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
40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
42 # set the version for version checking
44 $VERSION = 3.07.00.049;
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
50 C4::Search - Functions for searching the Koha catalog.
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
58 This module provides searching functions for Koha's bibliographic databases
74 &enabled_staff_search_views
78 # make all your functions, whether exported or not;
82 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
84 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
90 my $dbh = C4::Context->dbh;
91 my $result = TransformMarcToKoha( $dbh, $record, '' );
96 my ( $biblionumber, $title );
98 # search duplicate on ISBN, easy and fast..
100 if ( $result->{isbn} ) {
101 $result->{isbn} =~ s/\(.*$//;
102 $result->{isbn} =~ s/\s+$//;
103 $query = "isbn=$result->{isbn}";
106 $result->{title} =~ s /\\//g;
107 $result->{title} =~ s /\"//g;
108 $result->{title} =~ s /\(//g;
109 $result->{title} =~ s /\)//g;
111 # FIXME: instead of removing operators, could just do
112 # quotes around the value
113 $result->{title} =~ s/(and|or|not)//g;
114 $query = "ti,ext=$result->{title}";
115 $query .= " and itemtype=$result->{itemtype}"
116 if ( $result->{itemtype} );
117 if ( $result->{author} ) {
118 $result->{author} =~ s /\\//g;
119 $result->{author} =~ s /\"//g;
120 $result->{author} =~ s /\(//g;
121 $result->{author} =~ s /\)//g;
123 # remove valid operators
124 $result->{author} =~ s/(and|or|not)//g;
125 $query .= " and au,ext=$result->{author}";
129 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
131 if (!defined $error) {
132 foreach my $possible_duplicate_record (@{$searchresults}) {
134 MARC::Record->new_from_usmarc($possible_duplicate_record);
135 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
137 # FIXME :: why 2 $biblionumber ?
139 push @results, $result->{'biblionumber'};
140 push @results, $result->{'title'};
149 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
151 This function provides a simple search API on the bibliographic catalog
157 * $query can be a simple keyword or a complete CCL query
158 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
159 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
160 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
165 Returns an array consisting of three elements
166 * $error is undefined unless an error is detected
167 * $results is a reference to an array of records.
168 * $total_hits is the number of hits that would have been returned with no limit
170 If an error is returned the two other return elements are undefined. If error itself is undefined
171 the other two elements are always defined
173 =item C<usage in the script:>
177 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
179 if (defined $error) {
180 $template->param(query_error => $error);
181 warn "error: ".$error;
182 output_html_with_http_headers $input, $cookie, $template->output;
186 my $hits = @{$marcresults};
189 for my $r ( @{$marcresults} ) {
190 my $marcrecord = MARC::File::USMARC::decode($r);
191 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
193 #build the iarray of hashs for the template.
195 title => $biblio->{'title'},
196 subtitle => $biblio->{'subtitle'},
197 biblionumber => $biblio->{'biblionumber'},
198 author => $biblio->{'author'},
199 publishercode => $biblio->{'publishercode'},
200 publicationyear => $biblio->{'publicationyear'},
205 $template->param(result=>\@results);
210 my ( $query, $offset, $max_results, $servers ) = @_;
212 if ( C4::Context->preference('NoZebra') ) {
213 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
216 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
217 return ( undef, $search_result, scalar($result->{hits}) );
220 return ( 'No query entered', undef, undef ) unless $query;
221 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
222 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
229 # Initialize & Search Zebra
230 for ( my $i = 0 ; $i < @servers ; $i++ ) {
232 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
233 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
234 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
238 $zconns[$i]->errmsg() . " ("
239 . $zconns[$i]->errcode() . ") "
240 . $zconns[$i]->addinfo() . " "
241 . $zconns[$i]->diagset();
243 return ( $error, undef, undef ) if $zconns[$i]->errcode();
247 # caught a ZOOM::Exception
251 . $@->addinfo() . " "
253 warn $error." for query: $query";
254 return ( $error, undef, undef );
257 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
258 my $event = $zconns[ $i - 1 ]->last_event();
259 if ( $event == ZOOM::Event::ZEND ) {
261 my $first_record = defined( $offset ) ? $offset+1 : 1;
262 my $hits = $tmpresults[ $i - 1 ]->size();
263 $total_hits += $hits;
264 my $last_record = $hits;
265 if ( defined $max_results && $offset + $max_results < $hits ) {
266 $last_record = $offset + $max_results;
269 for my $j ( $first_record..$last_record ) {
270 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
271 push @{$results}, $record;
276 foreach my $result (@tmpresults) {
279 foreach my $zoom_query (@zoom_queries) {
280 $zoom_query->destroy();
283 return ( undef, $results, $total_hits );
289 ( undef, $results_hashref, \@facets_loop ) = getRecords (
291 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
292 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
296 The all singing, all dancing, multi-server, asynchronous, scanning,
297 searching, record nabbing, facet-building
299 See verbse embedded documentation.
305 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
306 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
310 my @servers = @$servers_ref;
311 my @sort_by = @$sort_by_ref;
313 # Initialize variables for the ZOOM connection and results object
317 my $results_hashref = ();
319 # Initialize variables for the faceted results objects
320 my $facets_counter = ();
321 my $facets_info = ();
322 my $facets = getFacets();
323 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
325 my @facets_loop; # stores the ref to array of hashes for template facets loop
327 ### LOOP THROUGH THE SERVERS
328 for ( my $i = 0 ; $i < @servers ; $i++ ) {
329 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
331 # perform the search, create the results objects
332 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
333 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
335 #$query_to_use = $simple_query if $scan;
336 warn $simple_query if ( $scan and $DEBUG );
338 # Check if we've got a query_type defined, if so, use it
341 if ($query_type =~ /^ccl/) {
342 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
343 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
344 } elsif ($query_type =~ /^cql/) {
345 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
346 } elsif ($query_type =~ /^pqf/) {
347 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
349 warn "Unknown query_type '$query_type'. Results undetermined.";
352 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
354 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
358 warn "WARNING: query problem with $query_to_use " . $@;
361 # Concatenate the sort_by limits and pass them to the results object
362 # Note: sort will override rank
364 foreach my $sort (@sort_by) {
365 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
366 $sort_by .= "1=1003 <i ";
368 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
369 $sort_by .= "1=1003 >i ";
371 elsif ( $sort eq "popularity_asc" ) {
372 $sort_by .= "1=9003 <i ";
374 elsif ( $sort eq "popularity_dsc" ) {
375 $sort_by .= "1=9003 >i ";
377 elsif ( $sort eq "call_number_asc" ) {
378 $sort_by .= "1=8007 <i ";
380 elsif ( $sort eq "call_number_dsc" ) {
381 $sort_by .= "1=8007 >i ";
383 elsif ( $sort eq "pubdate_asc" ) {
384 $sort_by .= "1=31 <i ";
386 elsif ( $sort eq "pubdate_dsc" ) {
387 $sort_by .= "1=31 >i ";
389 elsif ( $sort eq "acqdate_asc" ) {
390 $sort_by .= "1=32 <i ";
392 elsif ( $sort eq "acqdate_dsc" ) {
393 $sort_by .= "1=32 >i ";
395 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
396 $sort_by .= "1=4 <i ";
398 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
399 $sort_by .= "1=4 >i ";
402 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
405 if ($sort_by && !$scan) {
406 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
407 warn "WARNING sort $sort_by failed";
410 } # finished looping through servers
412 # The big moment: asynchronously retrieve results from all servers
413 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
414 my $ev = $zconns[ $i - 1 ]->last_event();
415 if ( $ev == ZOOM::Event::ZEND ) {
416 next unless $results[ $i - 1 ];
417 my $size = $results[ $i - 1 ]->size();
421 # loop through the results
422 $results_hash->{'hits'} = $size;
424 if ( $offset + $results_per_page <= $size ) {
425 $times = $offset + $results_per_page;
430 for ( my $j = $offset ; $j < $times ; $j++ ) {
