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" ) {
354 $sort_by .= "1=1003 <i ";
356 elsif ( $sort eq "author_za" ) {
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=20 <i ";
368 elsif ( $sort eq "call_number_dsc" ) {
369 $sort_by .= "1=20 >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" ) {
384 $sort_by .= "1=4 <i ";
386 elsif ( $sort eq "title_za" ) {
387 $sort_by .= "1=4 >i ";
390 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
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',
869 'Local-classification',
872 'Match-heading-see-from',
878 'Name-geographic-heading',
879 'Name-geographic-see',
880 'Name-geographic-seealso',
887 'Personal-name-heading',
889 'Personal-name-seealso',
905 'Subject-heading-thesaurus',
906 'Subject-name-personal',
907 'Subject-subdivision',
916 'Term-genre-form-heading',
917 'Term-genre-form-see',
918 'Term-genre-form-seealso',
924 'Title-uniform-heading',
926 'Title-uniform-seealso',
936 'classification-source',
938 'coded-location-qualifier',
949 'Local-classification',
952 'materials-specified',
961 'replacementpricedate',
977 $simple_query, $query_cgi,
979 $limit_cgi, $limit_desc,
980 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
982 Build queries and limits in CCL, CGI, Human,
983 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
985 See verbose embedded documentation.
991 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
993 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
996 my @operators = $operators ? @$operators : ();
997 my @indexes = $indexes ? @$indexes : ();
998 my @operands = $operands ? @$operands : ();
999 my @limits = $limits ? @$limits : ();
1000 my @sort_by = $sort_by ? @$sort_by : ();
1002 my $stemming = C4::Context->preference("QueryStemming") || 0;
1003 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1004 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1005 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1006 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1008 # no stemming/weight/fuzzy in NoZebra
1009 if ( C4::Context->preference("NoZebra") ) {
1013 $auto_truncation = 0;
1016 my $query = $operands[0];
1017 my $simple_query = $operands[0];
1019 # initialize the variables we're passing back
1028 my $stopwords_removed; # flag to determine if stopwords have been removed
1031 my $cclindexes = getIndexes();
1032 if( $query !~ /\s*ccl=/ ){
1033 for my $index (@$cclindexes){
1034 if($query =~ /($index)(,?\w)*[:=]/){
1038 $query = "ccl=$query" if($cclq);
1041 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1043 if ( $query =~ /^ccl=/ ) {
1044 return ( undef, $', $', "q=ccl=$'", $', '', '', '', '', 'ccl' );
1046 if ( $query =~ /^cql=/ ) {
1047 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1049 if ( $query =~ /^pqf=/ ) {
1050 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1053 # pass nested queries directly
1054 # FIXME: need better handling of some of these variables in this case
1055 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1056 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1057 # if ( $query =~ /(\(|\))/ ) {
1059 # undef, $query, $simple_query, $query_cgi,
1060 # $query, $limit, $limit_cgi, $limit_desc,
1061 # $stopwords_removed, 'ccl'
1065 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1066 # query operands and indexes and add stemming, truncation, field weighting, etc.
1067 # Once we do so, we'll end up with a value in $query, just like if we had an
1068 # incoming $query from the user
1071 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1072 my $previous_operand
1073 ; # a flag used to keep track if there was a previous query
1074 # if there was, we can apply the current operator
1076 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1078 # COMBINE OPERANDS, INDEXES AND OPERATORS
1079 if ( $operands[$i] ) {
1080 $operands[$i]=~s/^\s+//;
1082 # A flag to determine whether or not to add the index to the query
1085 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1086 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1089 $remove_stopwords = 0;
1091 my $operand = $operands[$i];
1092 my $index = $indexes[$i];
1094 # Add index-specific attributes
1095 # Date of Publication
1096 if ( $index eq 'yr' ) {
1097 $index .= ",st-numeric";
1099 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1102 # Date of Acquisition
1103 elsif ( $index eq 'acqdate' ) {
1104 $index .= ",st-date-normalized";
1106 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1108 # ISBN,ISSN,Standard Number, don't need special treatment
1109 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1112 $stemming, $auto_truncation,
1113 $weight_fields, $fuzzy_enabled,
1115 ) = ( 0, 0, 0, 0, 0 );
1123 # Set default structure attribute (word list)
1124 my $struct_attr = q{};
1125 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1126 $struct_attr = ",wrdl";
1129 # Some helpful index variants
1130 my $index_plus = $index . $struct_attr . ':';
1131 my $index_plus_comma = $index . $struct_attr . ',';
1134 if ($remove_stopwords) {
1135 ( $operand, $stopwords_removed ) =
1136 _remove_stopwords( $operand, $index );
1137 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1138 warn "REMOVED STOPWORDS: @$stopwords_removed"
1139 if ( $stopwords_removed && $DEBUG );
1142 if ($auto_truncation){
1143 unless ( $index =~ /(st-|phr|ext)/ ) {
1144 #FIXME only valid with LTR scripts
1145 $operand=join(" ",map{
1146 (index($_,"*")>0?"$_":"$_*")
1147 }split (/\s+/,$operand));
1148 warn $operand if $DEBUG;
1153 my $truncated_operand;
1154 my( $nontruncated, $righttruncated, $lefttruncated,
1155 $rightlefttruncated, $regexpr
1156 ) = _detect_truncation( $operand, $index );
