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',
896 'Record-control-number',
907 'Subject-heading-thesaurus',
908 'Subject-name-personal',
909 'Subject-subdivision',
918 'Term-genre-form-heading',
919 'Term-genre-form-see',
920 'Term-genre-form-seealso',
926 'Title-uniform-heading',
928 'Title-uniform-seealso',
938 'classification-source',
940 'coded-location-qualifier',
951 'Local-classification',
954 'materials-specified',
963 'replacementpricedate',
979 $simple_query, $query_cgi,
981 $limit_cgi, $limit_desc,
982 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
984 Build queries and limits in CCL, CGI, Human,
985 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
987 See verbose embedded documentation.
993 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
995 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
998 my @operators = $operators ? @$operators : ();
999 my @indexes = $indexes ? @$indexes : ();
1000 my @operands = $operands ? @$operands : ();
1001 my @limits = $limits ? @$limits : ();
1002 my @sort_by = $sort_by ? @$sort_by : ();
1004 my $stemming = C4::Context->preference("QueryStemming") || 0;
1005 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1006 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1007 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1008 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1010 # no stemming/weight/fuzzy in NoZebra
1011 if ( C4::Context->preference("NoZebra") ) {
1015 $auto_truncation = 0;
1018 my $query = $operands[0];
1019 my $simple_query = $operands[0];
1021 # initialize the variables we're passing back
1030 my $stopwords_removed; # flag to determine if stopwords have been removed
1033 my $cclindexes = getIndexes();
1034 if( $query !~ /\s*ccl=/ ){
1035 for my $index (@$cclindexes){
1036 if($query =~ /($index)(,?\w)*[:=]/){
1040 $query = "ccl=$query" if($cclq);
1043 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1045 if ( $query =~ /^ccl=/ ) {
1047 # This is needed otherwise ccl= and &limit won't work together, and
1048 # this happens when selecting a subject on the opac-detail page
1050 $q .= ' and '.join(' and ', @limits);
1052 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1054 if ( $query =~ /^cql=/ ) {
1055 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1057 if ( $query =~ /^pqf=/ ) {
1058 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1061 # pass nested queries directly
1062 # FIXME: need better handling of some of these variables in this case
1063 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1064 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1065 # if ( $query =~ /(\(|\))/ ) {
1067 # undef, $query, $simple_query, $query_cgi,
1068 # $query, $limit, $limit_cgi, $limit_desc,
1069 # $stopwords_removed, 'ccl'
1073 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1074 # query operands and indexes and add stemming, truncation, field weighting, etc.
1075 # Once we do so, we'll end up with a value in $query, just like if we had an
1076 # incoming $query from the user
1079 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1080 my $previous_operand
1081 ; # a flag used to keep track if there was a previous query
1082 # if there was, we can apply the current operator
1084 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1086 # COMBINE OPERANDS, INDEXES AND OPERATORS
1087 if ( $operands[$i] ) {
1088 $operands[$i]=~s/^\s+//;
1090 # A flag to determine whether or not to add the index to the query
1093 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1094 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1097 $remove_stopwords = 0;
1099 my $operand = $operands[$i];
1100 my $index = $indexes[$i];
1102 # Add index-specific attributes
1103 # Date of Publication
1104 if ( $index eq 'yr' ) {
1105 $index .= ",st-numeric";
1107 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1110 # Date of Acquisition
1111 elsif ( $index eq 'acqdate' ) {
1112 $index .= ",st-date-normalized";
1114 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1116 # ISBN,ISSN,Standard Number, don't need special treatment
1117 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1120 $stemming, $auto_truncation,
1121 $weight_fields, $fuzzy_enabled,
1123 ) = ( 0, 0, 0, 0, 0 );
1131 # Set default structure attribute (word list)
1132 my $struct_attr = q{};
1133 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1134 $struct_attr = ",wrdl";
1137 # Some helpful index variants
1138 my $index_plus = $index . $struct_attr . ':';
1139 my $index_plus_comma = $index . $struct_attr . ',';