434 ## Check if it's an index scan
436 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
438 # here we create a minimal MARC record and hand it off to the
439 # template just like a normal result ... perhaps not ideal, but
441 my $tmprecord = MARC::Record->new();
442 $tmprecord->encoding('UTF-8');
446 # the minimal record in author/title (depending on MARC flavour)
447 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
448 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
449 $tmprecord->append_fields($tmptitle);
451 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
452 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
453 $tmprecord->append_fields($tmptitle);
454 $tmprecord->append_fields($tmpauthor);
456 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
461 $record = $results[ $i - 1 ]->record($j)->raw();
463 # warn "RECORD $j:".$record;
464 $results_hash->{'RECORDS'}[$j] = $record;
468 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
470 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
471 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
473 my $jmax = $size>$facets_maxrecs? $facets_maxrecs: $size;
474 for my $facet ( @$facets ) {
475 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
476 my $render_record = $results[ $i - 1 ]->record($j)->render();
478 foreach my $tag ( @{$facet->{tags}} ) {
480 my $tag_num = substr($tag, 0, 3);
481 my $letters = substr($tag, 3);
482 my $field_pattern = '\n' . $tag_num . ' ([^z][^\n]+)';
483 $field_pattern = '\n' . $tag_num . ' ([^\n]+)' if (int($tag_num) < 10);
484 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
485 foreach my $field_token (@field_tokens) {
486 my @subf = ( $field_token =~ /\$([a-zA-Z0-9]) ([^\$]+)/g );
488 for (my $i = 0; $i < @subf; $i += 2) {
489 if ( $letters =~ $subf[$i] ) {
490 my $value = $subf[$i+1];
493 push @values, $value;
496 my $data = join($facet->{sep}, @values);
497 unless ( $data ~~ @used_datas ) {
498 $facets_counter->{ $facet->{idx} }->{$data}++;
499 push @used_datas, $data;
504 $facets_info->{ $facet->{idx} }->{label_value} = $facet->{label};
505 $facets_info->{ $facet->{idx} }->{expanded} = $facet->{expanded};
510 # warn "connection ", $i-1, ": $size hits";
511 # warn $results[$i-1]->record(0)->render() if $size > 0;
514 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
516 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
517 keys %$facets_counter )
520 my $number_of_facets;
521 my @this_facets_array;
524 $facets_counter->{$link_value}->{$b}
525 <=> $facets_counter->{$link_value}->{$a}
526 } keys %{ $facets_counter->{$link_value} }
530 if ( ( $number_of_facets < 6 )
531 || ( $expanded_facet eq $link_value )
532 || ( $facets_info->{$link_value}->{'expanded'} ) )
535 # Sanitize the link value ), ( will cause errors with CCL,
536 my $facet_link_value = $one_facet;
537 $facet_link_value =~ s/(\(|\))/ /g;
539 # fix the length that will display in the label,
540 my $facet_label_value = $one_facet;
541 my $facet_max_length =
542 C4::Context->preference('FacetLabelTruncationLength') || 20;
544 substr( $one_facet, 0, $facet_max_length ) . "..."
545 if length($facet_label_value) > $facet_max_length;
547 # if it's a branch, label by the name, not the code,
548 if ( $link_value =~ /branch/ ) {
549 if (defined $branches
550 && ref($branches) eq "HASH"
551 && defined $branches->{$one_facet}
552 && ref ($branches->{$one_facet}) eq "HASH")
555 $branches->{$one_facet}->{'branchname'};
558 $facet_label_value = "*";
561 # if it's a itemtype, label by the name, not the code,
562 if ( $link_value =~ /itype/ ) {
563 if (defined $itemtypes
564 && ref($itemtypes) eq "HASH"
565 && defined $itemtypes->{$one_facet}
566 && ref ($itemtypes->{$one_facet}) eq "HASH")
569 $itemtypes->{$one_facet}->{'description'};
573 # but we're down with the whole label being in the link's title.
574 push @this_facets_array, {
575 facet_count => $facets_counter->{$link_value}->{$one_facet},
576 facet_label_value => $facet_label_value,
577 facet_title_value => $one_facet,
578 facet_link_value => $facet_link_value,
579 type_link_value => $link_value,
584 # handle expanded option
585 unless ( $facets_info->{$link_value}->{'expanded'} ) {
587 if ( ( $number_of_facets > 6 )
588 && ( $expanded_facet ne $link_value ) );
591 type_link_value => $link_value,
592 type_id => $link_value . "_id",
593 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
594 facets => \@this_facets_array,
595 expandable => $expandable,
596 expand => $link_value,
597 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
602 return ( undef, $results_hashref, \@facets_loop );
607 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
608 $results_per_page, $offset, $expanded_facet, $branches,
612 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
614 $paz->search($simple_query);
615 sleep 1; # FIXME: WHY?
618 my $results_hashref = {};
619 my $stats = XMLin($paz->stat);
620 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
622 # for a grouped search result, the number of hits
623 # is the number of groups returned; 'bib_hits' will have
624 # the total number of bibs.
625 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
626 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
628 HIT: foreach my $hit (@{ $results->{'hit'} }) {
629 my $recid = $hit->{recid}->[0];
631 my $work_title = $hit->{'md-work-title'}->[0];
633 if (exists $hit->{'md-work-author'}) {
634 $work_author = $hit->{'md-work-author'}->[0];
636 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
638 my $result_group = {};
639 $result_group->{'group_label'} = $group_label;
640 $result_group->{'group_merge_key'} = $recid;
643 if (exists $hit->{count}) {
644 $count = $hit->{count}->[0];
646 $result_group->{'group_count'} = $count;
648 for (my $i = 0; $i < $count; $i++) {
649 # FIXME -- may need to worry about diacritics here
650 my $rec = $paz->record($recid, $i);
651 push @{ $result_group->{'RECORDS'} }, $rec;
654 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
657 # pass through facets
658 my $termlist_xml = $paz->termlist('author,subject');
659 my $terms = XMLin($termlist_xml, forcearray => 1);
660 my @facets_loop = ();
661 #die Dumper($results);
662 # foreach my $list (sort keys %{ $terms->{'list'} }) {
664 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
666 # facet_label_value => $facet->{'name'}->[0],
669 # push @facets_loop, ( {
670 # type_label => $list,
671 # facets => \@facets,
675 return ( undef, $results_hashref, \@facets_loop );
679 sub _remove_stopwords {
680 my ( $operand, $index ) = @_;
681 my @stopwords_removed;
683 # phrase and exact-qualified indexes shouldn't have stopwords removed
684 if ( $index !~ m/phr|ext/ ) {
686 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
687 # we use IsAlpha unicode definition, to deal correctly with diacritics.
688 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
689 # is a stopword, we'd get "çon" and wouldn't find anything...
691 foreach ( keys %{ C4::Context->stopwords } ) {
692 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
693 if ( my ($matched) = ($operand =~
694 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
696 $operand =~ s/\Q$matched\E/ /gi;
697 push @stopwords_removed, $_;
701 return ( $operand, \@stopwords_removed );
705 sub _detect_truncation {
706 my ( $operand, $index ) = @_;
707 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
710 my @wordlist = split( /\s/, $operand );
711 foreach my $word (@wordlist) {
712 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
713 push @rightlefttruncated, $word;
715 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
716 push @lefttruncated, $word;
718 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
719 push @righttruncated, $word;
721 elsif ( index( $word, "*" ) < 0 ) {
722 push @nontruncated, $word;
725 push @regexpr, $word;
729 \@nontruncated, \@righttruncated, \@lefttruncated,
730 \@rightlefttruncated, \@regexpr
735 sub _build_stemmed_operand {
736 my ($operand,$lang) = @_;
737 require Lingua::Stem::Snowball ;
738 my $stemmed_operand=q{};
740 # If operand contains a digit, it is almost certainly an identifier, and should
741 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
742 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
743 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
744 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
745 return $operand if $operand =~ /\d/;
747 # FIXME: the locale should be set based on the user's language and/or search choice
749 # Make sure we only use the first two letters from the language code
750 $lang = lc(substr($lang, 0, 2));
751 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
752 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
753 if ($lang eq 'nb' || $lang eq 'nn') {
756 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
757 encoding => "UTF-8" );
759 my @words = split( / /, $operand );
760 my @stems = $stemmer->stem(\@words);
761 for my $stem (@stems) {
762 $stemmed_operand .= "$stem";
763 $stemmed_operand .= "?"