1158 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1163 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1164 scalar(@$rightlefttruncated) > 0 )
1167 # Don't field weight or add the index to the query, we do it here
1169 undef $weight_fields;
1170 my $previous_truncation_operand;
1171 if (scalar @$nontruncated) {
1172 $truncated_operand .= "$index_plus @$nontruncated ";
1173 $previous_truncation_operand = 1;
1175 if (scalar @$righttruncated) {
1176 $truncated_operand .= "and " if $previous_truncation_operand;
1177 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1178 $previous_truncation_operand = 1;
1180 if (scalar @$lefttruncated) {
1181 $truncated_operand .= "and " if $previous_truncation_operand;
1182 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1183 $previous_truncation_operand = 1;
1185 if (scalar @$rightlefttruncated) {
1186 $truncated_operand .= "and " if $previous_truncation_operand;
1187 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1188 $previous_truncation_operand = 1;
1191 $operand = $truncated_operand if $truncated_operand;
1192 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1195 my $stemmed_operand;
1196 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1199 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1201 # Handle Field Weighting
1202 my $weighted_operand;
1203 if ($weight_fields) {
1204 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1205 $operand = $weighted_operand;
1209 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1211 # If there's a previous operand, we need to add an operator
1212 if ($previous_operand) {
1214 # User-specified operator
1215 if ( $operators[ $i - 1 ] ) {
1216 $query .= " $operators[$i-1] ";
1217 $query .= " $index_plus " unless $indexes_set;
1218 $query .= " $operand";
1219 $query_cgi .= "&op=$operators[$i-1]";
1220 $query_cgi .= "&idx=$index" if $index;
1221 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1223 " $operators[$i-1] $index_plus $operands[$i]";
1226 # Default operator is and
1229 $query .= "$index_plus " unless $indexes_set;
1230 $query .= "$operand";
1231 $query_cgi .= "&op=and&idx=$index" if $index;
1232 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1233 $query_desc .= " and $index_plus $operands[$i]";
1237 # There isn't a pervious operand, don't need an operator
1240 # Field-weighted queries already have indexes set
1241 $query .= " $index_plus " unless $indexes_set;
1243 $query_desc .= " $index_plus $operands[$i]";
1244 $query_cgi .= "&idx=$index" if $index;
1245 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1246 $previous_operand = 1;
1251 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1254 my $group_OR_limits;
1255 my $availability_limit;
1256 foreach my $this_limit (@limits) {
1257 # if ( $this_limit =~ /available/ ) {
1259 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1261 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1262 # $availability_limit .=
1263 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1264 # $limit_cgi .= "&limit=available";
1265 # $limit_desc .= "";
1268 # group_OR_limits, prefixed by mc-
1269 # OR every member of the group
1270 # elsif ( $this_limit =~ /mc/ ) {
1271 if ( $this_limit =~ /mc/ ) {
1272 $group_OR_limits .= " or " if $group_OR_limits;
1273 $limit_desc .= " or " if $group_OR_limits;
1274 $group_OR_limits .= "$this_limit";
1275 $limit_cgi .= "&limit=$this_limit";
1276 $limit_desc .= " $this_limit";
1279 # Regular old limits
1281 $limit .= " and " if $limit || $query;
1282 $limit .= "$this_limit";
1283 $limit_cgi .= "&limit=$this_limit";
1284 if ($this_limit =~ /^branch:(.+)/) {
1285 my $branchcode = $1;
1286 my $branchname = GetBranchName($branchcode);
1287 if (defined $branchname) {
1288 $limit_desc .= " branch:$branchname";
1290 $limit_desc .= " $this_limit";
1293 $limit_desc .= " $this_limit";
1297 if ($group_OR_limits) {
1298 $limit .= " and " if ( $query || $limit );
1299 $limit .= "($group_OR_limits)";
1301 if ($availability_limit) {
1302 $limit .= " and " if ( $query || $limit );
1303 $limit .= "($availability_limit)";
1306 # Normalize the query and limit strings
1307 # This is flawed , means we can't search anything with : in it
1308 # if user wants to do ccl or cql, start the query with that
1309 # $query =~ s/:/=/g;
1310 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc)):/=/g;
1311 $query =~ s/(?<=rtrn):/=/g;
1313 for ( $query, $query_desc, $limit, $limit_desc ) {
1314 s/ / /g; # remove extra spaces
1315 s/^ //g; # remove any beginning spaces
1316 s/ $//g; # remove any ending spaces
1317 s/==/=/g; # remove double == from query
1319 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1321 for ($query_cgi,$simple_query) {
1324 # append the limit to the query
1325 $query .= " " . $limit;
1329 warn "QUERY:" . $query;
1330 warn "QUERY CGI:" . $query_cgi;
1331 warn "QUERY DESC:" . $query_desc;
1332 warn "LIMIT:" . $limit;
1333 warn "LIMIT CGI:" . $limit_cgi;
1334 warn "LIMIT DESC:" . $limit_desc;
1335 warn "---------\nLeave buildQuery\n---------";
1338 undef, $query, $simple_query, $query_cgi,
1339 $query_desc, $limit, $limit_cgi, $limit_desc,
1340 $stopwords_removed, $query_type
1344 =head2 searchResults
1346 my @search_results = searchResults($search_context, $searchdesc, $hits,
1347 $results_per_page, $offset, $scan,
1348 @marcresults, $hidelostitems);
1350 Format results in a form suitable for passing to the template
1354 # IMO this subroutine is pretty messy still -- it's responsible for
1355 # building the HTML output for the template
1357 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1358 my $dbh = C4::Context->dbh;
1361 $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1363 #Build branchnames hash
1365 #get branch information.....
