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);
30 use C4::Reserves; # CheckReserves
34 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
36 # set the version for version checking
39 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
44 C4::Search - Functions for searching the Koha catalog.
48 See opac/opac-search.pl or catalogue/search.pl for example of usage
52 This module provides searching functions for Koha's bibliographic databases
70 #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)
72 # make all your functions, whether exported or not;
76 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
78 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
84 my $dbh = C4::Context->dbh;
85 my $result = TransformMarcToKoha( $dbh, $record, '' );
90 my ( $biblionumber, $title );
92 # search duplicate on ISBN, easy and fast..
94 if ( $result->{isbn} ) {
95 $result->{isbn} =~ s/\(.*$//;
96 $result->{isbn} =~ s/\s+$//;
97 $query = "isbn=$result->{isbn}";
100 $result->{title} =~ s /\\//g;
101 $result->{title} =~ s /\"//g;
102 $result->{title} =~ s /\(//g;
103 $result->{title} =~ s /\)//g;
105 # FIXME: instead of removing operators, could just do
106 # quotes around the value
107 $result->{title} =~ s/(and|or|not)//g;
108 $query = "ti,ext=$result->{title}";
109 $query .= " and itemtype=$result->{itemtype}"
110 if ( $result->{itemtype} );
111 if ( $result->{author} ) {
112 $result->{author} =~ s /\\//g;
113 $result->{author} =~ s /\"//g;
114 $result->{author} =~ s /\(//g;
115 $result->{author} =~ s /\)//g;
117 # remove valid operators
118 $result->{author} =~ s/(and|or|not)//g;
119 $query .= " and au,ext=$result->{author}";
123 # FIXME: add error handling
124 my ( $error, $searchresults ) = SimpleSearch($query); # FIXME :: hardcoded !
126 foreach my $possible_duplicate_record (@$searchresults) {
128 MARC::Record->new_from_usmarc($possible_duplicate_record);
129 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
131 # FIXME :: why 2 $biblionumber ?
133 push @results, $result->{'biblionumber'};
134 push @results, $result->{'title'};
142 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
144 This function provides a simple search API on the bibliographic catalog
150 * $query can be a simple keyword or a complete CCL query
151 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
152 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
153 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
158 * $error is a empty unless an error is detected
159 * \@results is an array of records.
160 * $total_hits is the number of hits that would have been returned with no limit
162 =item C<usage in the script:>
166 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
168 if (defined $error) {
169 $template->param(query_error => $error);
170 warn "error: ".$error;
171 output_html_with_http_headers $input, $cookie, $template->output;
175 my $hits = scalar @$marcresults;
178 for my $i (0..$hits) {
180 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
181 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
183 #build the hash for the template.
184 $resultsloop{title} = $biblio->{'title'};
185 $resultsloop{subtitle} = $biblio->{'subtitle'};
186 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
187 $resultsloop{author} = $biblio->{'author'};
188 $resultsloop{publishercode} = $biblio->{'publishercode'};
189 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
191 push @results, \%resultsloop;
194 $template->param(result=>\@results);
199 my ( $query, $offset, $max_results, $servers ) = @_;
201 if ( C4::Context->preference('NoZebra') ) {
202 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
205 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
206 return ( undef, $search_result, scalar($result->{hits}) );
209 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
210 my @servers = defined ( $servers ) ? @$servers : ( "biblioserver" );
216 return ( "No query entered", undef, undef ) unless $query;
218 # Initialize & Search Zebra
219 for ( my $i = 0 ; $i < @servers ; $i++ ) {
221 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
222 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
223 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
227 $zconns[$i]->errmsg() . " ("
228 . $zconns[$i]->errcode() . ") "
229 . $zconns[$i]->addinfo() . " "
230 . $zconns[$i]->diagset();
232 return ( $error, undef, undef ) if $zconns[$i]->errcode();
236 # caught a ZOOM::Exception
240 . $@->addinfo() . " "
243 return ( $error, undef, undef );
246 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
247 my $event = $zconns[ $i - 1 ]->last_event();
248 if ( $event == ZOOM::Event::ZEND ) {
250 my $first_record = defined( $offset ) ? $offset+1 : 1;
251 my $hits = $tmpresults[ $i - 1 ]->size();
252 $total_hits += $hits;
253 my $last_record = $hits;
254 if ( defined $max_results && $offset + $max_results < $hits ) {
255 $last_record = $offset + $max_results;
258 for my $j ( $first_record..$last_record ) {
259 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
260 push @results, $record;
265 foreach my $result (@tmpresults) {
268 foreach my $zoom_query (@zoom_queries) {
269 $zoom_query->destroy();
272 return ( undef, \@results, $total_hits );
278 ( undef, $results_hashref, \@facets_loop ) = getRecords (
280 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
281 $results_per_page, $offset, $expanded_facet, $branches,
285 The all singing, all dancing, multi-server, asynchronous, scanning,
286 searching, record nabbing, facet-building
288 See verbse embedded documentation.
294 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
295 $results_per_page, $offset, $expanded_facet, $branches,
299 my @servers = @$servers_ref;
300 my @sort_by = @$sort_by_ref;
302 # Initialize variables for the ZOOM connection and results object
306 my $results_hashref = ();
308 # Initialize variables for the faceted results objects
309 my $facets_counter = ();
310 my $facets_info = ();
311 my $facets = getFacets();
313 my @facets_loop; # stores the ref to array of hashes for template facets loop
315 ### LOOP THROUGH THE SERVERS
316 for ( my $i = 0 ; $i < @servers ; $i++ ) {
317 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
319 # perform the search, create the results objects
320 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
321 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
323 #$query_to_use = $simple_query if $scan;
324 warn $simple_query if ( $scan and $DEBUG );
326 # Check if we've got a query_type defined, if so, use it
329 if ($query_type =~ /^ccl/) {
330 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
331 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
332 } elsif ($query_type =~ /^cql/) {
333 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
334 } elsif ($query_type =~ /^pqf/) {
335 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
337 warn "Unknown query_type '$query_type'. Results undetermined.";
340 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
342 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
346 warn "WARNING: query problem with $query_to_use " . $@;
349 # Concatenate the sort_by limits and pass them to the results object
350 # Note: sort will override rank
352 foreach my $sort (@sort_by) {
353 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
354 $sort_by .= "1=1003 <i ";
356 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
357 $sort_by .= "1=1003 >i ";
359 elsif ( $sort eq "popularity_asc" ) {
360 $sort_by .= "1=9003 <i ";
362 elsif ( $sort eq "popularity_dsc" ) {
363 $sort_by .= "1=9003 >i ";
365 elsif ( $sort eq "call_number_asc" ) {
366 $sort_by .= "1=8007 <i ";
368 elsif ( $sort eq "call_number_dsc" ) {
369 $sort_by .= "1=8007 >i ";
371 elsif ( $sort eq "pubdate_asc" ) {
372 $sort_by .= "1=31 <i ";
374 elsif ( $sort eq "pubdate_dsc" ) {
375 $sort_by .= "1=31 >i ";
377 elsif ( $sort eq "acqdate_asc" ) {
378 $sort_by .= "1=32 <i ";
380 elsif ( $sort eq "acqdate_dsc" ) {
381 $sort_by .= "1=32 >i ";
383 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
384 $sort_by .= "1=4 <i ";
386 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
387 $sort_by .= "1=4 >i ";
390 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
393 if ($sort_by && !$scan) {
394 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
395 warn "WARNING sort $sort_by failed";
398 } # finished looping through servers
400 # The big moment: asynchronously retrieve results from all servers
