= " or wrdl,r9=\"$operand\"";
784 # embedded sorting: 0 a-z; 1 z-a
785 # $weighted_query .= ") or (sort1,aut=1";
788 # Barcode searches should skip this process
789 elsif ( $index eq 'bc' ) {
790 $weighted_query .= "bc=\"$operand\"";
793 # Authority-number searches should skip this process
794 elsif ( $index eq 'an' ) {
795 $weighted_query .= "an=\"$operand\"";
798 # If the index already has more than one qualifier, wrap the operand
799 # in quotes and pass it back (assumption is that the user knows what they
800 # are doing and won't appreciate us mucking up their query
801 elsif ( $index =~ ',' ) {
802 $weighted_query .= " $index=\"$operand\"";
805 #TODO: build better cases based on specific search indexes
807 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
808 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
809 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
811 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
814 $weighted_query .= "))"; # close rank specification
815 return $weighted_query;
820 Return an array with available indexes.
842 'Author-personal-bibliography',
852 'Chronological-subdivision',
862 'Conference-name-heading',
863 'Conference-name-see',
864 'Conference-name-seealso',
869 'Corporate-name-heading',
870 'Corporate-name-see',
871 'Corporate-name-seealso',
873 'date-entered-on-file',
874 'Date-of-acquisition',
875 'Date-of-publication',
876 'Dewey-classification',
883 'Geographic-subdivision',
886 'Heading-use-main-or-added-entry',
887 'Heading-use-series-added-entry ',
888 'Heading-use-subject-added-entry',
906 'Local-classification',
909 'Match-heading-see-from',
917 'Name-geographic-heading',
918 'Name-geographic-see',
919 'Name-geographic-seealso',
927 'Personal-name-heading',
929 'Personal-name-seealso',
936 'Record-control-number',
947 'Subject-heading-thesaurus',
948 'Subject-name-personal',
949 'Subject-subdivision',
959 'Term-genre-form-heading',
960 'Term-genre-form-see',
961 'Term-genre-form-seealso',
968 'Title-uniform-heading',
970 'Title-uniform-seealso',
980 'classification-source',
982 'coded-location-qualifier',
993 'Local-classification',
996 'materials-specified',
1005 'replacementpricedate',
1023 $simple_query, $query_cgi,
1024 $query_desc, $limit,
1025 $limit_cgi, $limit_desc,
1026 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1028 Build queries and limits in CCL, CGI, Human,
1029 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1031 See verbose embedded documentation.
1037 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1039 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1042 my @operators = $operators ? @$operators : ();
1043 my @indexes = $indexes ? @$indexes : ();
1044 my @operands = $operands ? @$operands : ();
1045 my @limits = $limits ? @$limits : ();
1046 my @sort_by = $sort_by ? @$sort_by : ();
1048 my $stemming = C4::Context->preference("QueryStemming") || 0;
1049 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1050 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1051 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1052 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1054 # no stemming/weight/fuzzy in NoZebra
1055 if ( C4::Context->preference("NoZebra") ) {
1059 $auto_truncation = 0;
1062 my $query = $operands[0];
1063 my $simple_query = $operands[0];
1065 # initialize the variables we're passing back
1074 my $stopwords_removed; # flag to determine if stopwords have been removed
1077 my $cclindexes = getIndexes();
1078 if ( $query !~ /\s*ccl=/ ) {
1079 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1081 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1083 $query = "ccl=$query" if $cclq;
1086 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1088 if ( $query =~ /^ccl=/ ) {
1090 # This is needed otherwise ccl= and &limit won't work together, and
1091 # this happens when selecting a subject on the opac-detail page
1093 $q .= ' and '.join(' and ', @limits);
1095 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1097 if ( $query =~ /^cql=/ ) {
1098 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1100 if ( $query =~ /^pqf=/ ) {
1101 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1104 # pass nested queries directly
1105 # FIXME: need better handling of some of these variables in this case
1106 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1107 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1108 # if ( $query =~ /(\(|\))/ ) {
1110 # undef, $query, $simple_query, $query_cgi,
1111 # $query, $limit, $limit_cgi, $limit_desc,
1112 # $stopwords_removed, 'ccl'
1116 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1117 # query operands and indexes and add stemming, truncation, field weighting, etc.