1367 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1369 while ( my $bdata = $bsth->fetchrow_hashref ) {
1370 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1372 # FIXME - We build an authorised values hash here, using the default framework
1373 # though it is possible to have different authvals for different fws.
1375 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1377 # get notforloan authorised value list (see $shelflocations FIXME)
1378 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1380 #Build itemtype hash
1381 #find itemtype & itemtype image
1385 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1388 while ( my $bdata = $bsth->fetchrow_hashref ) {
1389 foreach (qw(description imageurl summary notforloan)) {
1390 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1394 #search item field code
1397 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1400 my ($itemtag) = $sth->fetchrow;
1402 ## find column names of items related to MARC
1403 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1405 my %subfieldstosearch;
1406 while ( ( my $column ) = $sth2->fetchrow ) {
1407 my ( $tagfield, $tagsubfield ) =
1408 &GetMarcFromKohaField( "items." . $column, "" );
1409 $subfieldstosearch{$column} = $tagsubfield;
1412 # handle which records to actually retrieve
1414 if ( $hits && $offset + $results_per_page <= $hits ) {
1415 $times = $offset + $results_per_page;
1418 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1421 my $marcflavour = C4::Context->preference("marcflavour");
1422 # We get the biblionumber position in MARC
1423 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1426 # loop through all of the records we've retrieved
1427 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1428 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1430 $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1432 $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1435 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1436 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1437 $oldbiblio->{result_number} = $i + 1;
1439 # add imageurl to itemtype if there is one
1440 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1442 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1443 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1444 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1445 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1446 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1447 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1449 # edition information, if any
1450 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1451 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1452 # Build summary if there is one (the summary is defined in the itemtypes table)
1453 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1454 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1455 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1456 my @fields = $marcrecord->fields();
1459 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1461 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1462 $tag =~ /(.{3})(.)/;
1463 if($marcrecord->field($1)){
1464 my @abc = $marcrecord->field($1)->subfield($2);
1465 $tags->{$tag} = $#abc + 1 ;
1469 # We catch how many times to repeat this line
1471 foreach my $tag (keys(%$tags)){
1472 $max = $tags->{$tag} if($tags->{$tag} > $max);
1475 # we replace, and repeat each line
1476 for (my $i = 0 ; $i < $max ; $i++){
1477 my $newline = $line;
1479 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1480 $tag =~ /(.{3})(.)/;
1482 if($marcrecord->field($1)){
1483 my @repl = $marcrecord->field($1)->subfield($2);
1484 my $subfieldvalue = $repl[$i];
1486 if (! utf8::is_utf8($subfieldvalue)) {
1487 utf8::decode($subfieldvalue);
1490 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1493 $newsummary .= "$newline\n";
1497 $newsummary =~ s/\[(.*?)]//g;
1498 $newsummary =~ s/\n/<br\/>/g;
1499 $oldbiblio->{summary} = $newsummary;
1502 # Pull out the items fields
1503 my @fields = $marcrecord->field($itemtag);
1505 # Setting item statuses for display
1506 my @available_items_loop;
1507 my @onloan_items_loop;
1508 my @other_items_loop;
1510 my $available_items;
1514 my $ordered_count = 0;
1515 my $available_count = 0;
1516 my $onloan_count = 0;
1517 my $longoverdue_count = 0;
1518 my $other_count = 0;
1519 my $wthdrawn_count = 0;
1520 my $itemlost_count = 0;
1521 my $itembinding_count = 0;
1522 my $itemdamaged_count = 0;
1523 my $item_in_transit_count = 0;
1524 my $can_place_holds = 0;
1525 my $item_onhold_count = 0;
1526 my $items_count = scalar(@fields);
1528 ( C4::Context->preference('maxItemsinSearchResults') )
1529 ? C4::Context->preference('maxItemsinSearchResults') - 1
1532 # loop through every item
1533 foreach my $field (@fields) {
1536 # populate the items hash
1537 foreach my $code ( keys %subfieldstosearch ) {
1538 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1541 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1542 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1543 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1544 if ($item->{$hbranch}) {
1545 $item->{'branchname'} = $branches{$item->{$hbranch}};
1547 elsif ($item->{$otherbranch}) { # Last resort
1548 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1551 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1552 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1553 if ( $item->{onloan} ) {
1555 my $key = $prefix . $item->{onloan} . $item->{barcode};
1556 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1557 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1558 $onloan_items->{$key}->{branchname} = $item->{branchname};
1559 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1560 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1561 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1562 # if something's checked out and lost, mark it as 'long overdue'
1563 if ( $item->{itemlost} ) {
1564 $onloan_items->{$prefix}->{longoverdue}++;
1565 $longoverdue_count++;
1566 } else { # can place holds as long as item isn't lost
1567 $can_place_holds = 1;
1571 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1575 if ( $item->{notforloan} == -1 ) {
1579 # is item in transit?