764 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
765 $stemmed_operand .= " ";
767 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
768 return $stemmed_operand;
772 sub _build_weighted_query {
774 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
775 # pretty well but could work much better if we had a smarter query parser
776 my ( $operand, $stemmed_operand, $index ) = @_;
777 my $stemming = C4::Context->preference("QueryStemming") || 0;
778 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
779 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
781 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
783 # Keyword, or, no index specified
784 if ( ( $index eq 'kw' ) || ( !$index ) ) {
786 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
787 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
788 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
789 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
790 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
791 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
792 if $fuzzy_enabled; # add fuzzy, word list
793 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
794 if ( $stemming and $stemmed_operand )
795 ; # add stemming, right truncation
796 $weighted_query .= " or wrdl,r9=\"$operand\"";
798 # embedded sorting: 0 a-z; 1 z-a
799 # $weighted_query .= ") or (sort1,aut=1";
802 # Barcode searches should skip this process
803 elsif ( $index eq 'bc' ) {
804 $weighted_query .= "bc=\"$operand\"";
807 # Authority-number searches should skip this process
808 elsif ( $index eq 'an' ) {
809 $weighted_query .= "an=\"$operand\"";
812 # If the index already has more than one qualifier, wrap the operand
813 # in quotes and pass it back (assumption is that the user knows what they
814 # are doing and won't appreciate us mucking up their query
815 elsif ( $index =~ ',' ) {
816 $weighted_query .= " $index=\"$operand\"";
819 #TODO: build better cases based on specific search indexes
821 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
822 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
823 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
825 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
828 $weighted_query .= "))"; # close rank specification
829 return $weighted_query;
834 Return an array with available indexes.
856 'Author-personal-bibliography',
866 'Chronological-subdivision',
876 'Conference-name-heading',
877 'Conference-name-see',
878 'Conference-name-seealso',
883 'Corporate-name-heading',
884 'Corporate-name-see',
885 'Corporate-name-seealso',
887 'date-entered-on-file',
888 'Date-of-acquisition',
889 'Date-of-publication',
890 'Dewey-classification',
897 'Geographic-subdivision',
900 'Heading-use-main-or-added-entry',
901 'Heading-use-series-added-entry ',
902 'Heading-use-subject-added-entry',
920 'Local-classification',
923 'Match-heading-see-from',
931 'Name-geographic-heading',
932 'Name-geographic-see',
933 'Name-geographic-seealso',
941 'Personal-name-heading',
943 'Personal-name-seealso',
950 'Record-control-number',
961 'Subject-heading-thesaurus',
962 'Subject-name-personal',
963 'Subject-subdivision',
973 'Term-genre-form-heading',
974 'Term-genre-form-see',
975 'Term-genre-form-seealso',
982 'Title-uniform-heading',
984 'Title-uniform-seealso',
994 'classification-source',
996 'coded-location-qualifier',
1007 'Local-classification',
1010 'materials-specified',
1019 'replacementpricedate',
1034 =head2 _handle_exploding_index
1036 my $query = _handle_exploding_index($index, $term)
1038 Callback routine to generate the search for "exploding" indexes (i.e.
1039 those indexes which are turned into multiple or-connected searches based
1044 sub _handle_exploding_index {
1045 my ( $index, $term ) = @_;
1047 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1049 my $marcflavour = C4::Context->preference('marcflavour');
1051 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1052 my $wantedcodes = '';
1053 my @subqueries = ( "(su=\"$term\")");
1054 my ($error, $results, $total_hits) = SimpleSearch( "Heading,wrdl=$term", undef, undef, [ "authorityserver" ] );
1055 foreach my $auth (@$results) {
1056 my $record = MARC::Record->new_from_usmarc($auth);
1057 my @references = $record->field('5..');
1059 if ($index eq 'su-br') {
1061 } elsif ($index eq 'su-na') {
1063 } elsif ($index eq 'su-rl') {
1066 foreach my $reference (@references) {
1067 my $codes = $reference->subfield($codesubfield);
1068 push @subqueries, '(su="' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '")' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1072 return join(' or ', @subqueries);
1077 ( $operators, $operands, $indexes, $limits,
1078 $sort_by, $scan, $lang ) =
1079 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1081 Shim function to ease the transition from buildQuery to a new QueryParser.
1082 This function is called at the beginning of buildQuery, and modifies
1083 buildQuery's input. If it can handle the input, it returns a query that
1084 buildQuery will not try to parse.
1088 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1090 my @operators = $operators ? @$operators : ();
1091 my @indexes = $indexes ? @$indexes : ();
1092 my @operands = $operands ? @$operands : ();
1093 my @limits = $limits ? @$limits : ();
1094 my @sort_by = $sort_by ? @$sort_by : ();
1096 my $query = $operands[0];
1100 # TODO: once we are using QueryParser, all this special case code for
1101 # exploded search indexes will be replaced by a callback to
1102 # _handle_exploding_index
1103 if ( $query =~ m/^(.*)\b(su-br|su-na|su-rl)[:=](\w.*)$/ ) {
1109 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1110 if ($operands[$i] && $indexes[$i] =~ m/(su-br|su-na|su-rl)/) {
1111 $index = $indexes[$i];
1112 $term = $operands[$i];
1113 } elsif ($operands[$i]) {
1114 $query .= $operators[$i] eq 'or' ? ' or ' : ' and ' if ($query);
1115 $query .= "($indexes[$i]:$operands[$i])";
1121 my $queryPart = _handle_exploding_index($index, $term);
1123 $query .= "($queryPart)";
1126 $operands[0] = "ccl=$query";
1129 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang);
1135 $simple_query, $query_cgi,
1136 $query_desc, $limit,
1137 $limit_cgi, $limit_desc,
1138 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1140 Build queries and limits in CCL, CGI, Human,
1141 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1143 See verbose embedded documentation.
1149 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1151 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1153 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1156 my @operators = $operators ? @$operators : ();
1157 my @indexes = $indexes ? @$indexes : ();
1158 my @operands = $operands ? @$operands : ();
1159 my @limits = $limits ? @$limits : ();
1160 my @sort_by = $sort_by ? @$sort_by : ();
1162 my $stemming = C4::Context->preference("QueryStemming") || 0;
1163 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1164 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1165 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1166 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1168 # no stemming/weight/fuzzy in NoZebra
1169 if ( C4::Context->preference("NoZebra") ) {
1173 $auto_truncation = 0;
1176 my $query = $operands[0];
1177 my $simple_query = $operands[0];
1179 # initialize the variables we're passing back
1188 my $stopwords_removed; # flag to determine if stopwords have been removed
1191 my $cclindexes = getIndexes();
1192 if ( $query !~ /\s*ccl=/ ) {
1193 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1195 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1197 $query = "ccl=$query" if $cclq;
1200 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1202 if ( $query =~ /^ccl=/ ) {
1204 # This is needed otherwise ccl= and &limit won't work together, and
1205 # this happens when selecting a subject on the opac-detail page
1206 @limits = grep {!/^$/} @limits;
1208 $q .= ' and '.join(' and ', @limits);
1210 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1212 if ( $query =~ /^cql=/ ) {
1213 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1215 if ( $query =~ /^pqf=/ ) {
1216 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1219 # pass nested queries directly
1220 # FIXME: need better handling of some of these variables in this case
1221 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1222 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1223 # if ( $query =~ /(\(|\))/ ) {
1225 # undef, $query, $simple_query, $query_cgi,
1226 # $query, $limit, $limit_cgi, $limit_desc,
1227 # $stopwords_removed, 'ccl'
1231 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1232 # query operands and indexes and add stemming, truncation, field weighting, etc.