1118 # Once we do so, we'll end up with a value in $query, just like if we had an
1119 # incoming $query from the user
1122 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1123 my $previous_operand
1124 ; # a flag used to keep track if there was a previous query
1125 # if there was, we can apply the current operator
1127 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1129 # COMBINE OPERANDS, INDEXES AND OPERATORS
1130 if ( $operands[$i] ) {
1131 $operands[$i]=~s/^\s+//;
1133 # A flag to determine whether or not to add the index to the query
1136 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1137 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1140 $remove_stopwords = 0;
1142 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1144 my $operand = $operands[$i];
1145 my $index = $indexes[$i];
1147 # Add index-specific attributes
1148 # Date of Publication
1149 if ( $index eq 'yr' ) {
1150 $index .= ",st-numeric";
1152 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1155 # Date of Acquisition
1156 elsif ( $index eq 'acqdate' ) {
1157 $index .= ",st-date-normalized";
1159 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1161 # ISBN,ISSN,Standard Number, don't need special treatment
1162 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1164 $stemming, $auto_truncation,
1165 $weight_fields, $fuzzy_enabled,
1167 ) = ( 0, 0, 0, 0, 0 );
1175 # Set default structure attribute (word list)
1176 my $struct_attr = q{};
1177 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1178 $struct_attr = ",wrdl";
1181 # Some helpful index variants
1182 my $index_plus = $index . $struct_attr . ':';
1183 my $index_plus_comma = $index . $struct_attr . ',';
1186 if ($remove_stopwords) {
1187 ( $operand, $stopwords_removed ) =
1188 _remove_stopwords( $operand, $index );
1189 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1190 warn "REMOVED STOPWORDS: @$stopwords_removed"
1191 if ( $stopwords_removed && $DEBUG );
1194 if ($auto_truncation){
1195 unless ( $index =~ /(st-|phr|ext)/ ) {
1196 #FIXME only valid with LTR scripts
1197 $operand=join(" ",map{
1198 (index($_,"*")>0?"$_":"$_*")
1199 }split (/\s+/,$operand));
1200 warn $operand if $DEBUG;
1205 my $truncated_operand;
1206 my( $nontruncated, $righttruncated, $lefttruncated,
1207 $rightlefttruncated, $regexpr
1208 ) = _detect_truncation( $operand, $index );
1210 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1215 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1216 scalar(@$rightlefttruncated) > 0 )
1219 # Don't field weight or add the index to the query, we do it here
1221 undef $weight_fields;
1222 my $previous_truncation_operand;
1223 if (scalar @$nontruncated) {
1224 $truncated_operand .= "$index_plus @$nontruncated ";
1225 $previous_truncation_operand = 1;
1227 if (scalar @$righttruncated) {
1228 $truncated_operand .= "and " if $previous_truncation_operand;
1229 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1230 $previous_truncation_operand = 1;
1232 if (scalar @$lefttruncated) {
1233 $truncated_operand .= "and " if $previous_truncation_operand;
1234 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1235 $previous_truncation_operand = 1;
1237 if (scalar @$rightlefttruncated) {
1238 $truncated_operand .= "and " if $previous_truncation_operand;
1239 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1240 $previous_truncation_operand = 1;
1243 $operand = $truncated_operand if $truncated_operand;
1244 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1247 my $stemmed_operand;
1248 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1251 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1253 # Handle Field Weighting
1254 my $weighted_operand;
1255 if ($weight_fields) {
1256 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1257 $operand = $weighted_operand;
1261 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1263 # If there's a previous operand, we need to add an operator
1264 if ($previous_operand) {
1266 # User-specified operator
1267 if ( $operators[ $i - 1 ] ) {
1268 $query .= " $operators[$i-1] ";
1269 $query .= " $index_plus " unless $indexes_set;
1270 $query .= " $operand";
1271 $query_cgi .= "&op=$operators[$i-1]";
1272 $query_cgi .= "&idx=$index" if $index;
1273 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1275 " $operators[$i-1] $index_plus $operands[$i]";
1278 # Default operator is and
1281 $query .= "$index_plus " unless $indexes_set;
1282 $query .= "$operand";
1283 $query_cgi .= "&op=and&idx=$index" if $index;
1284 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1285 $query_desc .= " and $index_plus $operands[$i]";
1289 # There isn't a pervious operand, don't need an operator
1292 # Field-weighted queries already have indexes set
1293 $query .= " $index_plus " unless $indexes_set;
1295 $query_desc .= " $index_plus $operands[$i]";
1296 $query_cgi .= "&idx=$index" if $index;
1297 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1298 $previous_operand = 1;
1303 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1306 my $group_OR_limits;
1307 my $availability_limit;
1308 foreach my $this_limit (@limits) {
1309 if ( $this_limit =~ /available/ ) {
1311 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1313 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1314 $availability_limit .=
1315 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1316 $limit_cgi .= "&limit=available";
1320 # group_OR_limits, prefixed by mc-
1321 # OR every member of the group
1322 elsif ( $this_limit =~ /mc/ ) {
1324 if ( $this_limit =~ /mc-ccode:/ ) {
1325 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1326 $this_limit =~ tr/"//d;
1327 my ($k,$v) = split(/:/, $this_limit,2);
1328 $this_limit = $k.":\"".$v."\"";
1331 $group_OR_limits .= " or " if $group_OR_limits;
1332 $limit_desc .= " or " if $group_OR_limits;
1333 $group_OR_limits .= "$this_limit";
1334 $limit_cgi .= "&limit=$this_limit";
1335 $limit_desc .= " $this_limit";
1338 # Regular old limits
1340 $limit .= " and " if $limit || $query;
1341 $limit .= "$this_limit";
1342 $limit_cgi .= "&limit=$this_limit";
1343 if ($this_limit =~ /^branch:(.+)/) {
1344 my $branchcode = $1;
1345 my $branchname = GetBranchName($branchcode);
1346 if (defined $branchname) {
1347 $limit_desc .= " branch:$branchname";
1349 $limit_desc .= " $this_limit";
1352 $limit_desc .= " $this_limit";
1356 if ($group_OR_limits) {
1357 $limit .= " and " if ( $query || $limit );
1358 $limit .= "($group_OR_limits)";
1360 if ($availability_limit) {
1361 $limit .= " and " if ( $query || $limit );
1362 $limit .= "($availability_limit)";
1365 # Normalize the query and limit strings
1366 # This is flawed , means we can't search anything with : in it
1367 # if user wants to do ccl or cql, start the query with that
1368 # $query =~ s/:/=/g;
1369 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1370 $query =~ s/(?<=(wrdl)):/=/g;
1371 $query =~ s/(?<=(trn|phr)):/=/g;
1373 for ( $query, $query_desc, $limit, $limit_desc ) {
1374 s/ +/ /g; # remove extra spaces
1375 s/^ //g; # remove any beginning spaces
1376 s/ $//g; # remove any ending spaces
1377 s/==/=/g; # remove double == from query
1379 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1381 for ($query_cgi,$simple_query) {
1384 # append the limit to the query
1385 $query .= " " . $limit;
1389 warn "QUERY:" . $query;
1390 warn "QUERY CGI:" . $query_cgi;
1391 warn "QUERY DESC:" . $query_desc;
1392 warn "LIMIT:" . $limit;
1393 warn "LIMIT CGI:" . $limit_cgi;
1394 warn "LIMIT DESC:" . $limit_desc;
1395 warn "---------\nLeave buildQuery\n---------";
1398 undef, $query, $simple_query, $query_cgi,
1399 $query_desc, $limit, $limit_cgi, $limit_desc,
1400 $stopwords_removed, $query_type
1404 =head2 searchResults
1406 my @search_results = searchResults($search_context, $searchdesc, $hits,
1407 $results_per_page, $offset, $scan,
1408 @marcresults, $hidelostitems);
1410 Format results in a form suitable for passing to the template
1414 # IMO this subroutine is pretty messy still -- it's responsible for
1415 # building the HTML output for the template
1417 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1418 my $dbh = C4::Context->dbh;
1421 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1422 my ($is_opac, $hidelostitems);
1423 if ($search_context eq 'opac') {
1424 $hidelostitems = C4::Context->preference('hidelostitems');
1428 #Build branchnames hash
1430 #get branch information.....