401 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
402 my $ev = $zconns[ $i - 1 ]->last_event();
403 if ( $ev == ZOOM::Event::ZEND ) {
404 next unless $results[ $i - 1 ];
405 my $size = $results[ $i - 1 ]->size();
409 # loop through the results
410 $results_hash->{'hits'} = $size;
412 if ( $offset + $results_per_page <= $size ) {
413 $times = $offset + $results_per_page;
418 for ( my $j = $offset ; $j < $times ; $j++ ) {
423 ## Check if it's an index scan
425 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
427 # here we create a minimal MARC record and hand it off to the
428 # template just like a normal result ... perhaps not ideal, but
430 my $tmprecord = MARC::Record->new();
431 $tmprecord->encoding('UTF-8');
435 # the minimal record in author/title (depending on MARC flavour)
436 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
437 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
438 $tmprecord->append_fields($tmptitle);
440 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
441 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
442 $tmprecord->append_fields($tmptitle);
443 $tmprecord->append_fields($tmpauthor);
445 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
450 $record = $results[ $i - 1 ]->record($j)->raw();
452 # warn "RECORD $j:".$record;
453 $results_hash->{'RECORDS'}[$j] = $record;
455 # Fill the facets while we're looping, but only for the biblioserver
456 $facet_record = MARC::Record->new_from_usmarc($record)
457 if $servers[ $i - 1 ] =~ /biblioserver/;
459 #warn $servers[$i-1]."\n".$record; #.$facet_record->title();
461 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
462 ($facets->[$k]) or next;
463 my @fields = map {$facet_record->field($_)} @{$facets->[$k]->{'tags'}} ;
464 for my $field (@fields) {
465 my @subfields = $field->subfields();
466 for my $subfield (@subfields) {
467 my ( $code, $data ) = @$subfield;
468 ($code eq $facets->[$k]->{'subfield'}) or next;
469 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
472 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} =
473 $facets->[$k]->{'label_value'};
474 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} =
475 $facets->[$k]->{'expanded'};
480 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
483 # warn "connection ", $i-1, ": $size hits";
484 # warn $results[$i-1]->record(0)->render() if $size > 0;
487 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
489 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
490 keys %$facets_counter )
493 my $number_of_facets;
494 my @this_facets_array;
497 $facets_counter->{$link_value}->{$b}
498 <=> $facets_counter->{$link_value}->{$a}
499 } keys %{ $facets_counter->{$link_value} }
503 if ( ( $number_of_facets < 6 )
504 || ( $expanded_facet eq $link_value )
505 || ( $facets_info->{$link_value}->{'expanded'} ) )
508 # Sanitize the link value ), ( will cause errors with CCL,
509 my $facet_link_value = $one_facet;
510 $facet_link_value =~ s/(\(|\))/ /g;
512 # fix the length that will display in the label,
513 my $facet_label_value = $one_facet;
515 substr( $one_facet, 0, 20 ) . "..."
516 unless length($facet_label_value) <= 20;
518 # if it's a branch, label by the name, not the code,
519 if ( $link_value =~ /branch/ ) {
520 if (defined $branches
521 && ref($branches) eq "HASH"
522 && defined $branches->{$one_facet}
523 && ref ($branches->{$one_facet}) eq "HASH")
526 $branches->{$one_facet}->{'branchname'};
529 $facet_label_value = "*";
533 # but we're down with the whole label being in the link's title.
534 push @this_facets_array, {
535 facet_count => $facets_counter->{$link_value}->{$one_facet},
536 facet_label_value => $facet_label_value,
537 facet_title_value => $one_facet,
538 facet_link_value => $facet_link_value,
539 type_link_value => $link_value,
544 # handle expanded option
545 unless ( $facets_info->{$link_value}->{'expanded'} ) {
547 if ( ( $number_of_facets > 6 )
548 && ( $expanded_facet ne $link_value ) );
551 type_link_value => $link_value,
552 type_id => $link_value . "_id",
553 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
554 facets => \@this_facets_array,
555 expandable => $expandable,
556 expand => $link_value,
557 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
562 return ( undef, $results_hashref, \@facets_loop );
567 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
568 $results_per_page, $offset, $expanded_facet, $branches,
572 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
574 $paz->search($simple_query);
575 sleep 1; # FIXME: WHY?
578 my $results_hashref = {};
579 my $stats = XMLin($paz->stat);
580 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
582 # for a grouped search result, the number of hits
583 # is the number of groups returned; 'bib_hits' will have
584 # the total number of bibs.
585 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
586 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
588 HIT: foreach my $hit (@{ $results->{'hit'} }) {
589 my $recid = $hit->{recid}->[0];
591 my $work_title = $hit->{'md-work-title'}->[0];
593 if (exists $hit->{'md-work-author'}) {
594 $work_author = $hit->{'md-work-author'}->[0];
596 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
598 my $result_group = {};
599 $result_group->{'group_label'} = $group_label;
600 $result_group->{'group_merge_key'} = $recid;
603 if (exists $hit->{count}) {
604 $count = $hit->{count}->[0];
606 $result_group->{'group_count'} = $count;
608 for (my $i = 0; $i < $count; $i++) {
609 # FIXME -- may need to worry about diacritics here
610 my $rec = $paz->record($recid, $i);
611 push @{ $result_group->{'RECORDS'} }, $rec;
614 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
617 # pass through facets
618 my $termlist_xml = $paz->termlist('author,subject');
619 my $terms = XMLin($termlist_xml, forcearray => 1);
620 my @facets_loop = ();
621 #die Dumper($results);
622 # foreach my $list (sort keys %{ $terms->{'list'} }) {
624 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
626 # facet_label_value => $facet->{'name'}->[0],
629 # push @facets_loop, ( {
630 # type_label => $list,
631 # facets => \@facets,
635 return ( undef, $results_hashref, \@facets_loop );
639 sub _remove_stopwords {
640 my ( $operand, $index ) = @_;
641 my @stopwords_removed;
643 # phrase and exact-qualified indexes shouldn't have stopwords removed
644 if ( $index !~ m/phr|ext/ ) {
646 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
647 # we use IsAlpha unicode definition, to deal correctly with diacritics.
648 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
649 # is a stopword, we'd get "çon" and wouldn't find anything...
651 foreach ( keys %{ C4::Context->stopwords } ) {
652 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
653 if ( my ($matched) = ($operand =~
654 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
656 $operand =~ s/\Q$matched\E/ /gi;
657 push @stopwords_removed, $_;
661 return ( $operand, \@stopwords_removed );
665 sub _detect_truncation {
666 my ( $operand, $index ) = @_;
667 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
670 my @wordlist = split( /\s/, $operand );
671 foreach my $word (@wordlist) {
672 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
673 push @rightlefttruncated, $word;
675 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
676 push @lefttruncated, $word;
678 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
679 push @righttruncated, $word;
681 elsif ( index( $word, "*" ) < 0 ) {
682 push @nontruncated, $word;
685 push @regexpr, $word;
689 \@nontruncated, \@righttruncated, \@lefttruncated,
690 \@rightlefttruncated, \@regexpr
695 sub _build_stemmed_operand {
696 my ($operand,$lang) = @_;
697 require Lingua::Stem::Snowball ;
700 # If operand contains a digit, it is almost certainly an identifier, and should
701 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
702 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
703 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
704 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
705 return $operand if $operand =~ /\d/;
707 # FIXME: the locale should be set based on the user's language and/or search choice
709 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
710 encoding => "UTF-8" );
712 my @words = split( / /, $operand );
713 my @stems = $stemmer->stem(\@words);
714 for my $stem (@stems) {
715 $stemmed_operand .= "$stem";
716 $stemmed_operand .= "?"