1233 # Once we do so, we'll end up with a value in $query, just like if we had an
1234 # incoming $query from the user
1237 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1238 my $previous_operand
1239 ; # a flag used to keep track if there was a previous query
1240 # if there was, we can apply the current operator
1242 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1244 # COMBINE OPERANDS, INDEXES AND OPERATORS
1245 if ( $operands[$i] ) {
1246 $operands[$i]=~s/^\s+//;
1248 # A flag to determine whether or not to add the index to the query
1251 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1252 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1255 $remove_stopwords = 0;
1257 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1259 my $operand = $operands[$i];
1260 my $index = $indexes[$i];
1262 # Add index-specific attributes
1263 # Date of Publication
1264 if ( $index eq 'yr' ) {
1265 $index .= ",st-numeric";
1267 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1270 # Date of Acquisition
1271 elsif ( $index eq 'acqdate' ) {
1272 $index .= ",st-date-normalized";
1274 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1276 # ISBN,ISSN,Standard Number, don't need special treatment
1277 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1279 $stemming, $auto_truncation,
1280 $weight_fields, $fuzzy_enabled,
1282 ) = ( 0, 0, 0, 0, 0 );
1290 # Set default structure attribute (word list)
1291 my $struct_attr = q{};
1292 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1293 $struct_attr = ",wrdl";
1296 # Some helpful index variants
1297 my $index_plus = $index . $struct_attr . ':';
1298 my $index_plus_comma = $index . $struct_attr . ',';
1301 if ($remove_stopwords) {
1302 ( $operand, $stopwords_removed ) =
1303 _remove_stopwords( $operand, $index );
1304 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1305 warn "REMOVED STOPWORDS: @$stopwords_removed"
1306 if ( $stopwords_removed && $DEBUG );
1309 if ($auto_truncation){
1310 unless ( $index =~ /(st-|phr|ext)/ ) {
1311 #FIXME only valid with LTR scripts
1312 $operand=join(" ",map{
1313 (index($_,"*")>0?"$_":"$_*")
1314 }split (/\s+/,$operand));
1315 warn $operand if $DEBUG;
1320 my $truncated_operand;
1321 my( $nontruncated, $righttruncated, $lefttruncated,
1322 $rightlefttruncated, $regexpr
1323 ) = _detect_truncation( $operand, $index );
1325 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1330 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1331 scalar(@$rightlefttruncated) > 0 )
1334 # Don't field weight or add the index to the query, we do it here
1336 undef $weight_fields;
1337 my $previous_truncation_operand;
1338 if (scalar @$nontruncated) {
1339 $truncated_operand .= "$index_plus @$nontruncated ";
1340 $previous_truncation_operand = 1;
1342 if (scalar @$righttruncated) {
1343 $truncated_operand .= "and " if $previous_truncation_operand;
1344 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1345 $previous_truncation_operand = 1;
1347 if (scalar @$lefttruncated) {
1348 $truncated_operand .= "and " if $previous_truncation_operand;
1349 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1350 $previous_truncation_operand = 1;
1352 if (scalar @$rightlefttruncated) {
1353 $truncated_operand .= "and " if $previous_truncation_operand;
1354 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1355 $previous_truncation_operand = 1;
1358 $operand = $truncated_operand if $truncated_operand;
1359 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1362 my $stemmed_operand;
1363 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1366 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1368 # Handle Field Weighting
1369 my $weighted_operand;
1370 if ($weight_fields) {
1371 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1372 $operand = $weighted_operand;
1376 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1378 # If there's a previous operand, we need to add an operator
1379 if ($previous_operand) {
1381 # User-specified operator
1382 if ( $operators[ $i - 1 ] ) {
1383 $query .= " $operators[$i-1] ";
1384 $query .= " $index_plus " unless $indexes_set;
1385 $query .= " $operand";
1386 $query_cgi .= "&op=$operators[$i-1]";
1387 $query_cgi .= "&idx=$index" if $index;
1388 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1390 " $operators[$i-1] $index_plus $operands[$i]";
1393 # Default operator is and
1396 $query .= "$index_plus " unless $indexes_set;
1397 $query .= "$operand";
1398 $query_cgi .= "&op=and&idx=$index" if $index;
1399 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1400 $query_desc .= " and $index_plus $operands[$i]";
1404 # There isn't a pervious operand, don't need an operator
1407 # Field-weighted queries already have indexes set
1408 $query .= " $index_plus " unless $indexes_set;
1410 $query_desc .= " $index_plus $operands[$i]";
1411 $query_cgi .= "&idx=$index" if $index;
1412 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1413 $previous_operand = 1;
1418 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1421 my %group_OR_limits;
1422 my $availability_limit;
1423 foreach my $this_limit (@limits) {
1424 next unless $this_limit;
1425 if ( $this_limit =~ /available/ ) {
1427 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1429 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1430 $availability_limit .=
1431 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1432 $limit_cgi .= "&limit=available";
1436 # group_OR_limits, prefixed by mc-
1437 # OR every member of the group
1438 elsif ( $this_limit =~ /mc/ ) {
1439 my ($k,$v) = split(/:/, $this_limit,2);
1440 if ( $k !~ /mc-i(tem)?type/ ) {
1441 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1442 $this_limit =~ tr/"//d;
1443 $this_limit = $k.":\"".$v."\"";
1446 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1447 $limit_desc .= " or " if $group_OR_limits{$k};
1448 $group_OR_limits{$k} .= "$this_limit";
1449 $limit_cgi .= "&limit=$this_limit";
1450 $limit_desc .= " $this_limit";
1453 # Regular old limits
1455 $limit .= " and " if $limit || $query;
1456 $limit .= "$this_limit";
1457 $limit_cgi .= "&limit=$this_limit";
1458 if ($this_limit =~ /^branch:(.+)/) {
1459 my $branchcode = $1;
1460 my $branchname = GetBranchName($branchcode);
1461 if (defined $branchname) {
1462 $limit_desc .= " branch:$branchname";
1464 $limit_desc .= " $this_limit";
1467 $limit_desc .= " $this_limit";
1471 foreach my $k (keys (%group_OR_limits)) {
1472 $limit .= " and " if ( $query || $limit );
1473 $limit .= "($group_OR_limits{$k})";
1475 if ($availability_limit) {
1476 $limit .= " and " if ( $query || $limit );
1477 $limit .= "($availability_limit)";
1480 # Normalize the query and limit strings
1481 # This is flawed , means we can't search anything with : in it
1482 # if user wants to do ccl or cql, start the query with that
1483 # $query =~ s/:/=/g;
1484 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1485 $query =~ s/(?<=(wrdl)):/=/g;
1486 $query =~ s/(?<=(trn|phr)):/=/g;
1488 for ( $query, $query_desc, $limit, $limit_desc ) {
1489 s/ +/ /g; # remove extra spaces
1490 s/^ //g; # remove any beginning spaces
1491 s/ $//g; # remove any ending spaces
1492 s/==/=/g; # remove double == from query
1494 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1496 for ($query_cgi,$simple_query) {
1499 # append the limit to the query
1500 $query .= " " . $limit;
1504 warn "QUERY:" . $query;
1505 warn "QUERY CGI:" . $query_cgi;
1506 warn "QUERY DESC:" . $query_desc;
1507 warn "LIMIT:" . $limit;
1508 warn "LIMIT CGI:" . $limit_cgi;
1509 warn "LIMIT DESC:" . $limit_desc;
1510 warn "---------\nLeave buildQuery\n---------";
1513 undef, $query, $simple_query, $query_cgi,
1514 $query_desc, $limit, $limit_cgi, $limit_desc,
1515 $stopwords_removed, $query_type
1519 =head2 searchResults
1521 my @search_results = searchResults($search_context, $searchdesc, $hits,
1522 $results_per_page, $offset, $scan,
1525 Format results in a form suitable for passing to the template
1529 # IMO this subroutine is pretty messy still -- it's responsible for
1530 # building the HTML output for the template
1532 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1533 my $dbh = C4::Context->dbh;
1538 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1539 my ($is_opac, $hidelostitems);
1540 if ($search_context eq 'opac') {
1541 $hidelostitems = C4::Context->preference('hidelostitems');
1545 #Build branchnames hash
1547 #get branch information.....