1432 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1434 while ( my $bdata = $bsth->fetchrow_hashref ) {
1435 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1437 # FIXME - We build an authorised values hash here, using the default framework
1438 # though it is possible to have different authvals for different fws.
1440 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1442 # get notforloan authorised value list (see $shelflocations FIXME)
1443 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1445 #Build itemtype hash
1446 #find itemtype & itemtype image
1450 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1453 while ( my $bdata = $bsth->fetchrow_hashref ) {
1454 foreach (qw(description imageurl summary notforloan)) {
1455 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1459 #search item field code
1462 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1465 my ($itemtag) = $sth->fetchrow;
1467 ## find column names of items related to MARC
1468 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1470 my %subfieldstosearch;
1471 while ( ( my $column ) = $sth2->fetchrow ) {
1472 my ( $tagfield, $tagsubfield ) =
1473 &GetMarcFromKohaField( "items." . $column, "" );
1474 $subfieldstosearch{$column} = $tagsubfield;
1477 # handle which records to actually retrieve
1479 if ( $hits && $offset + $results_per_page <= $hits ) {
1480 $times = $offset + $results_per_page;
1483 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1486 my $marcflavour = C4::Context->preference("marcflavour");
1487 # We get the biblionumber position in MARC
1488 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1490 # loop through all of the records we've retrieved
1491 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1492 my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1496 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1497 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1498 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1499 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1500 $oldbiblio->{result_number} = $i + 1;
1502 # add imageurl to itemtype if there is one
1503 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1505 $oldbiblio->{'authorised_value_images'} = ($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 ) ) : [];
1506 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1507 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1508 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1509 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1510 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1512 # edition information, if any
1513 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1514 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1515 # Build summary if there is one (the summary is defined in the itemtypes table)
1516 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1517 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1518 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1519 my @fields = $marcrecord->fields();
1522 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1524 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1525 $tag =~ /(.{3})(.)/;
1526 if($marcrecord->field($1)){
1527 my @abc = $marcrecord->field($1)->subfield($2);
1528 $tags->{$tag} = $#abc + 1 ;
1532 # We catch how many times to repeat this line
1534 foreach my $tag (keys(%$tags)){
1535 $max = $tags->{$tag} if($tags->{$tag} > $max);
1538 # we replace, and repeat each line
1539 for (my $i = 0 ; $i < $max ; $i++){
1540 my $newline = $line;
1542 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1543 $tag =~ /(.{3})(.)/;
1545 if($marcrecord->field($1)){
1546 my @repl = $marcrecord->field($1)->subfield($2);
1547 my $subfieldvalue = $repl[$i];
1549 if (! utf8::is_utf8($subfieldvalue)) {
1550 utf8::decode($subfieldvalue);
1553 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1556 $newsummary .= "$newline\n";
1560 $newsummary =~ s/\[(.*?)]//g;
1561 $newsummary =~ s/\n/<br\/>/g;
1562 $oldbiblio->{summary} = $newsummary;
1565 # Pull out the items fields
1566 my @fields = $marcrecord->field($itemtag);
1567 my $marcflavor = C4::Context->preference("marcflavour");
1568 # adding linked items that belong to host records
1569 my $analyticsfield = '773';
1570 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1571 $analyticsfield = '773';
1572 } elsif ($marcflavor eq 'UNIMARC') {
1573 $analyticsfield = '461';
1575 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1576 my $hostbiblionumber = $hostfield->subfield("0");
1577 my $linkeditemnumber = $hostfield->subfield("9");
1578 if(!$hostbiblionumber eq undef){
1579 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1580 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1581 if(!$hostbiblio eq undef){
1582 my @hostitems = $hostbiblio->field($itemfield);
1583 foreach my $hostitem (@hostitems){
1584 if ($hostitem->subfield("9") eq $linkeditemnumber){
1585 my $linkeditem =$hostitem;
1586 # append linked items if they exist
1587 if (!$linkeditem eq undef){
1588 push (@fields, $linkeditem);}
1595 # Setting item statuses for display
1596 my @available_items_loop;
1597 my @onloan_items_loop;
1598 my @other_items_loop;
1600 my $available_items;
1604 my $ordered_count = 0;
1605 my $available_count = 0;
1606 my $onloan_count = 0;
1607 my $longoverdue_count = 0;
1608 my $other_count = 0;
1609 my $wthdrawn_count = 0;
1610 my $itemlost_count = 0;
1611 my $hideatopac_count = 0;
1612 my $itembinding_count = 0;
1613 my $itemdamaged_count = 0;
1614 my $item_in_transit_count = 0;
1615 my $can_place_holds = 0;
1616 my $item_onhold_count = 0;
1617 my $items_count = scalar(@fields);
1618 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1619 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1621 # loop through every item
1623 foreach my $field (@fields) {
1626 # populate the items hash
1627 foreach my $code ( keys %subfieldstosearch ) {
1628 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1633 my @hi = GetHiddenItemnumbers($item);
1634 $item->{'hideatopac'} = @hi;
1635 push @hiddenitems, @hi;
1638 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1639 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1641 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1642 if ($item->{$hbranch}) {
1643 $item->{'branchname'} = $branches{$item->{$hbranch}};
1645 elsif ($item->{$otherbranch}) { # Last resort
1646 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1649 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1650 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1651 my $userenv = C4::Context->userenv;
1652 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1654 my $key = $prefix . $item->{onloan} . $item->{barcode};
1655 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1656 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1657 $onloan_items->{$key}->{branchname} = $item->{branchname};
1658 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1659 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1660 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1661 # if something's checked out and lost, mark it as 'long overdue'
1662 if ( $item->{itemlost} ) {
1663 $onloan_items->{$prefix}->{longoverdue}++;
1664 $longoverdue_count++;
1665 } else { # can place holds as long as item isn't lost
1666 $can_place_holds = 1;
1670 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1674 if ( $item->{notforloan} == -1 ) {
1678 # is item in transit?