717 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
718 $stemmed_operand .= " ";
720 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
721 return $stemmed_operand;
725 sub _build_weighted_query {
727 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
728 # pretty well but could work much better if we had a smarter query parser
729 my ( $operand, $stemmed_operand, $index ) = @_;
730 my $stemming = C4::Context->preference("QueryStemming") || 0;
731 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
732 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
734 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
736 # Keyword, or, no index specified
737 if ( ( $index eq 'kw' ) || ( !$index ) ) {
739 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
740 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
741 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
742 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
743 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
744 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
745 if $fuzzy_enabled; # add fuzzy, word list
746 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
747 if ( $stemming and $stemmed_operand )
748 ; # add stemming, right truncation
749 $weighted_query .= " or wrdl,r9=\"$operand\"";
751 # embedded sorting: 0 a-z; 1 z-a
752 # $weighted_query .= ") or (sort1,aut=1";
755 # Barcode searches should skip this process
756 elsif ( $index eq 'bc' ) {
757 $weighted_query .= "bc=\"$operand\"";
760 # Authority-number searches should skip this process
761 elsif ( $index eq 'an' ) {
762 $weighted_query .= "an=\"$operand\"";
765 # If the index already has more than one qualifier, wrap the operand
766 # in quotes and pass it back (assumption is that the user knows what they
767 # are doing and won't appreciate us mucking up their query
768 elsif ( $index =~ ',' ) {
769 $weighted_query .= " $index=\"$operand\"";
772 #TODO: build better cases based on specific search indexes
774 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
775 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
776 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
778 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
781 $weighted_query .= "))"; # close rank specification
782 return $weighted_query;
787 Return an array with available indexes.
809 'Author-personal-bibliography',
818 'Chronological-subdivision',
828 'Conference-name-heading',
829 'Conference-name-see',
830 'Conference-name-seealso',
835 'Corporate-name-heading',
836 'Corporate-name-see',
837 'Corporate-name-seealso',
839 'date-entered-on-file',
840 'Date-of-acquisition',
841 'Date-of-publication',
842 'Dewey-classification',
848 'Geographic-subdivision',
851 'Heading-use-main-or-added-entry',
852 'Heading-use-series-added-entry ',
853 'Heading-use-subject-added-entry',
871 'Local-classification',
874 'Match-heading-see-from',
881 'Name-geographic-heading',
882 'Name-geographic-see',
883 'Name-geographic-seealso',
891 'Personal-name-heading',
893 'Personal-name-seealso',
909 'Subject-heading-thesaurus',
910 'Subject-name-personal',
911 'Subject-subdivision',
920 'Term-genre-form-heading',
921 'Term-genre-form-see',
922 'Term-genre-form-seealso',
929 'Title-uniform-heading',
931 'Title-uniform-seealso',
941 'classification-source',
943 'coded-location-qualifier',
954 'Local-classification',
957 'materials-specified',
966 'replacementpricedate',
982 $simple_query, $query_cgi,
984 $limit_cgi, $limit_desc,
985 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
987 Build queries and limits in CCL, CGI, Human,
988 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
990 See verbose embedded documentation.
996 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
998 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1001 my @operators = $operators ? @$operators : ();
1002 my @indexes = $indexes ? @$indexes : ();
1003 my @operands = $operands ? @$operands : ();
1004 my @limits = $limits ? @$limits : ();
1005 my @sort_by = $sort_by ? @$sort_by : ();
1007 my $stemming = C4::Context->preference("QueryStemming") || 0;
1008 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1009 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1010 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1011 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1013 # no stemming/weight/fuzzy in NoZebra
1014 if ( C4::Context->preference("NoZebra") ) {
1018 $auto_truncation = 0;
1021 my $query = $operands[0];
1022 my $simple_query = $operands[0];
1024 # initialize the variables we're passing back
1033 my $stopwords_removed; # flag to determine if stopwords have been removed
1036 my $cclindexes = getIndexes();
1037 if( $query !~ /\s*ccl=/ ){
1038 for my $index (@$cclindexes){
1039 if($query =~ /($index)(,?\w)*[:=]/){
1043 $query = "ccl=$query" if($cclq);
1046 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1048 if ( $query =~ /^ccl=/ ) {
1050 # This is needed otherwise ccl= and &limit won't work together, and
1051 # this happens when selecting a subject on the opac-detail page
1053 $q .= ' and '.join(' and ', @limits);
1055 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1057 if ( $query =~ /^cql=/ ) {
1058 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1060 if ( $query =~ /^pqf=/ ) {
1061 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1064 # pass nested queries directly
1065 # FIXME: need better handling of some of these variables in this case
1066 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1067 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1068 # if ( $query =~ /(\(|\))/ ) {
1070 # undef, $query, $simple_query, $query_cgi,
1071 # $query, $limit, $limit_cgi, $limit_desc,
1072 # $stopwords_removed, 'ccl'
1076 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1077 # query operands and indexes and add stemming, truncation, field weighting, etc.