1549 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1551 while ( my $bdata = $bsth->fetchrow_hashref ) {
1552 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1554 # FIXME - We build an authorised values hash here, using the default framework
1555 # though it is possible to have different authvals for different fws.
1557 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1559 # get notforloan authorised value list (see $shelflocations FIXME)
1560 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1562 #Build itemtype hash
1563 #find itemtype & itemtype image
1567 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1570 while ( my $bdata = $bsth->fetchrow_hashref ) {
1571 foreach (qw(description imageurl summary notforloan)) {
1572 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1576 #search item field code
1577 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1579 ## find column names of items related to MARC
1580 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1582 my %subfieldstosearch;
1583 while ( ( my $column ) = $sth2->fetchrow ) {
1584 my ( $tagfield, $tagsubfield ) =
1585 &GetMarcFromKohaField( "items." . $column, "" );
1586 $subfieldstosearch{$column} = $tagsubfield;
1589 # handle which records to actually retrieve
1591 if ( $hits && $offset + $results_per_page <= $hits ) {
1592 $times = $offset + $results_per_page;
1595 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1598 my $marcflavour = C4::Context->preference("marcflavour");
1599 # We get the biblionumber position in MARC
1600 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1602 # loop through all of the records we've retrieved
1603 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1604 my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1608 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1609 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1610 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1611 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1612 $oldbiblio->{result_number} = $i + 1;
1614 # add imageurl to itemtype if there is one
1615 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1617 $oldbiblio->{'authorised_value_images'} = ($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 ) ) : [];
1618 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1619 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1620 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1621 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1622 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1624 # edition information, if any
1625 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1626 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1627 # Build summary if there is one (the summary is defined in the itemtypes table)
1628 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1629 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1630 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1631 my @fields = $marcrecord->fields();
1634 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1636 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1637 $tag =~ /(.{3})(.)/;
1638 if($marcrecord->field($1)){
1639 my @abc = $marcrecord->field($1)->subfield($2);
1640 $tags->{$tag} = $#abc + 1 ;
1644 # We catch how many times to repeat this line
1646 foreach my $tag (keys(%$tags)){
1647 $max = $tags->{$tag} if($tags->{$tag} > $max);
1650 # we replace, and repeat each line
1651 for (my $i = 0 ; $i < $max ; $i++){
1652 my $newline = $line;
1654 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1655 $tag =~ /(.{3})(.)/;
1657 if($marcrecord->field($1)){
1658 my @repl = $marcrecord->field($1)->subfield($2);
1659 my $subfieldvalue = $repl[$i];
1661 if (! utf8::is_utf8($subfieldvalue)) {
1662 utf8::decode($subfieldvalue);
1665 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1668 $newsummary .= "$newline\n";
1672 $newsummary =~ s/\[(.*?)]//g;
1673 $newsummary =~ s/\n/<br\/>/g;
1674 $oldbiblio->{summary} = $newsummary;
1677 # Pull out the items fields
1678 my @fields = $marcrecord->field($itemtag);
1679 my $marcflavor = C4::Context->preference("marcflavour");
1680 # adding linked items that belong to host records
1681 my $analyticsfield = '773';
1682 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1683 $analyticsfield = '773';
1684 } elsif ($marcflavor eq 'UNIMARC') {
1685 $analyticsfield = '461';
1687 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1688 my $hostbiblionumber = $hostfield->subfield("0");
1689 my $linkeditemnumber = $hostfield->subfield("9");
1690 if(!$hostbiblionumber eq undef){
1691 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1692 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1693 if(!$hostbiblio eq undef){
1694 my @hostitems = $hostbiblio->field($itemfield);
1695 foreach my $hostitem (@hostitems){
1696 if ($hostitem->subfield("9") eq $linkeditemnumber){
1697 my $linkeditem =$hostitem;
1698 # append linked items if they exist
1699 if (!$linkeditem eq undef){
1700 push (@fields, $linkeditem);}
1707 # Setting item statuses for display
1708 my @available_items_loop;
1709 my @onloan_items_loop;
1710 my @other_items_loop;
1712 my $available_items;
1716 my $ordered_count = 0;
1717 my $available_count = 0;
1718 my $onloan_count = 0;
1719 my $longoverdue_count = 0;
1720 my $other_count = 0;
1721 my $wthdrawn_count = 0;
1722 my $itemlost_count = 0;
1723 my $hideatopac_count = 0;
1724 my $itembinding_count = 0;
1725 my $itemdamaged_count = 0;
1726 my $item_in_transit_count = 0;
1727 my $can_place_holds = 0;
1728 my $item_onhold_count = 0;
1729 my $items_count = scalar(@fields);
1730 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1731 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1732 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1734 # loop through every item
1735 foreach my $field (@fields) {
1738 # populate the items hash
1739 foreach my $code ( keys %subfieldstosearch ) {
1740 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1742 $item->{description} = $itemtypes{ $item->{itype} }{description};
1746 # hidden because lost
1747 if ($hidelostitems && $item->{itemlost}) {
1748 $hideatopac_count++;
1751 # hidden based on OpacHiddenItems syspref
1752 my @hi = C4::Items::GetHiddenItemnumbers($item);
1754 push @hiddenitems, @hi;
1755 $hideatopac_count++;
1760 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1761 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1763 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1764 if ($item->{$hbranch}) {
1765 $item->{'branchname'} = $branches{$item->{$hbranch}};
1767 elsif ($item->{$otherbranch}) { # Last resort
1768 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1771 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1772 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1773 my $userenv = C4::Context->userenv;
1774 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1776 my $key = $prefix . $item->{onloan} . $item->{barcode};
1777 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1778 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1779 $onloan_items->{$key}->{branchname} = $item->{branchname};
1780 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1781 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1782 $onloan_items->{$key}->{description} = $item->{description};
1783 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1784 # if something's checked out and lost, mark it as 'long overdue'
1785 if ( $item->{itemlost} ) {
1786 $onloan_items->{$prefix}->{longoverdue}++;
1787 $longoverdue_count++;
1788 } else { # can place holds as long as item isn't lost
1789 $can_place_holds = 1;
1793 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1797 if ( $item->{notforloan} == -1 ) {
1801 # is item in transit?
1802 my $transfertwhen = '';
1803 my ($transfertfrom, $transfertto);
1805 # is item on the reserve shelf?
1806 my $reservestatus = '';
1809 unless ($item->{wthdrawn}
1810 || $item->{itemlost}
1812 || $item->{notforloan}
1813 || $items_count > 20) {
1815 # A couple heuristics to limit how many times
1816 # we query the database for item transfer information, sacrificing
1817 # accuracy in some cases for speed;
1819 # 1. don't query if item has one of the other statuses
1820 # 2. don't check transit status if the bib has
1821 # more than 20 items
1823 # FIXME: to avoid having the query the database like this, and to make
1824 # the in transit status count as unavailable for search limiting,
1825 # should map transit status to record indexed in Zebra.