1142 if ($remove_stopwords) {
1143 ( $operand, $stopwords_removed ) =
1144 _remove_stopwords( $operand, $index );
1145 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1146 warn "REMOVED STOPWORDS: @$stopwords_removed"
1147 if ( $stopwords_removed && $DEBUG );
1150 if ($auto_truncation){
1151 unless ( $index =~ /(st-|phr|ext)/ ) {
1152 #FIXME only valid with LTR scripts
1153 $operand=join(" ",map{
1154 (index($_,"*")>0?"$_":"$_*")
1155 }split (/\s+/,$operand));
1156 warn $operand if $DEBUG;
1161 my $truncated_operand;
1162 my( $nontruncated, $righttruncated, $lefttruncated,
1163 $rightlefttruncated, $regexpr
1164 ) = _detect_truncation( $operand, $index );
1166 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1171 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1172 scalar(@$rightlefttruncated) > 0 )
1175 # Don't field weight or add the index to the query, we do it here
1177 undef $weight_fields;
1178 my $previous_truncation_operand;
1179 if (scalar @$nontruncated) {
1180 $truncated_operand .= "$index_plus @$nontruncated ";
1181 $previous_truncation_operand = 1;
1183 if (scalar @$righttruncated) {
1184 $truncated_operand .= "and " if $previous_truncation_operand;
1185 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1186 $previous_truncation_operand = 1;
1188 if (scalar @$lefttruncated) {
1189 $truncated_operand .= "and " if $previous_truncation_operand;
1190 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1191 $previous_truncation_operand = 1;
1193 if (scalar @$rightlefttruncated) {
1194 $truncated_operand .= "and " if $previous_truncation_operand;
1195 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1196 $previous_truncation_operand = 1;
1199 $operand = $truncated_operand if $truncated_operand;
1200 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1203 my $stemmed_operand;
1204 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1207 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1209 # Handle Field Weighting
1210 my $weighted_operand;
1211 if ($weight_fields) {
1212 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1213 $operand = $weighted_operand;
1217 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1219 # If there's a previous operand, we need to add an operator
1220 if ($previous_operand) {
1222 # User-specified operator
1223 if ( $operators[ $i - 1 ] ) {
1224 $query .= " $operators[$i-1] ";
1225 $query .= " $index_plus " unless $indexes_set;
1226 $query .= " $operand";
1227 $query_cgi .= "&op=$operators[$i-1]";
1228 $query_cgi .= "&idx=$index" if $index;
1229 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1231 " $operators[$i-1] $index_plus $operands[$i]";
1234 # Default operator is and
1237 $query .= "$index_plus " unless $indexes_set;
1238 $query .= "$operand";
1239 $query_cgi .= "&op=and&idx=$index" if $index;
1240 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1241 $query_desc .= " and $index_plus $operands[$i]";
1245 # There isn't a pervious operand, don't need an operator
1248 # Field-weighted queries already have indexes set
1249 $query .= " $index_plus " unless $indexes_set;
1251 $query_desc .= " $index_plus $operands[$i]";
1252 $query_cgi .= "&idx=$index" if $index;
1253 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1254 $previous_operand = 1;
1259 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1262 my $group_OR_limits;
1263 my $availability_limit;
1264 foreach my $this_limit (@limits) {
1265 if ( $this_limit =~ /available/ ) {
1267 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1269 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1270 $availability_limit .=
1271 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1272 $limit_cgi .= "&limit=available";
1276 # group_OR_limits, prefixed by mc-
1277 # OR every member of the group
1278 elsif ( $this_limit =~ /mc/ ) {
1280 if ( $this_limit =~ /mc-ccode:/ ) {
1281 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1282 $this_limit =~ tr/"//d;
1283 my ($k,$v) = split(/:/, $this_limit,2);
1284 $this_limit = $k.":\"".$v."\"";
1287 $group_OR_limits .= " or " if $group_OR_limits;
1288 $limit_desc .= " or " if $group_OR_limits;
1289 $group_OR_limits .= "$this_limit";
1290 $limit_cgi .= "&limit=$this_limit";
1291 $limit_desc .= " $this_limit";
1294 # Regular old limits
1296 $limit .= " and " if $limit || $query;
1297 $limit .= "$this_limit";
1298 $limit_cgi .= "&limit=$this_limit";
1299 if ($this_limit =~ /^branch:(.+)/) {
1300 my $branchcode = $1;
1301 my $branchname = GetBranchName($branchcode);
1302 if (defined $branchname) {