1078 # Once we do so, we'll end up with a value in $query, just like if we had an
1079 # incoming $query from the user
1082 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1083 my $previous_operand
1084 ; # a flag used to keep track if there was a previous query
1085 # if there was, we can apply the current operator
1087 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1089 # COMBINE OPERANDS, INDEXES AND OPERATORS
1090 if ( $operands[$i] ) {
1091 $operands[$i]=~s/^\s+//;
1093 # A flag to determine whether or not to add the index to the query
1096 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1097 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1100 $remove_stopwords = 0;
1102 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1104 my $operand = $operands[$i];
1105 my $index = $indexes[$i];
1107 # Add index-specific attributes
1108 # Date of Publication
1109 if ( $index eq 'yr' ) {
1110 $index .= ",st-numeric";
1112 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1115 # Date of Acquisition
1116 elsif ( $index eq 'acqdate' ) {
1117 $index .= ",st-date-normalized";
1119 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1121 # ISBN,ISSN,Standard Number, don't need special treatment
1122 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1124 $stemming, $auto_truncation,
1125 $weight_fields, $fuzzy_enabled,
1127 ) = ( 0, 0, 0, 0, 0 );
1135 # Set default structure attribute (word list)
1136 my $struct_attr = q{};
1137 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1138 $struct_attr = ",wrdl";
1141 # Some helpful index variants
1142 my $index_plus = $index . $struct_attr . ':';
1143 my $index_plus_comma = $index . $struct_attr . ',';
1146 if ($remove_stopwords) {
1147 ( $operand, $stopwords_removed ) =
1148 _remove_stopwords( $operand, $index );
1149 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1150 warn "REMOVED STOPWORDS: @$stopwords_removed"
1151 if ( $stopwords_removed && $DEBUG );
1154 if ($auto_truncation){
1155 unless ( $index =~ /(st-|phr|ext)/ ) {
1156 #FIXME only valid with LTR scripts
1157 $operand=join(" ",map{
1158 (index($_,"*")>0?"$_":"$_*")
1159 }split (/\s+/,$operand));
1160 warn $operand if $DEBUG;
1165 my $truncated_operand;
1166 my( $nontruncated, $righttruncated, $lefttruncated,
1167 $rightlefttruncated, $regexpr
1168 ) = _detect_truncation( $operand, $index );
1170 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1175 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1176 scalar(@$rightlefttruncated) > 0 )
1179 # Don't field weight or add the index to the query, we do it here
1181 undef $weight_fields;
1182 my $previous_truncation_operand;
1183 if (scalar @$nontruncated) {
1184 $truncated_operand .= "$index_plus @$nontruncated ";
1185 $previous_truncation_operand = 1;
1187 if (scalar @$righttruncated) {
1188 $truncated_operand .= "and " if $previous_truncation_operand;
1189 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1190 $previous_truncation_operand = 1;
1192 if (scalar @$lefttruncated) {
1193 $truncated_operand .= "and " if $previous_truncation_operand;
1194 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1195 $previous_truncation_operand = 1;
1197 if (scalar @$rightlefttruncated) {
1198 $truncated_operand .= "and " if $previous_truncation_operand;
1199 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1200 $previous_truncation_operand = 1;
1203 $operand = $truncated_operand if $truncated_operand;
1204 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1207 my $stemmed_operand;
1208 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1211 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1213 # Handle Field Weighting
1214 my $weighted_operand;
1215 if ($weight_fields) {
1216 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1217 $operand = $weighted_operand;
1221 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1223 # If there's a previous operand, we need to add an operator
1224 if ($previous_operand) {
1226 # User-specified operator
1227 if ( $operators[ $i - 1 ] ) {
1228 $query .= " $operators[$i-1] ";
1229 $query .= " $index_plus " unless $indexes_set;
1230 $query .= " $operand";
1231 $query_cgi .= "&op=$operators[$i-1]";
1232 $query_cgi .= "&idx=$index" if $index;
1233 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1235 " $operators[$i-1] $index_plus $operands[$i]";
1238 # Default operator is and
1241 $query .= "$index_plus " unless $indexes_set;
1242 $query .= "$operand";
1243 $query_cgi .= "&op=and&idx=$index" if $index;
1244 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1245 $query_desc .= " and $index_plus $operands[$i]";
1249 # There isn't a pervious operand, don't need an operator
1252 # Field-weighted queries already have indexes set
1253 $query .= " $index_plus " unless $indexes_set;
1255 $query_desc .= " $index_plus $operands[$i]";
1256 $query_cgi .= "&idx=$index" if $index;
1257 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1258 $previous_operand = 1;
1263 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1266 my $group_OR_limits;
1267 my $availability_limit;
1268 foreach my $this_limit (@limits) {
1269 if ( $this_limit =~ /available/ ) {
1271 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1273 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1274 $availability_limit .=
1275 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1276 $limit_cgi .= "&limit=available";
1280 # group_OR_limits, prefixed by mc-
1281 # OR every member of the group
1282 elsif ( $this_limit =~ /mc/ ) {
1284 if ( $this_limit =~ /mc-ccode:/ ) {
1285 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1286 $this_limit =~ tr/"//d;
1287 my ($k,$v) = split(/:/, $this_limit,2);
1288 $this_limit = $k.":\"".$v."\"";
1291 $group_OR_limits .= " or " if $group_OR_limits;
1292 $limit_desc .= " or " if $group_OR_limits;
1293 $group_OR_limits .= "$this_limit";
1294 $limit_cgi .= "&limit=$this_limit";
1295 $limit_desc .= " $this_limit";
1298 # Regular old limits
1300 $limit .= " and " if $limit || $query;
1301 $limit .= "$this_limit";
1302 $limit_cgi .= "&limit=$this_limit";
1303 if ($this_limit =~ /^branch:(.+)/) {
1304 my $branchcode = $1;
1305 my $branchname = GetBranchName($branchcode);
1306 if (defined $branchname) {
1307 $limit_desc .= " branch:$branchname";
1309 $limit_desc .= " $this_limit";
1312 $limit_desc .= " $this_limit";
1316 if ($group_OR_limits) {
1317 $limit .= " and " if ( $query || $limit );
1318 $limit .= "($group_OR_limits)";
1320 if ($availability_limit) {
1321 $limit .= " and " if ( $query || $limit );
1322 $limit .= "($availability_limit)";
1325 # Normalize the query and limit strings
1326 # This is flawed , means we can't search anything with : in it
1327 # if user wants to do ccl or cql, start the query with that
1328 # $query =~ s/:/=/g;
1329 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1330 $query =~ s/(?<=(wrdl)):/=/g;
1331 $query =~ s/(?<=(trn|phr)):/=/g;
1333 for ( $query, $query_desc, $limit, $limit_desc ) {
1334 s/ +/ /g; # remove extra spaces
1335 s/^ //g; # remove any beginning spaces
1336 s/ $//g; # remove any ending spaces
1337 s/==/=/g; # remove double == from query
1339 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1341 for ($query_cgi,$simple_query) {
1344 # append the limit to the query
1345 $query .= " " . $limit;
1349 warn "QUERY:" . $query;
1350 warn "QUERY CGI:" . $query_cgi;
1351 warn "QUERY DESC:" . $query_desc;
1352 warn "LIMIT:" . $limit;
1353 warn "LIMIT CGI:" . $limit_cgi;
1354 warn "LIMIT DESC:" . $limit_desc;
1355 warn "---------\nLeave buildQuery\n---------";
1358 undef, $query, $simple_query, $query_cgi,
1359 $query_desc, $limit, $limit_cgi, $limit_desc,
1360 $stopwords_removed, $query_type
1364 =head2 searchResults
1366 my @search_results = searchResults($search_context, $searchdesc, $hits,
1367 $results_per_page, $offset, $scan,
1368 @marcresults, $hidelostitems);
1370 Format results in a form suitable for passing to the template
1374 # IMO this subroutine is pretty messy still -- it's responsible for
1375 # building the HTML output for the template
1377 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1378 my $dbh = C4::Context->dbh;
1381 $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1383 #Build branchnames hash
1385 #get branch information.....
1387 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1389 while ( my $bdata = $bsth->fetchrow_hashref ) {
1390 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1392 # FIXME - We build an authorised values hash here, using the default framework
1393 # though it is possible to have different authvals for different fws.