1580 my $transfertwhen = '';
1581 my ($transfertfrom, $transfertto);
1583 # is item on the reserve shelf?
1584 my $reservestatus = 0;
1587 unless ($item->{wthdrawn}
1588 || $item->{itemlost}
1590 || $item->{notforloan}
1591 || $items_count > 20) {
1593 # A couple heuristics to limit how many times
1594 # we query the database for item transfer information, sacrificing
1595 # accuracy in some cases for speed;
1597 # 1. don't query if item has one of the other statuses
1598 # 2. don't check transit status if the bib has
1599 # more than 20 items
1601 # FIXME: to avoid having the query the database like this, and to make
1602 # the in transit status count as unavailable for search limiting,
1603 # should map transit status to record indexed in Zebra.
1605 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1606 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1609 # item is withdrawn, lost or damaged
1610 if ( $item->{wthdrawn}
1611 || $item->{itemlost}
1613 || $item->{notforloan}
1614 || $reservestatus eq 'Waiting'
1615 || ($transfertwhen ne ''))
1617 $wthdrawn_count++ if $item->{wthdrawn};
1618 $itemlost_count++ if $item->{itemlost};
1619 $itemdamaged_count++ if $item->{damaged};
1620 $item_in_transit_count++ if $transfertwhen ne '';
1621 $item_onhold_count++ if $reservestatus eq 'Waiting';
1622 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1625 my $key = $prefix . $item->{status};
1626 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1627 $other_items->{$key}->{$_} = $item->{$_};
1629 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1630 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1631 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1632 $other_items->{$key}->{count}++ if $item->{$hbranch};
1633 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1634 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1638 $can_place_holds = 1;
1640 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1641 foreach (qw(branchname itemcallnumber)) {
1642 $available_items->{$prefix}->{$_} = $item->{$_};
1644 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1645 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1648 } # notforloan, item level and biblioitem level
1649 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1651 ( C4::Context->preference('maxItemsinSearchResults') )
1652 ? C4::Context->preference('maxItemsinSearchResults') - 1
1654 for my $key ( sort keys %$onloan_items ) {
1655 (++$onloanitemscount > $maxitems) and last;
1656 push @onloan_items_loop, $onloan_items->{$key};
1658 for my $key ( sort keys %$other_items ) {
1659 (++$otheritemscount > $maxitems) and last;
1660 push @other_items_loop, $other_items->{$key};
1662 for my $key ( sort keys %$available_items ) {
1663 (++$availableitemscount > $maxitems) and last;
1664 push @available_items_loop, $available_items->{$key}
1667 # XSLT processing of some stuff
1669 SetUTF8Flag($marcrecord);
1670 $debug && warn $marcrecord->as_formatted;
1671 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1672 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1673 # is not implemented yet
1674 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1678 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1679 $can_place_holds = 0
1680 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1681 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1682 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1683 $oldbiblio->{items_count} = $items_count;
1684 $oldbiblio->{available_items_loop} = \@available_items_loop;
1685 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1686 $oldbiblio->{other_items_loop} = \@other_items_loop;
1687 $oldbiblio->{availablecount} = $available_count;
1688 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1689 $oldbiblio->{onloancount} = $onloan_count;
1690 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1691 $oldbiblio->{othercount} = $other_count;
1692 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1693 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1694 $oldbiblio->{itemlostcount} = $itemlost_count;
1695 $oldbiblio->{damagedcount} = $itemdamaged_count;
1696 $oldbiblio->{intransitcount} = $item_in_transit_count;
1697 $oldbiblio->{onholdcount} = $item_onhold_count;
1698 $oldbiblio->{orderedcount} = $ordered_count;
1699 $oldbiblio->{isbn} =~
1700 s/-//g; # deleting - in isbn to enable amazon content
1701 push( @newresults, $oldbiblio )
1702 if(not $hidelostitems
1703 or (($items_count > $itemlost_count )
1704 && $hidelostitems));
1710 =head2 SearchAcquisitions
1711 Search for acquisitions
1714 sub SearchAcquisitions{
1715 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1717 my $dbh=C4::Context->dbh;
1718 # Variable initialization
1722 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1723 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1724 WHERE dateaccessioned BETWEEN ? AND ?