1679 my $transfertwhen = '';
1680 my ($transfertfrom, $transfertto);
1682 # is item on the reserve shelf?
1683 my $reservestatus = '';
1686 unless ($item->{wthdrawn}
1687 || $item->{itemlost}
1689 || $item->{notforloan}
1690 || $items_count > 20) {
1692 # A couple heuristics to limit how many times
1693 # we query the database for item transfer information, sacrificing
1694 # accuracy in some cases for speed;
1696 # 1. don't query if item has one of the other statuses
1697 # 2. don't check transit status if the bib has
1698 # more than 20 items
1700 # FIXME: to avoid having the query the database like this, and to make
1701 # the in transit status count as unavailable for search limiting,
1702 # should map transit status to record indexed in Zebra.
1704 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1705 ($reservestatus, $reserveitem, undef) = C4::Reserves::CheckReserves($item->{itemnumber});
1708 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1709 if ( $item->{wthdrawn}
1710 || $item->{itemlost}
1712 || $item->{notforloan} > 0
1713 || $item->{hideatopac}
1714 || $reservestatus eq 'Waiting'
1715 || ($transfertwhen ne ''))
1717 $wthdrawn_count++ if $item->{wthdrawn};
1718 $itemlost_count++ if $item->{itemlost};
1719 $itemdamaged_count++ if $item->{damaged};
1720 $hideatopac_count++ if $item->{hideatopac};
1721 $item_in_transit_count++ if $transfertwhen ne '';
1722 $item_onhold_count++ if $reservestatus eq 'Waiting';
1723 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1725 # can place hold on item ?
1726 if ((!$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems'))
1727 && !$item->{itemlost}
1728 && !$item->{withdrawn}
1730 $can_place_holds = 1;
1735 my $key = $prefix . $item->{status};
1736 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber hideatopac)) {
1737 $other_items->{$key}->{$_} = $item->{$_};
1739 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1740 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1741 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1742 $other_items->{$key}->{count}++ if $item->{$hbranch};
1743 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1744 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1748 $can_place_holds = 1;
1750 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1751 foreach (qw(branchname itemcallnumber hideatopac)) {
1752 $available_items->{$prefix}->{$_} = $item->{$_};
1754 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1755 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1758 } # notforloan, item level and biblioitem level
1760 next if $is_opac && $hideatopac_count >= $items_count;
1761 next if $hidelostitems && $itemlost_count >= $items_count;
1763 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1764 for my $key ( sort keys %$onloan_items ) {
1765 (++$onloanitemscount > $maxitems) and last;
1766 push @onloan_items_loop, $onloan_items->{$key};
1768 for my $key ( sort keys %$other_items ) {
1769 (++$otheritemscount > $maxitems) and last;
1770 push @other_items_loop, $other_items->{$key};
1772 for my $key ( sort keys %$available_items ) {
1773 (++$availableitemscount > $maxitems) and last;
1774 push @available_items_loop, $available_items->{$key}
1777 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1778 if (!C4::Context->preference("item-level_itypes")) {
1779 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1780 $can_place_holds = 0;
1783 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1784 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1785 $oldbiblio->{items_count} = $items_count;
1786 $oldbiblio->{available_items_loop} = \@available_items_loop;
1787 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1788 $oldbiblio->{other_items_loop} = \@other_items_loop;
1789 $oldbiblio->{availablecount} = $available_count;
1790 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1791 $oldbiblio->{onloancount} = $onloan_count;
1792 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1793 $oldbiblio->{othercount} = $other_count;
1794 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1795 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1796 $oldbiblio->{itemlostcount} = $itemlost_count;
1797 $oldbiblio->{damagedcount} = $itemdamaged_count;
1798 $oldbiblio->{intransitcount} = $item_in_transit_count;
1799 $oldbiblio->{onholdcount} = $item_onhold_count;
1800 $oldbiblio->{orderedcount} = $ordered_count;
1801 # deleting - in isbn to enable amazon content
1802 $oldbiblio->{isbn} =~ s/-//g;
1804 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1805 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1806 my $subfields = substr $fieldspec, 3;
1807 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1808 my @alternateholdingsinfo = ();
1809 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1810 my $alternateholdingscount = 0;
1812 for my $field (@holdingsfields) {
1813 my %holding = ( holding => '' );
1814 my $havesubfield = 0;
1815 for my $subfield ($field->subfields()) {
1816 if ((index $subfields, $$subfield[0]) >= 0) {
1817 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1818 $holding{'holding'} .= $$subfield[1];
1822 if ($havesubfield) {
1823 push(@alternateholdingsinfo, \%holding);
1824 $alternateholdingscount++;
1828 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1829 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1832 # XSLT processing of some stuff
1833 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1834 SetUTF8Flag($marcrecord);
1835 $debug && warn $marcrecord->as_formatted;
1836 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1837 # is not implemented yet
1838 $oldbiblio->{XSLTResultsRecord}
1839 = XSLTParse4Display($oldbiblio->{biblionumber},
1843 1, # clean up the problematic ampersand entities that Zebra outputs
1849 push( @newresults, $oldbiblio );
1855 =head2 SearchAcquisitions
1856 Search for acquisitions
1859 sub SearchAcquisitions{
1860 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1862 my $dbh=C4::Context->dbh;
1863 # Variable initialization
1867 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1868 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1869 WHERE dateaccessioned BETWEEN ? AND ?