1395 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1397 # get notforloan authorised value list (see $shelflocations FIXME)
1398 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1400 #Build itemtype hash
1401 #find itemtype & itemtype image
1405 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1408 while ( my $bdata = $bsth->fetchrow_hashref ) {
1409 foreach (qw(description imageurl summary notforloan)) {
1410 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1414 #search item field code
1417 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1420 my ($itemtag) = $sth->fetchrow;
1422 ## find column names of items related to MARC
1423 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1425 my %subfieldstosearch;
1426 while ( ( my $column ) = $sth2->fetchrow ) {
1427 my ( $tagfield, $tagsubfield ) =
1428 &GetMarcFromKohaField( "items." . $column, "" );
1429 $subfieldstosearch{$column} = $tagsubfield;
1432 # handle which records to actually retrieve
1434 if ( $hits && $offset + $results_per_page <= $hits ) {
1435 $times = $offset + $results_per_page;
1438 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1441 my $marcflavour = C4::Context->preference("marcflavour");
1442 # We get the biblionumber position in MARC
1443 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1446 # loop through all of the records we've retrieved
1447 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1448 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1452 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1453 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1454 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1455 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1456 $oldbiblio->{result_number} = $i + 1;
1458 # add imageurl to itemtype if there is one
1459 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1461 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1462 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1463 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1464 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1465 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1466 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1468 # edition information, if any
1469 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1470 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1471 # Build summary if there is one (the summary is defined in the itemtypes table)
1472 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1473 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1474 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1475 my @fields = $marcrecord->fields();
1478 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1480 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1481 $tag =~ /(.{3})(.)/;
1482 if($marcrecord->field($1)){
1483 my @abc = $marcrecord->field($1)->subfield($2);
1484 $tags->{$tag} = $#abc + 1 ;
1488 # We catch how many times to repeat this line
1490 foreach my $tag (keys(%$tags)){
1491 $max = $tags->{$tag} if($tags->{$tag} > $max);
1494 # we replace, and repeat each line
1495 for (my $i = 0 ; $i < $max ; $i++){
1496 my $newline = $line;
1498 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1499 $tag =~ /(.{3})(.)/;
1501 if($marcrecord->field($1)){
1502 my @repl = $marcrecord->field($1)->subfield($2);
1503 my $subfieldvalue = $repl[$i];
1505 if (! utf8::is_utf8($subfieldvalue)) {
1506 utf8::decode($subfieldvalue);
1509 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1512 $newsummary .= "$newline\n";
1516 $newsummary =~ s/\[(.*?)]//g;
1517 $newsummary =~ s/\n/<br\/>/g;
1518 $oldbiblio->{summary} = $newsummary;
1521 # Pull out the items fields
1522 my @fields = $marcrecord->field($itemtag);
1524 # Setting item statuses for display
1525 my @available_items_loop;
1526 my @onloan_items_loop;
1527 my @other_items_loop;
1529 my $available_items;
1533 my $ordered_count = 0;
1534 my $available_count = 0;
1535 my $onloan_count = 0;
1536 my $longoverdue_count = 0;
1537 my $other_count = 0;
1538 my $wthdrawn_count = 0;
1539 my $itemlost_count = 0;
1540 my $itembinding_count = 0;
1541 my $itemdamaged_count = 0;
1542 my $item_in_transit_count = 0;
1543 my $can_place_holds = 0;
1544 my $item_onhold_count = 0;
1545 my $items_count = scalar(@fields);
1547 ( C4::Context->preference('maxItemsinSearchResults') )
1548 ? C4::Context->preference('maxItemsinSearchResults') - 1
1551 # loop through every item
1552 foreach my $field (@fields) {
1555 # populate the items hash
1556 foreach my $code ( keys %subfieldstosearch ) {
1557 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1560 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1561 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1562 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1563 if ($item->{$hbranch}) {
1564 $item->{'branchname'} = $branches{$item->{$hbranch}};
1566 elsif ($item->{$otherbranch}) { # Last resort
1567 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1570 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1571 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1572 if ( $item->{onloan} ) {
1574 my $key = $prefix . $item->{onloan} . $item->{barcode};
1575 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1576 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1577 $onloan_items->{$key}->{branchname} = $item->{branchname};
1578 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1579 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1580 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1581 # if something's checked out and lost, mark it as 'long overdue'
1582 if ( $item->{itemlost} ) {
1583 $onloan_items->{$prefix}->{longoverdue}++;
1584 $longoverdue_count++;
1585 } else { # can place holds as long as item isn't lost
1586 $can_place_holds = 1;
1590 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1594 if ( $item->{notforloan} == -1 ) {
1598 # is item in transit?
1599 my $transfertwhen = '';
1600 my ($transfertfrom, $transfertto);
1602 # is item on the reserve shelf?
1603 my $reservestatus = 0;
1606 unless ($item->{wthdrawn}
1607 || $item->{itemlost}
1609 || $item->{notforloan}
1610 || $items_count > 20) {
1612 # A couple heuristics to limit how many times
1613 # we query the database for item transfer information, sacrificing
1614 # accuracy in some cases for speed;
1616 # 1. don't query if item has one of the other statuses
1617 # 2. don't check transit status if the bib has
1618 # more than 20 items
1620 # FIXME: to avoid having the query the database like this, and to make
1621 # the in transit status count as unavailable for search limiting,
1622 # should map transit status to record indexed in Zebra.
1624 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1625 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1628 # item is withdrawn, lost or damaged
1629 if ( $item->{wthdrawn}
1630 || $item->{itemlost}
1632 || $item->{notforloan} > 0
1633 || $reservestatus eq 'Waiting'
1634 || ($transfertwhen ne ''))
1636 $wthdrawn_count++ if $item->{wthdrawn};
1637 $itemlost_count++ if $item->{itemlost};
1638 $itemdamaged_count++ if $item->{damaged};
1639 $item_in_transit_count++ if $transfertwhen ne '';
1640 $item_onhold_count++ if $reservestatus eq 'Waiting';
1641 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1644 my $key = $prefix . $item->{status};
1645 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1646 $other_items->{$key}->{$_} = $item->{$_};
1648 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1649 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1650 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1651 $other_items->{$key}->{count}++ if $item->{$hbranch};
1652 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1653 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1657 $can_place_holds = 1;
1659 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1660 foreach (qw(branchname itemcallnumber)) {
1661 $available_items->{$prefix}->{$_} = $item->{$_};
1663 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1664 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1667 } # notforloan, item level and biblioitem level
1668 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1670 ( C4::Context->preference('maxItemsinSearchResults') )
1671 ? C4::Context->preference('maxItemsinSearchResults') - 1
1673 for my $key ( sort keys %$onloan_items ) {
1674 (++$onloanitemscount > $maxitems) and last;
1675 push @onloan_items_loop, $onloan_items->{$key};
1677 for my $key ( sort keys %$other_items ) {
1678 (++$otheritemscount > $maxitems) and last;
1679 push @other_items_loop, $other_items->{$key};
1681 for my $key ( sort keys %$available_items ) {
1682 (++$availableitemscount > $maxitems) and last;
1683 push @available_items_loop, $available_items->{$key}
1686 # XSLT processing of some stuff
1688 SetUTF8Flag($marcrecord);
1689 $debug && warn $marcrecord->as_formatted;
1690 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1691 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1692 # is not implemented yet
1693 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1694 $search_context, 1);
1695 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1699 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1700 if (!C4::Context->preference("item-level_itypes")) {
1701 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1702 $can_place_holds = 0;
1705 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1706 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1707 $oldbiblio->{items_count} = $items_count;
1708 $oldbiblio->{available_items_loop} = \@available_items_loop;
1709 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1710 $oldbiblio->{other_items_loop} = \@other_items_loop;
1711 $oldbiblio->{availablecount} = $available_count;
1712 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1713 $oldbiblio->{onloancount} = $onloan_count;
1714 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1715 $oldbiblio->{othercount} = $other_count;
1716 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1717 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1718 $oldbiblio->{itemlostcount} = $itemlost_count;
1719 $oldbiblio->{damagedcount} = $itemdamaged_count;
1720 $oldbiblio->{intransitcount} = $item_in_transit_count;
1721 $oldbiblio->{onholdcount} = $item_onhold_count;
1722 $oldbiblio->{orderedcount} = $ordered_count;
1723 $oldbiblio->{isbn} =~
1724 s/-//g; # deleting - in isbn to enable amazon content
1725 push( @newresults, $oldbiblio )
1726 if(not $hidelostitems
1727 or (($items_count > $itemlost_count )
1728 && $hidelostitems));
1734 =head2 SearchAcquisitions
1735 Search for acquisitions
1738 sub SearchAcquisitions{
1739 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1741 my $dbh=C4::Context->dbh;
1742 # Variable initialization
1746 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1747 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1748 WHERE dateaccessioned BETWEEN ? AND ?