1727 my (@params,@loopcriteria);
1729 push @params, $datebegin->output("iso");
1730 push @params, $dateend->output("iso");
1732 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1733 if(C4::Context->preference("item-level_itypes")){
1734 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1736 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1738 push @params, @$itemtypes;
1741 if ($criteria =~/itemtype/){
1742 if(C4::Context->preference("item-level_itypes")){
1743 $str .= "AND items.itype=? ";
1745 $str .= "AND biblioitems.itemtype=? ";
1748 if(scalar(@$itemtypes) == 0){
1749 my $itypes = GetItemTypes();
1750 for my $key (keys %$itypes){
1751 push @$itemtypes, $key;
1755 @loopcriteria= @$itemtypes;
1756 }elsif ($criteria=~/itemcallnumber/){
1757 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1758 OR items.itemcallnumber is NULL
1759 OR items.itemcallnumber = '')";
1761 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1763 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1764 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1767 if ($orderby =~ /date_desc/){
1768 $str.=" ORDER BY dateaccessioned DESC";
1770 $str.=" ORDER BY title";
1773 my $qdataacquisitions=$dbh->prepare($str);
1775 my @loopacquisitions;
1776 foreach my $value(@loopcriteria){
1777 push @params,$value;
1779 $cell{"title"}=$value;
1780 $cell{"titlecode"}=$value;
1782 eval{$qdataacquisitions->execute(@params);};
1784 if ($@){ warn "recentacquisitions Error :$@";}
1787 while (my $data=$qdataacquisitions->fetchrow_hashref){
1788 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1790 $cell{"loopdata"}=\@loopdata;
1792 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1795 $qdataacquisitions->finish;
1796 return \@loopacquisitions;
1798 #----------------------------------------------------------------------
1800 # Non-Zebra GetRecords#
1801 #----------------------------------------------------------------------
1805 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1811 $query, $simple_query, $sort_by_ref, $servers_ref,
1812 $results_per_page, $offset, $expanded_facet, $branches,
1815 warn "query =$query" if $DEBUG;
1816 my $result = NZanalyse($query);
1817 warn "results =$result" if $DEBUG;
1819 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1825 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1826 the list is built from an inverted index in the nozebra SQL table
1827 note that title is here only for convenience : the sorting will be very fast when requested on title
1828 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1833 my ( $string, $server ) = @_;
1834 # warn "---------" if $DEBUG;
1835 warn " NZanalyse" if $DEBUG;
1836 # warn "---------" if $DEBUG;
1838 # $server contains biblioserver or authorities, depending on what we search on.
1839 #warn "querying : $string on $server";
1840 $server = 'biblioserver' unless $server;
1842 # if we have a ", replace the content to discard temporarily any and/or/not inside
1844 if ( $string =~ /"/ ) {
1845 $string =~ s/"(.*?)"/__X__/;
1847 warn "commacontent : $commacontent" if $DEBUG;
1850 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1851 # then, call again NZanalyse with $left and $right
1852 # (recursive until we find a leaf (=> something without and/or/not)
1853 # delete repeated operator... Would then go in infinite loop
1854 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1857 #process parenthesis before.
1858 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1861 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1863 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1865 my $leftresult = NZanalyse( $left, $server );
1867 my $rightresult = NZanalyse( $right, $server );
1869 # OK, we have the results for right and left part of the query
1870 # depending of operand, intersect, union or exclude both lists
1871 # to get a result list
1872 if ( $operator eq ' and ' ) {
1873 return NZoperatorAND($leftresult,$rightresult);
1875 elsif ( $operator eq ' or ' ) {
1877 # just merge the 2 strings
1878 return $leftresult . $rightresult;
1880 elsif ( $operator eq ' not ' ) {
1881 return NZoperatorNOT($leftresult,$rightresult);
1885 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1889 warn "string :" . $string if $DEBUG;
1893 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1896 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1898 warn "no parenthesis. left : $left operator: $operator right: $right"
1901 # it's not a leaf, we have a and/or/not
1904 # reintroduce comma content if needed
1905 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1906 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1907 warn "node : $left / $operator / $right\n" if $DEBUG;
1908 my $leftresult = NZanalyse( $left, $server );
1909 my $rightresult = NZanalyse( $right, $server );
1910 warn " leftresult : $leftresult" if $DEBUG;
1911 warn " rightresult : $rightresult" if $DEBUG;
1912 # OK, we have the results for right and left part of the query
1913 # depending of operand, intersect, union or exclude both lists
1914 # to get a result list
1915 if ( $operator eq ' and ' ) {
1917 return NZoperatorAND($leftresult,$rightresult);
1919 elsif ( $operator eq ' or ' ) {
1921 # just merge the 2 strings
1922 return $leftresult . $rightresult;
1924 elsif ( $operator eq ' not ' ) {
1925 return NZoperatorNOT($leftresult,$rightresult);
1929 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1930 die "error : operand unknown : $operator for $string";
1933 # it's a leaf, do the real SQL query and return the result
1936 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1937 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1938 #remove trailing blank at the beginning
1940 warn "leaf:$string" if $DEBUG;
1942 # parse the string in in operator/operand/value again
1946 if ($string =~ /(.*)(>=|<=)(.*)/) {
1953 # warn "handling leaf... left:$left operator:$operator right:$right"
1955 unless ($operator) {
1956 if ($string =~ /(.*)(>|<|=)(.*)/) {
1961 "handling unless (operator)... left:$left operator:$operator right:$right"
1969 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1972 # automatic replace for short operators
1973 $left = 'title' if $left =~ '^ti$';
1974 $left = 'author' if $left =~ '^au$';
1975 $left = 'publisher' if $left =~ '^pb$';
1976 $left = 'subject' if $left =~ '^su$';
1977 $left = 'koha-Auth-Number' if $left =~ '^an$';
1978 $left = 'keyword' if $left =~ '^kw$';
1979 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1980 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1981 my $dbh = C4::Context->dbh;
1982 if ( $operator && $left ne 'keyword' ) {
1983 #do a specific search
1984 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1985 my $sth = $dbh->prepare(
1986 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1988 warn "$left / $operator / $right\n" if $DEBUG;
1990 # split each word, query the DB and build the biblionumbers result
1991 #sanitizing leftpart
1992 $left =~ s/^\s+|\s+$//;
1993 foreach ( split / /, $right ) {
1995 $_ =~ s/^\s+|\s+$//;
1997 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1998 $sth->execute( $server, $left, $_ )
1999 or warn "execute failed: $!";
2000 while ( my ( $line, $value ) = $sth->fetchrow ) {
2002 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2003 # otherwise, fill the result
2004 $biblionumbers .= $line
2005 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2006 warn "result : $value "
2007 . ( $right =~ /\d/ ) . "=="
2008 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2011 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2013 warn "NZAND" if $DEBUG;
2014 $results = NZoperatorAND($biblionumbers,$results);
2016 $results = $biblionumbers;
2021 #do a complete search (all indexes), if index='kw' do complete search too.