1303 $limit_desc .= " branch:$branchname";
1305 $limit_desc .= " $this_limit";
1308 $limit_desc .= " $this_limit";
1312 if ($group_OR_limits) {
1313 $limit .= " and " if ( $query || $limit );
1314 $limit .= "($group_OR_limits)";
1316 if ($availability_limit) {
1317 $limit .= " and " if ( $query || $limit );
1318 $limit .= "($availability_limit)";
1321 # Normalize the query and limit strings
1322 # This is flawed , means we can't search anything with : in it
1323 # if user wants to do ccl or cql, start the query with that
1324 # $query =~ s/:/=/g;
1325 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc)):/=/g;
1326 $query =~ s/(?<=(wrdl)):/=/g;
1327 $query =~ s/(?<=(trn|phr)):/=/g;
1329 for ( $query, $query_desc, $limit, $limit_desc ) {
1330 s/ +/ /g; # remove extra spaces
1331 s/^ //g; # remove any beginning spaces
1332 s/ $//g; # remove any ending spaces
1333 s/==/=/g; # remove double == from query
1335 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1337 for ($query_cgi,$simple_query) {
1340 # append the limit to the query
1341 $query .= " " . $limit;
1345 warn "QUERY:" . $query;
1346 warn "QUERY CGI:" . $query_cgi;
1347 warn "QUERY DESC:" . $query_desc;
1348 warn "LIMIT:" . $limit;
1349 warn "LIMIT CGI:" . $limit_cgi;
1350 warn "LIMIT DESC:" . $limit_desc;
1351 warn "---------\nLeave buildQuery\n---------";
1354 undef, $query, $simple_query, $query_cgi,
1355 $query_desc, $limit, $limit_cgi, $limit_desc,
1356 $stopwords_removed, $query_type
1360 =head2 searchResults
1362 my @search_results = searchResults($search_context, $searchdesc, $hits,
1363 $results_per_page, $offset, $scan,
1364 @marcresults, $hidelostitems);
1366 Format results in a form suitable for passing to the template
1370 # IMO this subroutine is pretty messy still -- it's responsible for
1371 # building the HTML output for the template
1373 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1374 my $dbh = C4::Context->dbh;
1377 $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1379 #Build branchnames hash
1381 #get branch information.....
1383 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1385 while ( my $bdata = $bsth->fetchrow_hashref ) {
1386 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1388 # FIXME - We build an authorised values hash here, using the default framework
1389 # though it is possible to have different authvals for different fws.
1391 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1393 # get notforloan authorised value list (see $shelflocations FIXME)
1394 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1396 #Build itemtype hash
1397 #find itemtype & itemtype image
1401 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1404 while ( my $bdata = $bsth->fetchrow_hashref ) {
1405 foreach (qw(description imageurl summary notforloan)) {
1406 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1410 #search item field code
1413 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1416 my ($itemtag) = $sth->fetchrow;
1418 ## find column names of items related to MARC
1419 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1421 my %subfieldstosearch;
1422 while ( ( my $column ) = $sth2->fetchrow ) {
1423 my ( $tagfield, $tagsubfield ) =
1424 &GetMarcFromKohaField( "items." . $column, "" );
1425 $subfieldstosearch{$column} = $tagsubfield;
1428 # handle which records to actually retrieve
1430 if ( $hits && $offset + $results_per_page <= $hits ) {
1431 $times = $offset + $results_per_page;
1434 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1437 my $marcflavour = C4::Context->preference("marcflavour");
1438 # We get the biblionumber position in MARC
1439 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1442 # loop through all of the records we've retrieved
1443 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1444 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1448 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1449 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1450 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1451 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1452 $oldbiblio->{result_number} = $i + 1;
1454 # add imageurl to itemtype if there is one
1455 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1457 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1458 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1459 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1460 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1461 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1462 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1464 # edition information, if any