1751 my (@params,@loopcriteria);
1753 push @params, $datebegin->output("iso");
1754 push @params, $dateend->output("iso");
1756 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1757 if(C4::Context->preference("item-level_itypes")){
1758 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1760 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1762 push @params, @$itemtypes;
1765 if ($criteria =~/itemtype/){
1766 if(C4::Context->preference("item-level_itypes")){
1767 $str .= "AND items.itype=? ";
1769 $str .= "AND biblioitems.itemtype=? ";
1772 if(scalar(@$itemtypes) == 0){
1773 my $itypes = GetItemTypes();
1774 for my $key (keys %$itypes){
1775 push @$itemtypes, $key;
1779 @loopcriteria= @$itemtypes;
1780 }elsif ($criteria=~/itemcallnumber/){
1781 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1782 OR items.itemcallnumber is NULL
1783 OR items.itemcallnumber = '')";
1785 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1787 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1788 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1791 if ($orderby =~ /date_desc/){
1792 $str.=" ORDER BY dateaccessioned DESC";
1794 $str.=" ORDER BY title";
1797 my $qdataacquisitions=$dbh->prepare($str);
1799 my @loopacquisitions;
1800 foreach my $value(@loopcriteria){
1801 push @params,$value;
1803 $cell{"title"}=$value;
1804 $cell{"titlecode"}=$value;
1806 eval{$qdataacquisitions->execute(@params);};
1808 if ($@){ warn "recentacquisitions Error :$@";}
1811 while (my $data=$qdataacquisitions->fetchrow_hashref){
1812 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1814 $cell{"loopdata"}=\@loopdata;
1816 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1819 $qdataacquisitions->finish;
1820 return \@loopacquisitions;
1822 #----------------------------------------------------------------------
1824 # Non-Zebra GetRecords#
1825 #----------------------------------------------------------------------
1829 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1835 $query, $simple_query, $sort_by_ref, $servers_ref,
1836 $results_per_page, $offset, $expanded_facet, $branches,
1839 warn "query =$query" if $DEBUG;
1840 my $result = NZanalyse($query);
1841 warn "results =$result" if $DEBUG;
1843 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1849 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1850 the list is built from an inverted index in the nozebra SQL table
1851 note that title is here only for convenience : the sorting will be very fast when requested on title
1852 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1857 my ( $string, $server ) = @_;
1858 # warn "---------" if $DEBUG;
1859 warn " NZanalyse" if $DEBUG;
1860 # warn "---------" if $DEBUG;
1862 # $server contains biblioserver or authorities, depending on what we search on.
1863 #warn "querying : $string on $server";
1864 $server = 'biblioserver' unless $server;
1866 # if we have a ", replace the content to discard temporarily any and/or/not inside
1868 if ( $string =~ /"/ ) {
1869 $string =~ s/"(.*?)"/__X__/;
1871 warn "commacontent : $commacontent" if $DEBUG;
1874 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1875 # then, call again NZanalyse with $left and $right
1876 # (recursive until we find a leaf (=> something without and/or/not)
1877 # delete repeated operator... Would then go in infinite loop
1878 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1881 #process parenthesis before.
1882 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1885 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1887 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1889 my $leftresult = NZanalyse( $left, $server );
1891 my $rightresult = NZanalyse( $right, $server );
1893 # OK, we have the results for right and left part of the query
1894 # depending of operand, intersect, union or exclude both lists
1895 # to get a result list
1896 if ( $operator eq ' and ' ) {
1897 return NZoperatorAND($leftresult,$rightresult);
1899 elsif ( $operator eq ' or ' ) {
1901 # just merge the 2 strings
1902 return $leftresult . $rightresult;
1904 elsif ( $operator eq ' not ' ) {
1905 return NZoperatorNOT($leftresult,$rightresult);
1909 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1913 warn "string :" . $string if $DEBUG;
1917 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1920 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1922 warn "no parenthesis. left : $left operator: $operator right: $right"
1925 # it's not a leaf, we have a and/or/not
1928 # reintroduce comma content if needed
1929 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1930 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1931 warn "node : $left / $operator / $right\n" if $DEBUG;
1932 my $leftresult = NZanalyse( $left, $server );
1933 my $rightresult = NZanalyse( $right, $server );
1934 warn " leftresult : $leftresult" if $DEBUG;
1935 warn " rightresult : $rightresult" if $DEBUG;
1936 # OK, we have the results for right and left part of the query
1937 # depending of operand, intersect, union or exclude both lists
1938 # to get a result list
1939 if ( $operator eq ' and ' ) {
1940 return NZoperatorAND($leftresult,$rightresult);
1942 elsif ( $operator eq ' or ' ) {
1944 # just merge the 2 strings
1945 return $leftresult . $rightresult;
1947 elsif ( $operator eq ' not ' ) {
1948 return NZoperatorNOT($leftresult,$rightresult);
1952 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1953 die "error : operand unknown : $operator for $string";
1956 # it's a leaf, do the real SQL query and return the result
1959 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1960 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1961 #remove trailing blank at the beginning
1963 warn "leaf:$string" if $DEBUG;
1965 # parse the string in in operator/operand/value again
1969 if ($string =~ /(.*)(>=|<=)(.*)/) {
1976 # warn "handling leaf... left:$left operator:$operator right:$right"
1978 unless ($operator) {
1979 if ($string =~ /(.*)(>|<|=)(.*)/) {
1984 "handling unless (operator)... left:$left operator:$operator right:$right"
1992 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1995 # automatic replace for short operators
1996 $left = 'title' if $left =~ '^ti$';
1997 $left = 'author' if $left =~ '^au$';
1998 $left = 'publisher' if $left =~ '^pb$';
1999 $left = 'subject' if $left =~ '^su$';
2000 $left = 'koha-Auth-Number' if $left =~ '^an$';
2001 $left = 'keyword' if $left =~ '^kw$';
2002 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2003 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2004 my $dbh = C4::Context->dbh;
2005 if ( $operator && $left ne 'keyword' ) {
2006 #do a specific search
2007 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2008 my $sth = $dbh->prepare(
2009 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2011 warn "$left / $operator / $right\n" if $DEBUG;
2013 # split each word, query the DB and build the biblionumbers result
2014 #sanitizing leftpart
2015 $left =~ s/^\s+|\s+$//;
2016 foreach ( split / /, $right ) {
2018 $_ =~ s/^\s+|\s+$//;
2020 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2021 $sth->execute( $server, $left, $_ )
2022 or warn "execute failed: $!";
2023 while ( my ( $line, $value ) = $sth->fetchrow ) {
2025 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2026 # otherwise, fill the result
2027 $biblionumbers .= $line
2028 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2029 warn "result : $value "
2030 . ( $right =~ /\d/ ) . "=="
2031 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2034 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2036 warn "NZAND" if $DEBUG;
2037 $results = NZoperatorAND($biblionumbers,$results);
2039 $results = $biblionumbers;
2044 #do a complete search (all indexes), if index='kw' do complete search too.