2022 my $sth = $dbh->prepare(
2023 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2026 # split each word, query the DB and build the biblionumbers result
2027 foreach ( split / /, $string ) {
2028 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2029 warn "search on all indexes on $_" if $DEBUG;
2032 $sth->execute( $server, $_ );
2033 while ( my $line = $sth->fetchrow ) {
2034 $biblionumbers .= $line;
2037 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2039 $results = NZoperatorAND($biblionumbers,$results);
2042 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2043 $results = $biblionumbers;
2047 warn "return : $results for LEAF : $string" if $DEBUG;
2050 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2054 my ($rightresult, $leftresult)=@_;
2056 my @leftresult = split /;/, $leftresult;
2057 warn " @leftresult / $rightresult \n" if $DEBUG;
2059 # my @rightresult = split /;/,$leftresult;
2062 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2063 # the result is stored twice, to have the same weight for AND than OR.
2064 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2065 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2066 foreach (@leftresult) {
2069 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2070 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2071 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2073 "$value-$countvalue;$value-$countvalue;";
2076 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2077 return $finalresult;
2081 my ($rightresult, $leftresult)=@_;
2082 return $rightresult.$leftresult;
2086 my ($leftresult, $rightresult)=@_;
2088 my @leftresult = split /;/, $leftresult;
2090 # my @rightresult = split /;/,$leftresult;
2092 foreach (@leftresult) {
2094 $value=$1 if $value=~m/(.*)-\d+$/;
2095 unless ($rightresult =~ "$value-") {
2096 $finalresult .= "$_;";
2099 return $finalresult;
2104 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2111 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2112 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2114 # order title asc by default
2115 # $ordering = '1=36 <i' unless $ordering;
2116 $results_per_page = 20 unless $results_per_page;
2117 $offset = 0 unless $offset;
2118 my $dbh = C4::Context->dbh;
2121 # order by POPULARITY
2123 if ( $ordering =~ /popularity/ ) {
2127 # popularity is not in MARC record, it's builded from a specific query
2129 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2130 foreach ( split /;/, $biblionumbers ) {
2131 my ( $biblionumber, $title ) = split /,/, $_;
2132 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2133 $sth->execute($biblionumber);
2134 my $popularity = $sth->fetchrow || 0;
2136 # hint : the key is popularity.title because we can have
2137 # many results with the same popularity. In this case, sub-ordering is done by title
2138 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2139 # (un-frequent, I agree, but we won't forget anything that way ;-)
2140 $popularity{ sprintf( "%10d", $popularity ) . $title
2141 . $biblionumber } = $biblionumber;
2144 # sort the hash and return the same structure as GetRecords (Zebra querying)
2147 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2148 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2149 $result_hash->{'RECORDS'}[ $numbers++ ] =
2150 $result{ $popularity{$key} }->as_usmarc();
2153 else { # sort popularity ASC
2154 foreach my $key ( sort ( keys %popularity ) ) {
2155 $result_hash->{'RECORDS'}[ $numbers++ ] =
2156 $result{ $popularity{$key} }->as_usmarc();
2159 my $finalresult = ();
2160 $result_hash->{'hits'} = $numbers;
2161 $finalresult->{'biblioserver'} = $result_hash;
2162 return $finalresult;
2168 elsif ( $ordering =~ /author/ ) {
2170 foreach ( split /;/, $biblionumbers ) {
2171 my ( $biblionumber, $title ) = split /,/, $_;
2172 my $record = GetMarcBiblio($biblionumber);
2174 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2175 $author = $record->subfield( '200', 'f' );
2176 $author = $record->subfield( '700', 'a' ) unless $author;
2179 $author = $record->subfield( '100', 'a' );
2182 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2183 # and we don't want to get only 1 result for each of them !!!