1872 my (@params,@loopcriteria);
1874 push @params, $datebegin->output("iso");
1875 push @params, $dateend->output("iso");
1877 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1878 if(C4::Context->preference("item-level_itypes")){
1879 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1881 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1883 push @params, @$itemtypes;
1886 if ($criteria =~/itemtype/){
1887 if(C4::Context->preference("item-level_itypes")){
1888 $str .= "AND items.itype=? ";
1890 $str .= "AND biblioitems.itemtype=? ";
1893 if(scalar(@$itemtypes) == 0){
1894 my $itypes = GetItemTypes();
1895 for my $key (keys %$itypes){
1896 push @$itemtypes, $key;
1900 @loopcriteria= @$itemtypes;
1901 }elsif ($criteria=~/itemcallnumber/){
1902 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1903 OR items.itemcallnumber is NULL
1904 OR items.itemcallnumber = '')";
1906 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1908 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1909 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1912 if ($orderby =~ /date_desc/){
1913 $str.=" ORDER BY dateaccessioned DESC";
1915 $str.=" ORDER BY title";
1918 my $qdataacquisitions=$dbh->prepare($str);
1920 my @loopacquisitions;
1921 foreach my $value(@loopcriteria){
1922 push @params,$value;
1924 $cell{"title"}=$value;
1925 $cell{"titlecode"}=$value;
1927 eval{$qdataacquisitions->execute(@params);};
1929 if ($@){ warn "recentacquisitions Error :$@";}
1932 while (my $data=$qdataacquisitions->fetchrow_hashref){
1933 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1935 $cell{"loopdata"}=\@loopdata;
1937 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1940 $qdataacquisitions->finish;
1941 return \@loopacquisitions;
1943 #----------------------------------------------------------------------
1945 # Non-Zebra GetRecords#
1946 #----------------------------------------------------------------------
1950 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1956 $query, $simple_query, $sort_by_ref, $servers_ref,
1957 $results_per_page, $offset, $expanded_facet, $branches,
1960 warn "query =$query" if $DEBUG;
1961 my $result = NZanalyse($query);
1962 warn "results =$result" if $DEBUG;
1964 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1970 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1971 the list is built from an inverted index in the nozebra SQL table
1972 note that title is here only for convenience : the sorting will be very fast when requested on title
1973 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1978 my ( $string, $server ) = @_;
1979 # warn "---------" if $DEBUG;
1980 warn " NZanalyse" if $DEBUG;
1981 # warn "---------" if $DEBUG;
1983 # $server contains biblioserver or authorities, depending on what we search on.
1984 #warn "querying : $string on $server";
1985 $server = 'biblioserver' unless $server;
1987 # if we have a ", replace the content to discard temporarily any and/or/not inside
1989 if ( $string =~ /"/ ) {
1990 $string =~ s/"(.*?)"/__X__/;
1992 warn "commacontent : $commacontent" if $DEBUG;
1995 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1996 # then, call again NZanalyse with $left and $right
1997 # (recursive until we find a leaf (=> something without and/or/not)
1998 # delete repeated operator... Would then go in infinite loop
1999 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
2002 #process parenthesis before.