2045 my $sth = $dbh->prepare(
2046 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2049 # split each word, query the DB and build the biblionumbers result
2050 foreach ( split / /, $string ) {
2051 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2052 warn "search on all indexes on $_" if $DEBUG;
2055 $sth->execute( $server, $_ );
2056 while ( my $line = $sth->fetchrow ) {
2057 $biblionumbers .= $line;
2060 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2062 $results = NZoperatorAND($biblionumbers,$results);
2065 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2066 $results = $biblionumbers;
2070 warn "return : $results for LEAF : $string" if $DEBUG;
2073 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2077 my ($rightresult, $leftresult)=@_;
2079 my @leftresult = split /;/, $leftresult;
2080 warn " @leftresult / $rightresult \n" if $DEBUG;
2082 # my @rightresult = split /;/,$leftresult;
2085 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2086 # the result is stored twice, to have the same weight for AND than OR.
2087 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2088 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2089 foreach (@leftresult) {
2092 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2093 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2094 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2096 "$value-$countvalue;$value-$countvalue;";
2099 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2100 return $finalresult;
2104 my ($rightresult, $leftresult)=@_;
2105 return $rightresult.$leftresult;
2109 my ($leftresult, $rightresult)=@_;
2111 my @leftresult = split /;/, $leftresult;
2113 # my @rightresult = split /;/,$leftresult;
2115 foreach (@leftresult) {
2117 $value=$1 if $value=~m/(.*)-\d+$/;
2118 unless ($rightresult =~ "$value-") {
2119 $finalresult .= "$_;";
2122 return $finalresult;
2127 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2134 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2135 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2137 # order title asc by default
2138 # $ordering = '1=36 <i' unless $ordering;
2139 $results_per_page = 20 unless $results_per_page;
2140 $offset = 0 unless $offset;
2141 my $dbh = C4::Context->dbh;
2144 # order by POPULARITY
2146 if ( $ordering =~ /popularity/ ) {
2150 # popularity is not in MARC record, it's builded from a specific query
2152 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2153 foreach ( split /;/, $biblionumbers ) {
2154 my ( $biblionumber, $title ) = split /,/, $_;
2155 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2156 $sth->execute($biblionumber);
2157 my $popularity = $sth->fetchrow || 0;
2159 # hint : the key is popularity.title because we can have
2160 # many results with the same popularity. In this case, sub-ordering is done by title
2161 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2162 # (un-frequent, I agree, but we won't forget anything that way ;-)
2163 $popularity{ sprintf( "%10d", $popularity ) . $title
2164 . $biblionumber } = $biblionumber;
2167 # sort the hash and return the same structure as GetRecords (Zebra querying)
2170 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2171 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2172 $result_hash->{'RECORDS'}[ $numbers++ ] =
2173 $result{ $popularity{$key} }->as_usmarc();
2176 else { # sort popularity ASC
2177 foreach my $key ( sort ( keys %popularity ) ) {
2178 $result_hash->{'RECORDS'}[ $numbers++ ] =
2179 $result{ $popularity{$key} }->as_usmarc();
2182 my $finalresult = ();
2183 $result_hash->{'hits'} = $numbers;
2184 $finalresult->{'biblioserver'} = $result_hash;
2185 return $finalresult;
2191 elsif ( $ordering =~ /author/ ) {
2193 foreach ( split /;/, $biblionumbers ) {
2194 my ( $biblionumber, $title ) = split /,/, $_;
2195 my $record = GetMarcBiblio($biblionumber);
2197 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2198 $author = $record->subfield( '200', 'f' );
2199 $author = $record->subfield( '700', 'a' ) unless $author;
2202 $author = $record->subfield( '100', 'a' );
2205 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2206 # and we don't want to get only 1 result for each of them !!!
2207 $result{ $author . $biblionumber } = $record;
2210 # sort the hash and return the same structure as GetRecords (Zebra querying)
2213 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2214 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2215 $result_hash->{'RECORDS'}[ $numbers++ ] =
2216 $result{$key}->as_usmarc();
2219 else { # sort by author ASC
2220 foreach my $key ( sort ( keys %result ) ) {
2221 $result_hash->{'RECORDS'}[ $numbers++ ] =
2222 $result{$key}->as_usmarc();
2225 my $finalresult = ();
2226 $result_hash->{'hits'} = $numbers;
2227 $finalresult->{'biblioserver'} = $result_hash;
2228 return $finalresult;
2231 # ORDER BY callnumber
2234 elsif ( $ordering =~ /callnumber/ ) {
2236 foreach ( split /;/, $biblionumbers ) {
2237 my ( $biblionumber, $title ) = split /,/, $_;
2238 my $record = GetMarcBiblio($biblionumber);
2240 my $frameworkcode = GetFrameworkCode($biblionumber);
2241 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2242 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2243 unless $callnumber_tag;
2244 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2245 $callnumber = $record->subfield( '200', 'f' );
2247 $callnumber = $record->subfield( '100', 'a' );
2250 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2251 # and we don't want to get only 1 result for each of them !!!
2252 $result{ $callnumber . $biblionumber } = $record;
2255 # sort the hash and return the same structure as GetRecords (Zebra querying)
2258 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2259 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2260 $result_hash->{'RECORDS'}[ $numbers++ ] =
2261 $result{$key}->as_usmarc();
2264 else { # sort by title ASC
2265 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2266 $result_hash->{'RECORDS'}[ $numbers++ ] =
2267 $result{$key}->as_usmarc();
2270 my $finalresult = ();
2271 $result_hash->{'hits'} = $numbers;
2272 $finalresult->{'biblioserver'} = $result_hash;
2273 return $finalresult;
2275 elsif ( $ordering =~ /pubdate/ ) { #pub year
2277 foreach ( split /;/, $biblionumbers ) {
2278 my ( $biblionumber, $title ) = split /,/, $_;
2279 my $record = GetMarcBiblio($biblionumber);
2280 my ( $publicationyear_tag, $publicationyear_subfield ) =
2281 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2282 my $publicationyear =
2283 $record->subfield( $publicationyear_tag,
2284 $publicationyear_subfield );
2286 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2287 # and we don't want to get only 1 result for each of them !!!
2288 $result{ $publicationyear . $biblionumber } = $record;
2291 # sort the hash and return the same structure as GetRecords (Zebra querying)
2294 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2295 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2296 $result_hash->{'RECORDS'}[ $numbers++ ] =
2297 $result{$key}->as_usmarc();
2300 else { # sort by pub year ASC
2301 foreach my $key ( sort ( keys %result ) ) {
2302 $result_hash->{'RECORDS'}[ $numbers++ ] =
2303 $result{$key}->as_usmarc();
2306 my $finalresult = ();
2307 $result_hash->{'hits'} = $numbers;
2308 $finalresult->{'biblioserver'} = $result_hash;
2309 return $finalresult;
2315 elsif ( $ordering =~ /title/ ) {
2317 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2319 foreach ( split /;/, $biblionumbers ) {
2320 my ( $biblionumber, $title ) = split /,/, $_;
2322 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2323 # and we don't want to get only 1 result for each of them !!!
2324 # hint & speed improvement : we can order without reading the record
2325 # so order, and read records only for the requested page !