2184 $result{ $author . $biblionumber } = $record;
2187 # sort the hash and return the same structure as GetRecords (Zebra querying)
2190 if ( $ordering eq 'author_za' ) { # sort by author desc
2191 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2192 $result_hash->{'RECORDS'}[ $numbers++ ] =
2193 $result{$key}->as_usmarc();
2196 else { # sort by author ASC
2197 foreach my $key ( sort ( keys %result ) ) {
2198 $result_hash->{'RECORDS'}[ $numbers++ ] =
2199 $result{$key}->as_usmarc();
2202 my $finalresult = ();
2203 $result_hash->{'hits'} = $numbers;
2204 $finalresult->{'biblioserver'} = $result_hash;
2205 return $finalresult;
2208 # ORDER BY callnumber
2211 elsif ( $ordering =~ /callnumber/ ) {
2213 foreach ( split /;/, $biblionumbers ) {
2214 my ( $biblionumber, $title ) = split /,/, $_;
2215 my $record = GetMarcBiblio($biblionumber);
2217 my $frameworkcode = GetFrameworkCode($biblionumber);
2218 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2219 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2220 unless $callnumber_tag;
2221 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2222 $callnumber = $record->subfield( '200', 'f' );
2224 $callnumber = $record->subfield( '100', 'a' );
2227 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2228 # and we don't want to get only 1 result for each of them !!!
2229 $result{ $callnumber . $biblionumber } = $record;
2232 # sort the hash and return the same structure as GetRecords (Zebra querying)
2235 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2236 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2237 $result_hash->{'RECORDS'}[ $numbers++ ] =
2238 $result{$key}->as_usmarc();
2241 else { # sort by title ASC
2242 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2243 $result_hash->{'RECORDS'}[ $numbers++ ] =
2244 $result{$key}->as_usmarc();
2247 my $finalresult = ();
2248 $result_hash->{'hits'} = $numbers;
2249 $finalresult->{'biblioserver'} = $result_hash;
2250 return $finalresult;
2252 elsif ( $ordering =~ /pubdate/ ) { #pub year
2254 foreach ( split /;/, $biblionumbers ) {
2255 my ( $biblionumber, $title ) = split /,/, $_;
2256 my $record = GetMarcBiblio($biblionumber);
2257 my ( $publicationyear_tag, $publicationyear_subfield ) =
2258 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2259 my $publicationyear =
2260 $record->subfield( $publicationyear_tag,
2261 $publicationyear_subfield );
2263 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2264 # and we don't want to get only 1 result for each of them !!!
2265 $result{ $publicationyear . $biblionumber } = $record;
2268 # sort the hash and return the same structure as GetRecords (Zebra querying)
2271 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2272 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2273 $result_hash->{'RECORDS'}[ $numbers++ ] =
2274 $result{$key}->as_usmarc();
2277 else { # sort by pub year ASC
2278 foreach my $key ( sort ( keys %result ) ) {
2279 $result_hash->{'RECORDS'}[ $numbers++ ] =
2280 $result{$key}->as_usmarc();
2283 my $finalresult = ();
2284 $result_hash->{'hits'} = $numbers;
2285 $finalresult->{'biblioserver'} = $result_hash;
2286 return $finalresult;
2292 elsif ( $ordering =~ /title/ ) {
2294 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2296 foreach ( split /;/, $biblionumbers ) {
2297 my ( $biblionumber, $title ) = split /,/, $_;
2299 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2300 # and we don't want to get only 1 result for each of them !!!
2301 # hint & speed improvement : we can order without reading the record
2302 # so order, and read records only for the requested page !
2303 $result{ $title . $biblionumber } = $biblionumber;
2306 # sort the hash and return the same structure as GetRecords (Zebra querying)
2309 if ( $ordering eq 'title_az' ) { # sort by title desc
2310 foreach my $key ( sort ( keys %result ) ) {
2311 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2314 else { # sort by title ASC
2315 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2316 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2320 # limit the $results_per_page to result size if it's more
2321 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2323 # for the requested page, replace biblionumber by the complete record
2324 # speed improvement : avoid reading too much things
2326 my $counter = $offset ;
2327 $counter <= $offset + $results_per_page ;
2331 $result_hash->{'RECORDS'}[$counter] =
2332 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2334 my $finalresult = ();
2335 $result_hash->{'hits'} = $numbers;
2336 $finalresult->{'biblioserver'} = $result_hash;
2337 return $finalresult;
2344 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2347 foreach ( split /;/, $biblionumbers ) {
2348 my ( $biblionumber, $title ) = split /,/, $_;
2349 $title =~ /(.*)-(\d)/;
2354 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2355 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2356 # biblio N has ranking = 6
2357 $count_ranking{$biblionumber} += $ranking;
2360 # build the result by "inverting" the count_ranking hash
2361 # 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
2363 foreach ( keys %count_ranking ) {
2364 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2367 # sort the hash and return the same structure as GetRecords (Zebra querying)
2370 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2371 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2374 # limit the $results_per_page to result size if it's more
2375 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2377 # for the requested page, replace biblionumber by the complete record
2378 # speed improvement : avoid reading too much things
2380 my $counter = $offset ;
2381 $counter <= $offset + $results_per_page ;
2385 $result_hash->{'RECORDS'}[$counter] =
2386 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2387 if $result_hash->{'RECORDS'}[$counter];
2389 my $finalresult = ();
2390 $result_hash->{'hits'} = $numbers;
2391 $finalresult->{'biblioserver'} = $result_hash;
2392 return $finalresult;
2396 =head2 enabled_staff_search_views
2398 %hash = enabled_staff_search_views()
2400 This function returns a hash that contains three flags obtained from the system
2401 preferences, used to determine whether a particular staff search results view
2406 =item C<Output arg:>
2408 * $hash{can_view_MARC} is true only if the MARC view is enabled
2409 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2410 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2412 =item C<usage in the script:>
2416 $template->param ( C4::Search::enabled_staff_search_views );
2420 sub enabled_staff_search_views
2423 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2424 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2425 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2429 sub AddSearchHistory{
2430 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2431 my $dbh = C4::Context->dbh;
2433 # Add the request the user just made
2434 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2435 my $sth = $dbh->prepare($sql);
2436 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2437 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2440 sub GetSearchHistory{
2441 my ($borrowernumber,$session)=@_;
2442 my $dbh = C4::Context->dbh;
2444 # Add the request the user just made
2445 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2446 my $sth = $dbh->prepare($query);
2447 $sth->execute($borrowernumber, $session);
2448 return $sth->fetchall_hashref({});
2451 =head2 z3950_search_args
2453 $arrayref = z3950_search_args($matchpoints)
2455 This function returns an array reference that contains the search parameters to be
2456 passed to the Z39.50 search script (z3950_search.pl). The array elements
2457 are hash refs whose keys are name, value and encvalue, and whose values are the
2458 name of a search parameter, the value of that search parameter and the URL encoded
2459 value of that parameter.