1465 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1466 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1467 # Build summary if there is one (the summary is defined in the itemtypes table)
1468 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1469 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1470 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1471 my @fields = $marcrecord->fields();
1474 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1476 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1477 $tag =~ /(.{3})(.)/;
1478 if($marcrecord->field($1)){
1479 my @abc = $marcrecord->field($1)->subfield($2);
1480 $tags->{$tag} = $#abc + 1 ;
1484 # We catch how many times to repeat this line
1486 foreach my $tag (keys(%$tags)){
1487 $max = $tags->{$tag} if($tags->{$tag} > $max);
1490 # we replace, and repeat each line
1491 for (my $i = 0 ; $i < $max ; $i++){
1492 my $newline = $line;
1494 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1495 $tag =~ /(.{3})(.)/;
1497 if($marcrecord->field($1)){
1498 my @repl = $marcrecord->field($1)->subfield($2);
1499 my $subfieldvalue = $repl[$i];
1501 if (! utf8::is_utf8($subfieldvalue)) {
1502 utf8::decode($subfieldvalue);
1505 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1508 $newsummary .= "$newline\n";
1512 $newsummary =~ s/\[(.*?)]//g;
1513 $newsummary =~ s/\n/<br\/>/g;
1514 $oldbiblio->{summary} = $newsummary;
1517 # Pull out the items fields
1518 my @fields = $marcrecord->field($itemtag);
1520 # Setting item statuses for display
1521 my @available_items_loop;
1522 my @onloan_items_loop;
1523 my @other_items_loop;
1525 my $available_items;
1529 my $ordered_count = 0;
1530 my $available_count = 0;
1531 my $onloan_count = 0;
1532 my $longoverdue_count = 0;
1533 my $other_count = 0;
1534 my $wthdrawn_count = 0;
1535 my $itemlost_count = 0;
1536 my $itembinding_count = 0;
1537 my $itemdamaged_count = 0;
1538 my $item_in_transit_count = 0;
1539 my $can_place_holds = 0;
1540 my $item_onhold_count = 0;
1541 my $items_count = scalar(@fields);
1543 ( C4::Context->preference('maxItemsinSearchResults') )
1544 ? C4::Context->preference('maxItemsinSearchResults') - 1
1547 # loop through every item
1548 foreach my $field (@fields) {
1551 # populate the items hash
1552 foreach my $code ( keys %subfieldstosearch ) {
1553 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1556 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1557 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1558 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1559 if ($item->{$hbranch}) {
1560 $item->{'branchname'} = $branches{$item->{$hbranch}};
1562 elsif ($item->{$otherbranch}) { # Last resort
1563 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1566 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1567 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1568 if ( $item->{onloan} ) {
1570 my $key = $prefix . $item->{onloan} . $item->{barcode};
1571 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1572 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1573 $onloan_items->{$key}->{branchname} = $item->{branchname};
1574 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1575 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1576 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1577 # if something's checked out and lost, mark it as 'long overdue'
1578 if ( $item->{itemlost} ) {
1579 $onloan_items->{$prefix}->{longoverdue}++;
1580 $longoverdue_count++;
1581 } else { # can place holds as long as item isn't lost
1582 $can_place_holds = 1;
1586 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1590 if ( $item->{notforloan} == -1 ) {
1594 # is item in transit?
1595 my $transfertwhen = '';
1596 my ($transfertfrom, $transfertto);
1598 # is item on the reserve shelf?
1599 my $reservestatus = 0;
1602 unless ($item->{wthdrawn}
1603 || $item->{itemlost}
1605 || $item->{notforloan}
1606 || $items_count > 20) {
1608 # A couple heuristics to limit how many times
1609 # we query the database for item transfer information, sacrificing
1610 # accuracy in some cases for speed;
1612 # 1. don't query if item has one of the other statuses
1613 # 2. don't check transit status if the bib has
1614 # more than 20 items
1616 # FIXME: to avoid having the query the database like this, and to make
1617 # the in transit status count as unavailable for search limiting,
1618 # should map transit status to record indexed in Zebra.