2326 $result{ $title . $biblionumber } = $biblionumber;
2329 # sort the hash and return the same structure as GetRecords (Zebra querying)
2332 if ( $ordering eq 'title_az' ) { # sort by title desc
2333 foreach my $key ( sort ( keys %result ) ) {
2334 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2337 else { # sort by title ASC
2338 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2339 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2343 # limit the $results_per_page to result size if it's more
2344 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2346 # for the requested page, replace biblionumber by the complete record
2347 # speed improvement : avoid reading too much things
2349 my $counter = $offset ;
2350 $counter <= $offset + $results_per_page ;
2354 $result_hash->{'RECORDS'}[$counter] =
2355 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2357 my $finalresult = ();
2358 $result_hash->{'hits'} = $numbers;
2359 $finalresult->{'biblioserver'} = $result_hash;
2360 return $finalresult;
2367 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2370 foreach ( split /;/, $biblionumbers ) {
2371 my ( $biblionumber, $title ) = split /,/, $_;
2372 $title =~ /(.*)-(\d)/;
2377 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2378 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2379 # biblio N has ranking = 6
2380 $count_ranking{$biblionumber} += $ranking;
2383 # build the result by "inverting" the count_ranking hash
2384 # 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
2386 foreach ( keys %count_ranking ) {
2387 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2390 # sort the hash and return the same structure as GetRecords (Zebra querying)
2393 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2394 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2397 # limit the $results_per_page to result size if it's more
2398 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2400 # for the requested page, replace biblionumber by the complete record
2401 # speed improvement : avoid reading too much things
2403 my $counter = $offset ;
2404 $counter <= $offset + $results_per_page ;
2408 $result_hash->{'RECORDS'}[$counter] =
2409 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2410 if $result_hash->{'RECORDS'}[$counter];
2412 my $finalresult = ();
2413 $result_hash->{'hits'} = $numbers;
2414 $finalresult->{'biblioserver'} = $result_hash;
2415 return $finalresult;
2419 =head2 enabled_staff_search_views
2421 %hash = enabled_staff_search_views()
2423 This function returns a hash that contains three flags obtained from the system
2424 preferences, used to determine whether a particular staff search results view
2429 =item C<Output arg:>
2431 * $hash{can_view_MARC} is true only if the MARC view is enabled
2432 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2433 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2435 =item C<usage in the script:>
2439 $template->param ( C4::Search::enabled_staff_search_views );
2443 sub enabled_staff_search_views
2446 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2447 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2448 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2452 sub AddSearchHistory{
2453 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2454 my $dbh = C4::Context->dbh;
2456 # Add the request the user just made
2457 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2458 my $sth = $dbh->prepare($sql);
2459 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2460 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2463 sub GetSearchHistory{
2464 my ($borrowernumber,$session)=@_;
2465 my $dbh = C4::Context->dbh;
2467 # Add the request the user just made
2468 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2469 my $sth = $dbh->prepare($query);
2470 $sth->execute($borrowernumber, $session);
2471 return $sth->fetchall_hashref({});
2474 =head2 z3950_search_args
2476 $arrayref = z3950_search_args($matchpoints)
2478 This function returns an array reference that contains the search parameters to be
2479 passed to the Z39.50 search script (z3950_search.pl). The array elements
2480 are hash refs whose keys are name, value and encvalue, and whose values are the
2481 name of a search parameter, the value of that search parameter and the URL encoded
2482 value of that parameter.
2484 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2486 The search parameter values are obtained from the bibliographic record whose
2487 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2489 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2490 a general purpose search argument. In this case, the returned array contains only
2491 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2493 If a search parameter value is undefined or empty, it is not included in the returned
2496 The returned array reference may be passed directly to the template parameters.
2500 =item C<Output arg:>
2502 * $array containing hash refs as described above
2504 =item C<usage in the script:>
2508 $data = Biblio::GetBiblioData($bibno);
2509 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2513 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2517 sub z3950_search_args {
2519 $bibrec = { title => $bibrec } if !ref $bibrec;
2521 for my $field (qw/ lccn isbn issn title author dewey subject /)
2523 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2524 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2529 =head2 BiblioAddAuthorities
2531 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2533 this function finds the authorities linked to the biblio
2534 * search in the authority DB for the same authid (in $9 of the biblio)
2535 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2536 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2537 OR adds a new authority record
2543 * $record is the MARC record in question (marc blob)
2544 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2546 =item C<Output arg:>
2548 * $countlinked is the number of authorities records that are linked to this authority
2552 * 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)
2559 sub BiblioAddAuthorities{
2560 my ( $record, $frameworkcode ) = @_;
2561 my $dbh=C4::Context->dbh;
2562 my $query=$dbh->prepare(qq|
2563 SELECT authtypecode,tagfield
2564 FROM marc_subfield_structure
2565 WHERE frameworkcode=?
2566 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2567 # SELECT authtypecode,tagfield
2568 # FROM marc_subfield_structure
2569 # WHERE frameworkcode=?
2570 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2571 $query->execute($frameworkcode);
2572 my ($countcreated,$countlinked);
2573 while (my $data=$query->fetchrow_hashref){
2574 foreach my $field ($record->field($data->{tagfield})){
2575 next if ($field->subfield('3')||$field->subfield('9'));
2576 # No authorities id in the tag.
2577 # Search if there is any authorities to link to.
2578 my $query='at='.$data->{authtypecode}.' ';
2579 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2580 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2581 # there is only 1 result
2583 warn "BIBLIOADDSAUTHORITIES: $error";
2586 if ($results && scalar(@$results)==1) {
2587 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2588 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2590 } elsif (scalar(@$results)>1) {
2591 #More than One result
2592 #This can comes out of a lack of a subfield.
2593 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2594 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2597 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2598 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2599 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2600 my $authtypedata=C4::AuthoritiesMarc::GetAuthType($data->{authtypecode});
2601 next unless $authtypedata;
2602 my $marcrecordauth=MARC::Record->new();
2603 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2604 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2605 $marcrecordauth->insert_fields_ordered($authfield);
2607 # bug 2317: ensure new authority knows it's using UTF-8; currently
2608 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2609 # automatically for UNIMARC (by not transcoding)
2610 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2611 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2612 # of change to a core API just before the 3.0 release.
2613 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2614 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2617 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2619 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2621 $field->add_subfields('9'=>$authid);
2625 return ($countlinked,$countcreated);
2628 =head2 GetDistinctValues($field);
2630 C<$field> is a reference to the fields array
2634 sub GetDistinctValues {
2635 my ($fieldname,$string)=@_;
2636 # returns a reference to a hash of references to branches...
2637 if ($fieldname=~/\./){
2638 my ($table,$column)=split /\./, $fieldname;
2639 my $dbh = C4::Context->dbh;
2640 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2641 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 ");
2643 my $elements=$sth->fetchall_arrayref({});
2648 my @servers=qw<biblioserver authorityserver>;
2649 my (@zconns,@results);
2650 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2651 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2654 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2657 # The big moment: asynchronously retrieve results from all servers
2659 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2660 my $ev = $zconns[ $i - 1 ]->last_event();
2661 if ( $ev == ZOOM::Event::ZEND ) {
2662 next unless $results[ $i - 1 ];
2663 my $size = $results[ $i - 1 ]->size();
2665 for (my $j=0;$j<$size;$j++){
2667 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2668 push @elements, \%hashscan;
2678 END { } # module clean-up code here (global destructor)
2685 Koha Development Team <http://koha-community.org/>