1827 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1828 ($reservestatus, $reserveitem, undef) = C4::Reserves::CheckReserves($item->{itemnumber});
1831 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1832 if ( $item->{wthdrawn}
1833 || $item->{itemlost}
1835 || $item->{notforloan} > 0
1836 || $reservestatus eq 'Waiting'
1837 || ($transfertwhen ne ''))
1839 $wthdrawn_count++ if $item->{wthdrawn};
1840 $itemlost_count++ if $item->{itemlost};
1841 $itemdamaged_count++ if $item->{damaged};
1842 $item_in_transit_count++ if $transfertwhen ne '';
1843 $item_onhold_count++ if $reservestatus eq 'Waiting';
1844 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1846 # can place hold on item ?
1847 if ((!$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems'))
1848 && !$item->{itemlost}
1849 && !$item->{withdrawn}
1851 $can_place_holds = 1;
1856 my $key = $prefix . $item->{status};
1857 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1858 $other_items->{$key}->{$_} = $item->{$_};
1860 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1861 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1862 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1863 $other_items->{$key}->{count}++ if $item->{$hbranch};
1864 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1865 $other_items->{$key}->{description} = $item->{description};
1866 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1870 $can_place_holds = 1;
1872 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1873 foreach (qw(branchname itemcallnumber description)) {
1874 $available_items->{$prefix}->{$_} = $item->{$_};
1876 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1877 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1880 } # notforloan, item level and biblioitem level
1882 # if all items are hidden, do not show the record
1883 if ($items_count > 0 && $hideatopac_count == $items_count) {
1887 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1888 for my $key ( sort keys %$onloan_items ) {
1889 (++$onloanitemscount > $maxitems) and last;
1890 push @onloan_items_loop, $onloan_items->{$key};
1892 for my $key ( sort keys %$other_items ) {
1893 (++$otheritemscount > $maxitems) and last;
1894 push @other_items_loop, $other_items->{$key};
1896 for my $key ( sort keys %$available_items ) {
1897 (++$availableitemscount > $maxitems) and last;
1898 push @available_items_loop, $available_items->{$key}
1901 # XSLT processing of some stuff
1903 SetUTF8Flag($marcrecord);
1904 warn $marcrecord->as_formatted if $DEBUG;
1905 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1906 if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
1907 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
1908 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1911 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1912 if (!C4::Context->preference("item-level_itypes")) {
1913 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1914 $can_place_holds = 0;
1917 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1918 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1919 $oldbiblio->{items_count} = $items_count;
1920 $oldbiblio->{available_items_loop} = \@available_items_loop;
1921 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1922 $oldbiblio->{other_items_loop} = \@other_items_loop;
1923 $oldbiblio->{availablecount} = $available_count;
1924 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1925 $oldbiblio->{onloancount} = $onloan_count;
1926 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1927 $oldbiblio->{othercount} = $other_count;
1928 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1929 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1930 $oldbiblio->{itemlostcount} = $itemlost_count;
1931 $oldbiblio->{damagedcount} = $itemdamaged_count;
1932 $oldbiblio->{intransitcount} = $item_in_transit_count;
1933 $oldbiblio->{onholdcount} = $item_onhold_count;
1934 $oldbiblio->{orderedcount} = $ordered_count;
1936 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1937 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1938 my $subfields = substr $fieldspec, 3;
1939 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1940 my @alternateholdingsinfo = ();
1941 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1942 my $alternateholdingscount = 0;
1944 for my $field (@holdingsfields) {
1945 my %holding = ( holding => '' );
1946 my $havesubfield = 0;
1947 for my $subfield ($field->subfields()) {
1948 if ((index $subfields, $$subfield[0]) >= 0) {
1949 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1950 $holding{'holding'} .= $$subfield[1];
1954 if ($havesubfield) {
1955 push(@alternateholdingsinfo, \%holding);
1956 $alternateholdingscount++;
1960 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1961 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1964 push( @newresults, $oldbiblio );
1970 =head2 SearchAcquisitions
1971 Search for acquisitions
1974 sub SearchAcquisitions{
1975 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1977 my $dbh=C4::Context->dbh;
1978 # Variable initialization
1982 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1983 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1984 WHERE dateaccessioned BETWEEN ? AND ?
1987 my (@params,@loopcriteria);
1989 push @params, $datebegin->output("iso");
1990 push @params, $dateend->output("iso");
1992 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1993 if(C4::Context->preference("item-level_itypes")){
1994 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1996 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1998 push @params, @$itemtypes;
2001 if ($criteria =~/itemtype/){
2002 if(C4::Context->preference("item-level_itypes")){
2003 $str .= "AND items.itype=? ";
2005 $str .= "AND biblioitems.itemtype=? ";
2008 if(scalar(@$itemtypes) == 0){
2009 my $itypes = GetItemTypes();
2010 for my $key (keys %$itypes){
2011 push @$itemtypes, $key;
2015 @loopcriteria= @$itemtypes;
2016 }elsif ($criteria=~/itemcallnumber/){
2017 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2018 OR items.itemcallnumber is NULL
2019 OR items.itemcallnumber = '')";
2021 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2023 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2024 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2027 if ($orderby =~ /date_desc/){
2028 $str.=" ORDER BY dateaccessioned DESC";
2030 $str.=" ORDER BY title";
2033 my $qdataacquisitions=$dbh->prepare($str);
2035 my @loopacquisitions;
2036 foreach my $value(@loopcriteria){
2037 push @params,$value;
2039 $cell{"title"}=$value;
2040 $cell{"titlecode"}=$value;
2042 eval{$qdataacquisitions->execute(@params);};
2044 if ($@){ warn "recentacquisitions Error :$@";}
2047 while (my $data=$qdataacquisitions->fetchrow_hashref){
2048 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2050 $cell{"loopdata"}=\@loopdata;
2052 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2055 $qdataacquisitions->finish;
2056 return \@loopacquisitions;
2058 #----------------------------------------------------------------------
2060 # Non-Zebra GetRecords#
2061 #----------------------------------------------------------------------
2065 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
2071 $query, $simple_query, $sort_by_ref, $servers_ref,
2072 $results_per_page, $offset, $expanded_facet, $branches,
2075 warn "query =$query" if $DEBUG;
2076 my $result = NZanalyse($query);
2077 warn "results =$result" if $DEBUG;
2079 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
2085 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
2086 the list is built from an inverted index in the nozebra SQL table
2087 note that title is here only for convenience : the sorting will be very fast when requested on title
2088 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
2093 my ( $string, $server ) = @_;
2094 # warn "---------" if $DEBUG;
2095 warn " NZanalyse" if $DEBUG;
2096 # warn "---------" if $DEBUG;
2098 # $server contains biblioserver or authorities, depending on what we search on.
2099 #warn "querying : $string on $server";
2100 $server = 'biblioserver' unless $server;
2102 # if we have a ", replace the content to discard temporarily any and/or/not inside
2104 if ( $string =~ /"/ ) {
2105 $string =~ s/"(.*?)"/__X__/;
2107 warn "commacontent : $commacontent" if $DEBUG;
2110 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
2111 # then, call again NZanalyse with $left and $right
2112 # (recursive until we find a leaf (=> something without and/or/not)
2113 # delete repeated operator... Would then go in infinite loop
2114 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
2117 #process parenthesis before.