2003 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
2006 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
2008 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2010 my $leftresult = NZanalyse( $left, $server );
2012 my $rightresult = NZanalyse( $right, $server );
2014 # OK, we have the results for right and left part of the query
2015 # depending of operand, intersect, union or exclude both lists
2016 # to get a result list
2017 if ( $operator eq ' and ' ) {
2018 return NZoperatorAND($leftresult,$rightresult);
2020 elsif ( $operator eq ' or ' ) {
2022 # just merge the 2 strings
2023 return $leftresult . $rightresult;
2025 elsif ( $operator eq ' not ' ) {
2026 return NZoperatorNOT($leftresult,$rightresult);
2030 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2034 warn "string :" . $string if $DEBUG;
2038 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2041 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2043 warn "no parenthesis. left : $left operator: $operator right: $right"
2046 # it's not a leaf, we have a and/or/not
2049 # reintroduce comma content if needed
2050 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2051 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2052 warn "node : $left / $operator / $right\n" if $DEBUG;
2053 my $leftresult = NZanalyse( $left, $server );
2054 my $rightresult = NZanalyse( $right, $server );
2055 warn " leftresult : $leftresult" if $DEBUG;
2056 warn " rightresult : $rightresult" if $DEBUG;
2057 # OK, we have the results for right and left part of the query
2058 # depending of operand, intersect, union or exclude both lists
2059 # to get a result list
2060 if ( $operator eq ' and ' ) {
2061 return NZoperatorAND($leftresult,$rightresult);
2063 elsif ( $operator eq ' or ' ) {
2065 # just merge the 2 strings
2066 return $leftresult . $rightresult;
2068 elsif ( $operator eq ' not ' ) {
2069 return NZoperatorNOT($leftresult,$rightresult);
2073 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2074 die "error : operand unknown : $operator for $string";
2077 # it's a leaf, do the real SQL query and return the result
2080 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2081 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2082 #remove trailing blank at the beginning
2084 warn "leaf:$string" if $DEBUG;
2086 # parse the string in in operator/operand/value again
2090 if ($string =~ /(.*)(>=|<=)(.*)/) {
2097 # warn "handling leaf... left:$left operator:$operator right:$right"
2099 unless ($operator) {
2100 if ($string =~ /(.*)(>|<|=)(.*)/) {
2105 "handling unless (operator)... left:$left operator:$operator right:$right"
2113 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2116 # automatic replace for short operators
2117 $left = 'title' if $left =~ '^ti$';
2118 $left = 'author' if $left =~ '^au$';
2119 $left = 'publisher' if $left =~ '^pb$';
2120 $left = 'subject' if $left =~ '^su$';
2121 $left = 'koha-Auth-Number' if $left =~ '^an$';
2122 $left = 'keyword' if $left =~ '^kw$';
2123 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2124 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2125 my $dbh = C4::Context->dbh;
2126 if ( $operator && $left ne 'keyword' ) {
2127 #do a specific search
2128 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2129 my $sth = $dbh->prepare(
2130 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2132 warn "$left / $operator / $right\n" if $DEBUG;
2134 # split each word, query the DB and build the biblionumbers result
2135 #sanitizing leftpart
2136 $left =~ s/^\s+|\s+$//;
2137 foreach ( split / /, $right ) {
2139 $_ =~ s/^\s+|\s+$//;
2141 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2142 $sth->execute( $server, $left, $_ )
2143 or warn "execute failed: $!";
2144 while ( my ( $line, $value ) = $sth->fetchrow ) {
2146 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2147 # otherwise, fill the result
2148 $biblionumbers .= $line
2149 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2150 warn "result : $value "
2151 . ( $right =~ /\d/ ) . "=="
2152 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2155 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2157 warn "NZAND" if $DEBUG;
2158 $results = NZoperatorAND($biblionumbers,$results);
2160 $results = $biblionumbers;
2165 #do a complete search (all indexes), if index='kw' do complete search too.
2166 my $sth = $dbh->prepare(
2167 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2170 # split each word, query the DB and build the biblionumbers result
2171 foreach ( split / /, $string ) {
2172 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2173 warn "search on all indexes on $_" if $DEBUG;
2176 $sth->execute( $server, $_ );
2177 while ( my $line = $sth->fetchrow ) {
2178 $biblionumbers .= $line;
2181 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2183 $results = NZoperatorAND($biblionumbers,$results);
2186 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2187 $results = $biblionumbers;
2191 warn "return : $results for LEAF : $string" if $DEBUG;
2194 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2198 my ($rightresult, $leftresult)=@_;
2200 my @leftresult = split /;/, $leftresult;
2201 warn " @leftresult / $rightresult \n" if $DEBUG;
2203 # my @rightresult = split /;/,$leftresult;
2206 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2207 # the result is stored twice, to have the same weight for AND than OR.
2208 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2209 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2210 foreach (@leftresult) {
2213 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2214 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2215 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2217 "$value-$countvalue;$value-$countvalue;";
2220 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2221 return $finalresult;
2225 my ($rightresult, $leftresult)=@_;
2226 return $rightresult.$leftresult;
2230 my ($leftresult, $rightresult)=@_;
2232 my @leftresult = split /;/, $leftresult;
2234 # my @rightresult = split /;/,$leftresult;
2236 foreach (@leftresult) {
2238 $value=$1 if $value=~m/(.*)-\d+$/;
2239 unless ($rightresult =~ "$value-") {
2240 $finalresult .= "$_;";
2243 return $finalresult;
2248 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2255 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2256 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2258 # order title asc by default
2259 # $ordering = '1=36 <i' unless $ordering;
2260 $results_per_page = 20 unless $results_per_page;
2261 $offset = 0 unless $offset;
2262 my $dbh = C4::Context->dbh;
2265 # order by POPULARITY
2267 if ( $ordering =~ /popularity/ ) {
2271 # popularity is not in MARC record, it's builded from a specific query
2273 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2274 foreach ( split /;/, $biblionumbers ) {
2275 my ( $biblionumber, $title ) = split /,/, $_;
2276 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2277 $sth->execute($biblionumber);
2278 my $popularity = $sth->fetchrow || 0;
2280 # hint : the key is popularity.title because we can have
2281 # many results with the same popularity. In this case, sub-ordering is done by title
2282 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2283 # (un-frequent, I agree, but we won't forget anything that way ;-)
2284 $popularity{ sprintf( "%10d", $popularity ) . $title
2285 . $biblionumber } = $biblionumber;
2288 # sort the hash and return the same structure as GetRecords (Zebra querying)
2291 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2292 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2293 $result_hash->{'RECORDS'}[ $numbers++ ] =
2294 $result{ $popularity{$key} }->as_usmarc();
2297 else { # sort popularity ASC
2298 foreach my $key ( sort ( keys %popularity ) ) {
2299 $result_hash->{'RECORDS'}[ $numbers++ ] =
2300 $result{ $popularity{$key} }->as_usmarc();
2303 my $finalresult = ();
2304 $result_hash->{'hits'} = $numbers;
2305 $finalresult->{'biblioserver'} = $result_hash;
2306 return $finalresult;
2312 elsif ( $ordering =~ /author/ ) {
2314 foreach ( split /;/, $biblionumbers ) {
2315 my ( $biblionumber, $title ) = split /,/, $_;
2316 my $record = GetMarcBiblio($biblionumber);
2318 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2319 $author = $record->subfield( '200', 'f' );
2320 $author = $record->subfield( '700', 'a' ) unless $author;
2323 $author = $record->subfield( '100', 'a' );
2326 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2327 # and we don't want to get only 1 result for each of them !!!