2461 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2463 The search parameter values are obtained from the bibliographic record whose
2464 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2466 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2467 a general purpose search argument. In this case, the returned array contains only
2468 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2470 If a search parameter value is undefined or empty, it is not included in the returned
2473 The returned array reference may be passed directly to the template parameters.
2477 =item C<Output arg:>
2479 * $array containing hash refs as described above
2481 =item C<usage in the script:>
2485 $data = Biblio::GetBiblioData($bibno);
2486 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2490 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2494 sub z3950_search_args {
2496 $bibrec = { title => $bibrec } if !ref $bibrec;
2498 for my $field (qw/ lccn isbn issn title author dewey subject /)
2500 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2501 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2506 =head2 BiblioAddAuthorities
2508 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2510 this function finds the authorities linked to the biblio
2511 * search in the authority DB for the same authid (in $9 of the biblio)
2512 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2513 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2514 OR adds a new authority record
2520 * $record is the MARC record in question (marc blob)
2521 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2523 =item C<Output arg:>
2525 * $countlinked is the number of authorities records that are linked to this authority
2529 * 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)
2536 sub BiblioAddAuthorities{
2537 my ( $record, $frameworkcode ) = @_;
2538 my $dbh=C4::Context->dbh;
2539 my $query=$dbh->prepare(qq|
2540 SELECT authtypecode,tagfield
2541 FROM marc_subfield_structure
2542 WHERE frameworkcode=?
2543 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2544 # SELECT authtypecode,tagfield
2545 # FROM marc_subfield_structure
2546 # WHERE frameworkcode=?
2547 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2548 $query->execute($frameworkcode);
2549 my ($countcreated,$countlinked);
2550 while (my $data=$query->fetchrow_hashref){
2551 foreach my $field ($record->field($data->{tagfield})){
2552 next if ($field->subfield('3')||$field->subfield('9'));
2553 # No authorities id in the tag.
2554 # Search if there is any authorities to link to.
2555 my $query='at='.$data->{authtypecode}.' ';
2556 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2557 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2558 # there is only 1 result
2560 warn "BIBLIOADDSAUTHORITIES: $error";
2563 if ($results && scalar(@$results)==1) {
2564 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2565 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2567 } elsif (scalar(@$results)>1) {
2568 #More than One result
2569 #This can comes out of a lack of a subfield.
2570 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2571 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2574 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2575 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2576 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2577 my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2578 next unless $authtypedata;
2579 my $marcrecordauth=MARC::Record->new();
2580 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2581 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2582 $marcrecordauth->insert_fields_ordered($authfield);
2584 # bug 2317: ensure new authority knows it's using UTF-8; currently
2585 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2586 # automatically for UNIMARC (by not transcoding)
2587 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2588 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2589 # of change to a core API just before the 3.0 release.
2590 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2591 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2594 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2596 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2598 $field->add_subfields('9'=>$authid);
2602 return ($countlinked,$countcreated);
2605 =head2 GetDistinctValues($field);
2607 C<$field> is a reference to the fields array
2611 sub GetDistinctValues {
2612 my ($fieldname,$string)=@_;
2613 # returns a reference to a hash of references to branches...
2614 if ($fieldname=~/\./){
2615 my ($table,$column)=split /\./, $fieldname;
2616 my $dbh = C4::Context->dbh;
2617 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2618 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 ");
2620 my $elements=$sth->fetchall_arrayref({});
2625 my @servers=qw<biblioserver authorityserver>;
2626 my (@zconns,@results);
2627 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2628 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2631 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2634 # The big moment: asynchronously retrieve results from all servers
2636 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2637 my $ev = $zconns[ $i - 1 ]->last_event();
2638 if ( $ev == ZOOM::Event::ZEND ) {
2639 next unless $results[ $i - 1 ];
2640 my $size = $results[ $i - 1 ]->size();
2642 for (my $j=0;$j<$size;$j++){
2644 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2645 push @elements, \%hashscan;
2655 END { } # module clean-up code here (global destructor)
2662 Koha Development Team <http://koha-community.org/>