2118 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
2121 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
2123 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2125 my $leftresult = NZanalyse( $left, $server );
2127 my $rightresult = NZanalyse( $right, $server );
2129 # OK, we have the results for right and left part of the query
2130 # depending of operand, intersect, union or exclude both lists
2131 # to get a result list
2132 if ( $operator eq ' and ' ) {
2133 return NZoperatorAND($leftresult,$rightresult);
2135 elsif ( $operator eq ' or ' ) {
2137 # just merge the 2 strings
2138 return $leftresult . $rightresult;
2140 elsif ( $operator eq ' not ' ) {
2141 return NZoperatorNOT($leftresult,$rightresult);
2145 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2149 warn "string :" . $string if $DEBUG;
2153 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2156 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2158 warn "no parenthesis. left : $left operator: $operator right: $right"
2161 # it's not a leaf, we have a and/or/not
2164 # reintroduce comma content if needed
2165 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2166 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2167 warn "node : $left / $operator / $right\n" if $DEBUG;
2168 my $leftresult = NZanalyse( $left, $server );
2169 my $rightresult = NZanalyse( $right, $server );
2170 warn " leftresult : $leftresult" if $DEBUG;
2171 warn " rightresult : $rightresult" if $DEBUG;
2172 # OK, we have the results for right and left part of the query
2173 # depending of operand, intersect, union or exclude both lists
2174 # to get a result list
2175 if ( $operator eq ' and ' ) {
2176 return NZoperatorAND($leftresult,$rightresult);
2178 elsif ( $operator eq ' or ' ) {
2180 # just merge the 2 strings
2181 return $leftresult . $rightresult;
2183 elsif ( $operator eq ' not ' ) {
2184 return NZoperatorNOT($leftresult,$rightresult);
2188 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2189 die "error : operand unknown : $operator for $string";
2192 # it's a leaf, do the real SQL query and return the result
2195 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2196 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2197 #remove trailing blank at the beginning
2199 warn "leaf:$string" if $DEBUG;
2201 # parse the string in in operator/operand/value again
2205 if ($string =~ /(.*)(>=|<=)(.*)/) {
2212 # warn "handling leaf... left:$left operator:$operator right:$right"
2214 unless ($operator) {
2215 if ($string =~ /(.*)(>|<|=)(.*)/) {
2220 "handling unless (operator)... left:$left operator:$operator right:$right"
2228 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2231 # automatic replace for short operators
2232 $left = 'title' if $left =~ '^ti$';
2233 $left = 'author' if $left =~ '^au$';
2234 $left = 'publisher' if $left =~ '^pb$';
2235 $left = 'subject' if $left =~ '^su$';
2236 $left = 'koha-Auth-Number' if $left =~ '^an$';
2237 $left = 'keyword' if $left =~ '^kw$';
2238 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2239 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2240 my $dbh = C4::Context->dbh;
2241 if ( $operator && $left ne 'keyword' ) {
2242 #do a specific search
2243 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2244 my $sth = $dbh->prepare(
2245 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2247 warn "$left / $operator / $right\n" if $DEBUG;
2249 # split each word, query the DB and build the biblionumbers result
2250 #sanitizing leftpart
2251 $left =~ s/^\s+|\s+$//;
2252 foreach ( split / /, $right ) {
2254 $_ =~ s/^\s+|\s+$//;
2256 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2257 $sth->execute( $server, $left, $_ )
2258 or warn "execute failed: $!";
2259 while ( my ( $line, $value ) = $sth->fetchrow ) {
2261 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2262 # otherwise, fill the result
2263 $biblionumbers .= $line
2264 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2265 warn "result : $value "
2266 . ( $right =~ /\d/ ) . "=="
2267 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2270 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2272 warn "NZAND" if $DEBUG;
2273 $results = NZoperatorAND($biblionumbers,$results);
2275 $results = $biblionumbers;
2280 #do a complete search (all indexes), if index='kw' do complete search too.
2281 my $sth = $dbh->prepare(
2282 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2285 # split each word, query the DB and build the biblionumbers result
2286 foreach ( split / /, $string ) {
2287 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2288 warn "search on all indexes on $_" if $DEBUG;
2291 $sth->execute( $server, $_ );
2292 while ( my $line = $sth->fetchrow ) {
2293 $biblionumbers .= $line;
2296 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2298 $results = NZoperatorAND($biblionumbers,$results);
2301 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2302 $results = $biblionumbers;
2306 warn "return : $results for LEAF : $string" if $DEBUG;
2309 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2313 my ($rightresult, $leftresult)=@_;
2315 my @leftresult = split /;/, $leftresult;
2316 warn " @leftresult / $rightresult \n" if $DEBUG;
2318 # my @rightresult = split /;/,$leftresult;
2321 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2322 # the result is stored twice, to have the same weight for AND than OR.
2323 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2324 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2325 foreach (@leftresult) {
2328 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2329 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2330 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2332 "$value-$countvalue;$value-$countvalue;";
2335 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2336 return $finalresult;
2340 my ($rightresult, $leftresult)=@_;
2341 return $rightresult.$leftresult;
2345 my ($leftresult, $rightresult)=@_;
2347 my @leftresult = split /;/, $leftresult;
2349 # my @rightresult = split /;/,$leftresult;
2351 foreach (@leftresult) {
2353 $value=$1 if $value=~m/(.*)-\d+$/;
2354 unless ($rightresult =~ "$value-") {
2355 $finalresult .= "$_;";
2358 return $finalresult;
2363 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2370 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2371 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2373 # order title asc by default
2374 # $ordering = '1=36 <i' unless $ordering;
2375 $results_per_page = 20 unless $results_per_page;
2376 $offset = 0 unless $offset;
2377 my $dbh = C4::Context->dbh;
2380 # order by POPULARITY
2382 if ( $ordering =~ /popularity/ ) {
2386 # popularity is not in MARC record, it's builded from a specific query
2388 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2389 foreach ( split /;/, $biblionumbers ) {
2390 my ( $biblionumber, $title ) = split /,/, $_;
2391 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2392 $sth->execute($biblionumber);
2393 my $popularity = $sth->fetchrow || 0;
2395 # hint : the key is popularity.title because we can have
2396 # many results with the same popularity. In this case, sub-ordering is done by title
2397 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2398 # (un-frequent, I agree, but we won't forget anything that way ;-)
2399 $popularity{ sprintf( "%10d", $popularity ) . $title
2400 . $biblionumber } = $biblionumber;
2403 # sort the hash and return the same structure as GetRecords (Zebra querying)
2406 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2407 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2408 $result_hash->{'RECORDS'}[ $numbers++ ] =
2409 $result{ $popularity{$key} }->as_usmarc();
2412 else { # sort popularity ASC
2413 foreach my $key ( sort ( keys %popularity ) ) {
2414 $result_hash->{'RECORDS'}[ $numbers++ ] =
2415 $result{ $popularity{$key} }->as_usmarc();
2418 my $finalresult = ();
2419 $result_hash->{'hits'} = $numbers;
2420 $finalresult->{'biblioserver'} = $result_hash;
2421 return $finalresult;
2427 elsif ( $ordering =~ /author/ ) {
2429 foreach ( split /;/, $biblionumbers ) {
2430 my ( $biblionumber, $title ) = split /,/, $_;
2431 my $record = GetMarcBiblio($biblionumber);
2433 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2434 $author = $record->subfield( '200', 'f' );
2435 $author = $record->subfield( '700', 'a' ) unless $author;
2438 $author = $record->subfield( '100', 'a' );
2441 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2442 # and we don't want to get only 1 result for each of them !!!
2443 $result{ $author . $biblionumber } = $record;
2446 # sort the hash and return the same structure as GetRecords (Zebra querying)
2449 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2450 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2451 $result_hash->{'RECORDS'}[ $numbers++ ] =
2452 $result{$key}->as_usmarc();
2455 else { # sort by author ASC
2456 foreach my $key ( sort ( keys %result ) ) {
2457 $result_hash->{'RECORDS'}[ $numbers++ ] =
2458 $result{$key}->as_usmarc();
2461 my $finalresult = ();
2462 $result_hash->{'hits'} = $numbers;
2463 $finalresult->{'biblioserver'} = $result_hash;
2464 return $finalresult;
2467 # ORDER BY callnumber
2470 elsif ( $ordering =~ /callnumber/ ) {
2472 foreach ( split /;/, $biblionumbers ) {
2473 my ( $biblionumber, $title ) = split /,/, $_;
2474 my $record = GetMarcBiblio($biblionumber);
2476 my $frameworkcode = GetFrameworkCode($biblionumber);
2477 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2478 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2479 unless $callnumber_tag;
2480 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2481 $callnumber = $record->subfield( '200', 'f' );
2483 $callnumber = $record->subfield( '100', 'a' );
2486 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2487 # and we don't want to get only 1 result for each of them !!!