2328 $result{ $author . $biblionumber } = $record;
2331 # sort the hash and return the same structure as GetRecords (Zebra querying)
2334 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2335 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2336 $result_hash->{'RECORDS'}[ $numbers++ ] =
2337 $result{$key}->as_usmarc();
2340 else { # sort by author ASC
2341 foreach my $key ( sort ( keys %result ) ) {
2342 $result_hash->{'RECORDS'}[ $numbers++ ] =
2343 $result{$key}->as_usmarc();
2346 my $finalresult = ();
2347 $result_hash->{'hits'} = $numbers;
2348 $finalresult->{'biblioserver'} = $result_hash;
2349 return $finalresult;
2352 # ORDER BY callnumber
2355 elsif ( $ordering =~ /callnumber/ ) {
2357 foreach ( split /;/, $biblionumbers ) {
2358 my ( $biblionumber, $title ) = split /,/, $_;
2359 my $record = GetMarcBiblio($biblionumber);
2361 my $frameworkcode = GetFrameworkCode($biblionumber);
2362 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2363 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2364 unless $callnumber_tag;
2365 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2366 $callnumber = $record->subfield( '200', 'f' );
2368 $callnumber = $record->subfield( '100', 'a' );
2371 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2372 # and we don't want to get only 1 result for each of them !!!
2373 $result{ $callnumber . $biblionumber } = $record;
2376 # sort the hash and return the same structure as GetRecords (Zebra querying)
2379 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2380 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2381 $result_hash->{'RECORDS'}[ $numbers++ ] =
2382 $result{$key}->as_usmarc();
2385 else { # sort by title ASC
2386 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2387 $result_hash->{'RECORDS'}[ $numbers++ ] =
2388 $result{$key}->as_usmarc();
2391 my $finalresult = ();
2392 $result_hash->{'hits'} = $numbers;
2393 $finalresult->{'biblioserver'} = $result_hash;
2394 return $finalresult;
2396 elsif ( $ordering =~ /pubdate/ ) { #pub year
2398 foreach ( split /;/, $biblionumbers ) {
2399 my ( $biblionumber, $title ) = split /,/, $_;
2400 my $record = GetMarcBiblio($biblionumber);
2401 my ( $publicationyear_tag, $publicationyear_subfield ) =
2402 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2403 my $publicationyear =
2404 $record->subfield( $publicationyear_tag,
2405 $publicationyear_subfield );
2407 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2408 # and we don't want to get only 1 result for each of them !!!
2409 $result{ $publicationyear . $biblionumber } = $record;
2412 # sort the hash and return the same structure as GetRecords (Zebra querying)
2415 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2416 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2417 $result_hash->{'RECORDS'}[ $numbers++ ] =
2418 $result{$key}->as_usmarc();
2421 else { # sort by pub year ASC
2422 foreach my $key ( sort ( keys %result ) ) {
2423 $result_hash->{'RECORDS'}[ $numbers++ ] =
2424 $result{$key}->as_usmarc();
2427 my $finalresult = ();
2428 $result_hash->{'hits'} = $numbers;
2429 $finalresult->{'biblioserver'} = $result_hash;
2430 return $finalresult;
2436 elsif ( $ordering =~ /title/ ) {
2438 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2440 foreach ( split /;/, $biblionumbers ) {
2441 my ( $biblionumber, $title ) = split /,/, $_;
2443 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2444 # and we don't want to get only 1 result for each of them !!!
2445 # hint & speed improvement : we can order without reading the record
2446 # so order, and read records only for the requested page !
2447 $result{ $title . $biblionumber } = $biblionumber;
2450 # sort the hash and return the same structure as GetRecords (Zebra querying)
2453 if ( $ordering eq 'title_az' ) { # sort by title desc
2454 foreach my $key ( sort ( keys %result ) ) {
2455 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2458 else { # sort by title ASC
2459 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2460 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2464 # limit the $results_per_page to result size if it's more
2465 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2467 # for the requested page, replace biblionumber by the complete record
2468 # speed improvement : avoid reading too much things
2470 my $counter = $offset ;
2471 $counter <= $offset + $results_per_page ;
2475 $result_hash->{'RECORDS'}[$counter] =
2476 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2478 my $finalresult = ();
2479 $result_hash->{'hits'} = $numbers;
2480 $finalresult->{'biblioserver'} = $result_hash;
2481 return $finalresult;
2488 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2491 foreach ( split /;/, $biblionumbers ) {
2492 my ( $biblionumber, $title ) = split /,/, $_;
2493 $title =~ /(.*)-(\d)/;
2498 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2499 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2500 # biblio N has ranking = 6
2501 $count_ranking{$biblionumber} += $ranking;
2504 # build the result by "inverting" the count_ranking hash
2505 # 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
2507 foreach ( keys %count_ranking ) {
2508 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2511 # sort the hash and return the same structure as GetRecords (Zebra querying)
2514 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2515 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2518 # limit the $results_per_page to result size if it's more
2519 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2521 # for the requested page, replace biblionumber by the complete record
2522 # speed improvement : avoid reading too much things
2524 my $counter = $offset ;
2525 $counter <= $offset + $results_per_page ;
2529 $result_hash->{'RECORDS'}[$counter] =
2530 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2531 if $result_hash->{'RECORDS'}[$counter];
2533 my $finalresult = ();
2534 $result_hash->{'hits'} = $numbers;
2535 $finalresult->{'biblioserver'} = $result_hash;
2536 return $finalresult;
2540 =head2 enabled_staff_search_views
2542 %hash = enabled_staff_search_views()
2544 This function returns a hash that contains three flags obtained from the system
2545 preferences, used to determine whether a particular staff search results view
2550 =item C<Output arg:>
2552 * $hash{can_view_MARC} is true only if the MARC view is enabled
2553 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2554 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2556 =item C<usage in the script:>
2560 $template->param ( C4::Search::enabled_staff_search_views );
2564 sub enabled_staff_search_views
2567 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2568 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2569 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2573 sub AddSearchHistory{
2574 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2575 my $dbh = C4::Context->dbh;
2577 # Add the request the user just made
2578 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2579 my $sth = $dbh->prepare($sql);
2580 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2581 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2584 sub GetSearchHistory{
2585 my ($borrowernumber,$session)=@_;
2586 my $dbh = C4::Context->dbh;
2588 # Add the request the user just made
2589 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2590 my $sth = $dbh->prepare($query);
2591 $sth->execute($borrowernumber, $session);
2592 return $sth->fetchall_hashref({});
2595 =head2 z3950_search_args
2597 $arrayref = z3950_search_args($matchpoints)
2599 This function returns an array reference that contains the search parameters to be
2600 passed to the Z39.50 search script (z3950_search.pl). The array elements
2601 are hash refs whose keys are name, value and encvalue, and whose values are the
2602 name of a search parameter, the value of that search parameter and the URL encoded
2603 value of that parameter.