2488 $result{ $callnumber . $biblionumber } = $record;
2491 # sort the hash and return the same structure as GetRecords (Zebra querying)
2494 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2495 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2496 $result_hash->{'RECORDS'}[ $numbers++ ] =
2497 $result{$key}->as_usmarc();
2500 else { # sort by title ASC
2501 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2502 $result_hash->{'RECORDS'}[ $numbers++ ] =
2503 $result{$key}->as_usmarc();
2506 my $finalresult = ();
2507 $result_hash->{'hits'} = $numbers;
2508 $finalresult->{'biblioserver'} = $result_hash;
2509 return $finalresult;
2511 elsif ( $ordering =~ /pubdate/ ) { #pub year
2513 foreach ( split /;/, $biblionumbers ) {
2514 my ( $biblionumber, $title ) = split /,/, $_;
2515 my $record = GetMarcBiblio($biblionumber);
2516 my ( $publicationyear_tag, $publicationyear_subfield ) =
2517 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2518 my $publicationyear =
2519 $record->subfield( $publicationyear_tag,
2520 $publicationyear_subfield );
2522 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2523 # and we don't want to get only 1 result for each of them !!!
2524 $result{ $publicationyear . $biblionumber } = $record;
2527 # sort the hash and return the same structure as GetRecords (Zebra querying)
2530 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2531 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2532 $result_hash->{'RECORDS'}[ $numbers++ ] =
2533 $result{$key}->as_usmarc();
2536 else { # sort by pub year ASC
2537 foreach my $key ( sort ( keys %result ) ) {
2538 $result_hash->{'RECORDS'}[ $numbers++ ] =
2539 $result{$key}->as_usmarc();
2542 my $finalresult = ();
2543 $result_hash->{'hits'} = $numbers;
2544 $finalresult->{'biblioserver'} = $result_hash;
2545 return $finalresult;
2551 elsif ( $ordering =~ /title/ ) {
2553 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2555 foreach ( split /;/, $biblionumbers ) {
2556 my ( $biblionumber, $title ) = split /,/, $_;
2558 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2559 # and we don't want to get only 1 result for each of them !!!
2560 # hint & speed improvement : we can order without reading the record
2561 # so order, and read records only for the requested page !
2562 $result{ $title . $biblionumber } = $biblionumber;
2565 # sort the hash and return the same structure as GetRecords (Zebra querying)
2568 if ( $ordering eq 'title_az' ) { # sort by title desc
2569 foreach my $key ( sort ( keys %result ) ) {
2570 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2573 else { # sort by title ASC
2574 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2575 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2579 # limit the $results_per_page to result size if it's more
2580 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2582 # for the requested page, replace biblionumber by the complete record
2583 # speed improvement : avoid reading too much things
2585 my $counter = $offset ;
2586 $counter <= $offset + $results_per_page ;
2590 $result_hash->{'RECORDS'}[$counter] =
2591 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2593 my $finalresult = ();
2594 $result_hash->{'hits'} = $numbers;
2595 $finalresult->{'biblioserver'} = $result_hash;
2596 return $finalresult;
2603 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2606 foreach ( split /;/, $biblionumbers ) {
2607 my ( $biblionumber, $title ) = split /,/, $_;
2608 $title =~ /(.*)-(\d)/;
2613 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2614 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2615 # biblio N has ranking = 6
2616 $count_ranking{$biblionumber} += $ranking;
2619 # build the result by "inverting" the count_ranking hash
2620 # 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
2622 foreach ( keys %count_ranking ) {
2623 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2626 # sort the hash and return the same structure as GetRecords (Zebra querying)
2629 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2630 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2633 # limit the $results_per_page to result size if it's more
2634 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2636 # for the requested page, replace biblionumber by the complete record
2637 # speed improvement : avoid reading too much things
2639 my $counter = $offset ;
2640 $counter <= $offset + $results_per_page ;
2644 $result_hash->{'RECORDS'}[$counter] =
2645 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2646 if $result_hash->{'RECORDS'}[$counter];
2648 my $finalresult = ();
2649 $result_hash->{'hits'} = $numbers;
2650 $finalresult->{'biblioserver'} = $result_hash;
2651 return $finalresult;
2655 =head2 enabled_staff_search_views
2657 %hash = enabled_staff_search_views()
2659 This function returns a hash that contains three flags obtained from the system
2660 preferences, used to determine whether a particular staff search results view
2665 =item C<Output arg:>
2667 * $hash{can_view_MARC} is true only if the MARC view is enabled
2668 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2669 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2671 =item C<usage in the script:>
2675 $template->param ( C4::Search::enabled_staff_search_views );
2679 sub enabled_staff_search_views
2682 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2683 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2684 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2688 sub AddSearchHistory{
2689 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2690 my $dbh = C4::Context->dbh;
2692 # Add the request the user just made
2693 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2694 my $sth = $dbh->prepare($sql);
2695 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2696 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2699 sub GetSearchHistory{
2700 my ($borrowernumber,$session)=@_;
2701 my $dbh = C4::Context->dbh;
2703 # Add the request the user just made
2704 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2705 my $sth = $dbh->prepare($query);
2706 $sth->execute($borrowernumber, $session);
2707 return $sth->fetchall_hashref({});
2710 =head2 z3950_search_args
2712 $arrayref = z3950_search_args($matchpoints)
2714 This function returns an array reference that contains the search parameters to be
2715 passed to the Z39.50 search script (z3950_search.pl). The array elements
2716 are hash refs whose keys are name, value and encvalue, and whose values are the
2717 name of a search parameter, the value of that search parameter and the URL encoded
2718 value of that parameter.
2720 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2722 The search parameter values are obtained from the bibliographic record whose
2723 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2725 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2726 a general purpose search argument. In this case, the returned array contains only
2727 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2729 If a search parameter value is undefined or empty, it is not included in the returned
2732 The returned array reference may be passed directly to the template parameters.
2736 =item C<Output arg:>
2738 * $array containing hash refs as described above
2740 =item C<usage in the script:>
2744 $data = Biblio::GetBiblioData($bibno);
2745 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2749 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2753 sub z3950_search_args {
2755 my $isbn = Business::ISBN->new($bibrec);
2757 if (defined $isbn && $isbn->is_valid)
2759 $bibrec = { isbn => $bibrec } if !ref $bibrec;
2762 $bibrec = { title => $bibrec } if !ref $bibrec;
2765 for my $field (qw/ lccn isbn issn title author dewey subject /)
2767 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2768 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2773 =head2 GetDistinctValues($field);
2775 C<$field> is a reference to the fields array
2779 sub GetDistinctValues {
2780 my ($fieldname,$string)=@_;
2781 # returns a reference to a hash of references to branches...
2782 if ($fieldname=~/\./){
2783 my ($table,$column)=split /\./, $fieldname;
2784 my $dbh = C4::Context->dbh;
2785 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2786 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 ");
2788 my $elements=$sth->fetchall_arrayref({});
2793 my @servers=qw<biblioserver authorityserver>;
2794 my (@zconns,@results);
2795 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2796 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2799 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2802 # The big moment: asynchronously retrieve results from all servers
2804 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2805 my $ev = $zconns[ $i - 1 ]->last_event();
2806 if ( $ev == ZOOM::Event::ZEND ) {
2807 next unless $results[ $i - 1 ];
2808 my $size = $results[ $i - 1 ]->size();
2810 for (my $j=0;$j<$size;$j++){
2812 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2813 push @elements, \%hashscan;
2823 END { } # module clean-up code here (global destructor)
2830 Koha Development Team <http://koha-community.org/>