2605 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2607 The search parameter values are obtained from the bibliographic record whose
2608 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2610 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2611 a general purpose search argument. In this case, the returned array contains only
2612 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2614 If a search parameter value is undefined or empty, it is not included in the returned
2617 The returned array reference may be passed directly to the template parameters.
2621 =item C<Output arg:>
2623 * $array containing hash refs as described above
2625 =item C<usage in the script:>
2629 $data = Biblio::GetBiblioData($bibno);
2630 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2634 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2638 sub z3950_search_args {
2640 $bibrec = { title => $bibrec } if !ref $bibrec;
2642 for my $field (qw/ lccn isbn issn title author dewey subject /)
2644 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2645 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2650 =head2 BiblioAddAuthorities
2652 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2654 this function finds the authorities linked to the biblio
2655 * search in the authority DB for the same authid (in $9 of the biblio)
2656 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2657 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2658 OR adds a new authority record
2664 * $record is the MARC record in question (marc blob)
2665 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2667 =item C<Output arg:>
2669 * $countlinked is the number of authorities records that are linked to this authority
2673 * 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)
2680 sub BiblioAddAuthorities{
2681 my ( $record, $frameworkcode ) = @_;
2682 my $dbh=C4::Context->dbh;
2683 my $query=$dbh->prepare(qq|
2684 SELECT authtypecode,tagfield
2685 FROM marc_subfield_structure
2686 WHERE frameworkcode=?
2687 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2688 # SELECT authtypecode,tagfield
2689 # FROM marc_subfield_structure
2690 # WHERE frameworkcode=?
2691 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2692 $query->execute($frameworkcode);
2693 my ($countcreated,$countlinked);
2694 while (my $data=$query->fetchrow_hashref){
2695 foreach my $field ($record->field($data->{tagfield})){
2696 next if ($field->subfield('3')||$field->subfield('9'));
2697 # No authorities id in the tag.
2698 # Search if there is any authorities to link to.
2699 my $query='at='.$data->{authtypecode}.' ';
2700 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2701 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2702 # there is only 1 result
2704 warn "BIBLIOADDSAUTHORITIES: $error";
2707 if ( @{$results} == 1 ) {
2708 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2709 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2711 } elsif ( @{$results} > 1 ) {
2712 #More than One result
2713 #This can comes out of a lack of a subfield.
2714 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2715 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2718 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2719 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2720 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2721 my $authtypedata=C4::AuthoritiesMarc::GetAuthType($data->{authtypecode});
2722 next unless $authtypedata;
2723 my $marcrecordauth=MARC::Record->new();
2724 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2725 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2726 $marcrecordauth->insert_fields_ordered($authfield);
2728 # bug 2317: ensure new authority knows it's using UTF-8; currently
2729 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2730 # automatically for UNIMARC (by not transcoding)
2731 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2732 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2733 # of change to a core API just before the 3.0 release.
2734 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2735 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2738 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2740 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2742 $field->add_subfields('9'=>$authid);
2746 return ($countlinked,$countcreated);
2749 =head2 GetDistinctValues($field);
2751 C<$field> is a reference to the fields array
2755 sub GetDistinctValues {
2756 my ($fieldname,$string)=@_;
2757 # returns a reference to a hash of references to branches...
2758 if ($fieldname=~/\./){
2759 my ($table,$column)=split /\./, $fieldname;
2760 my $dbh = C4::Context->dbh;
2761 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2762 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 ");
2764 my $elements=$sth->fetchall_arrayref({});
2769 my @servers=qw<biblioserver authorityserver>;
2770 my (@zconns,@results);
2771 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2772 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2775 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2778 # The big moment: asynchronously retrieve results from all servers
2780 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2781 my $ev = $zconns[ $i - 1 ]->last_event();
2782 if ( $ev == ZOOM::Event::ZEND ) {
2783 next unless $results[ $i - 1 ];
2784 my $size = $results[ $i - 1 ]->size();
2786 for (my $j=0;$j<$size;$j++){
2788 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2789 push @elements, \%hashscan;
2799 END { } # module clean-up code here (global destructor)
2806 Koha Development Team <http://koha-community.org/>