1620 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1621 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1624 # item is withdrawn, lost or damaged
1625 if ( $item->{wthdrawn}
1626 || $item->{itemlost}
1628 || $item->{notforloan} > 0
1629 || $reservestatus eq 'Waiting'
1630 || ($transfertwhen ne ''))
1632 $wthdrawn_count++ if $item->{wthdrawn};
1633 $itemlost_count++ if $item->{itemlost};
1634 $itemdamaged_count++ if $item->{damaged};
1635 $item_in_transit_count++ if $transfertwhen ne '';
1636 $item_onhold_count++ if $reservestatus eq 'Waiting';
1637 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1640 my $key = $prefix . $item->{status};
1641 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1642 $other_items->{$key}->{$_} = $item->{$_};
1644 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1645 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1646 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1647 $other_items->{$key}->{count}++ if $item->{$hbranch};
1648 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1649 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1653 $can_place_holds = 1;
1655 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1656 foreach (qw(branchname itemcallnumber)) {
1657 $available_items->{$prefix}->{$_} = $item->{$_};
1659 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1660 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1663 } # notforloan, item level and biblioitem level
1664 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1666 ( C4::Context->preference('maxItemsinSearchResults') )
1667 ? C4::Context->preference('maxItemsinSearchResults') - 1
1669 for my $key ( sort keys %$onloan_items ) {
1670 (++$onloanitemscount > $maxitems) and last;
1671 push @onloan_items_loop, $onloan_items->{$key};
1673 for my $key ( sort keys %$other_items ) {
1674 (++$otheritemscount > $maxitems) and last;
1675 push @other_items_loop, $other_items->{$key};
1677 for my $key ( sort keys %$available_items ) {
1678 (++$availableitemscount > $maxitems) and last;
1679 push @available_items_loop, $available_items->{$key}
1682 # XSLT processing of some stuff
1684 SetUTF8Flag($marcrecord);
1685 $debug && warn $marcrecord->as_formatted;
1686 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1687 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1688 # is not implemented yet
1689 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1693 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1694 $can_place_holds = 0
1695 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1696 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1697 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1698 $oldbiblio->{items_count} = $items_count;
1699 $oldbiblio->{available_items_loop} = \@available_items_loop;
1700 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1701 $oldbiblio->{other_items_loop} = \@other_items_loop;
1702 $oldbiblio->{availablecount} = $available_count;
1703 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1704 $oldbiblio->{onloancount} = $onloan_count;
1705 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1706 $oldbiblio->{othercount} = $other_count;
1707 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1708 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1709 $oldbiblio->{itemlostcount} = $itemlost_count;
1710 $oldbiblio->{damagedcount} = $itemdamaged_count;
1711 $oldbiblio->{intransitcount} = $item_in_transit_count;
1712 $oldbiblio->{onholdcount} = $item_onhold_count;
1713 $oldbiblio->{orderedcount} = $ordered_count;
1714 $oldbiblio->{isbn} =~
1715 s/-//g; # deleting - in isbn to enable amazon content
1716 push( @newresults, $oldbiblio )
1717 if(not $hidelostitems
1718 or (($items_count > $itemlost_count )
1719 && $hidelostitems));
1725 =head2 SearchAcquisitions
1726 Search for acquisitions
1729 sub SearchAcquisitions{
1730 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1732 my $dbh=C4::Context->dbh;
1733 # Variable initialization
1737 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1738 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1739 WHERE dateaccessioned BETWEEN ? AND ?
1742 my (@params,@loopcriteria);
1744 push @params, $datebegin->output("iso");
1745 push @params, $dateend->output("iso");
1747 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1748 if(C4::Context->preference("item-level_itypes")){
1749 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1751 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1753 push @params, @$itemtypes;
1756 if ($criteria =~/itemtype/){
1757 if(C4::Context->preference("item-level_itypes")){
1758 $str .= "AND items.itype=? ";
1760 $str .= "AND biblioitems.itemtype=? ";
1763 if(scalar(@$itemtypes) == 0){
1764 my $itypes = GetItemTypes();
1765 for my $key (keys %$itypes){
1766 push @$itemtypes, $key;
1770 @loopcriteria= @$itemtypes;
1771 }elsif ($criteria=~/itemcallnumber/){
1772 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1773 OR items.itemcallnumber is NULL
1774 OR items.itemcallnumber = '')";
1776 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1778 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1779 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1782 if ($orderby =~ /date_desc/){
1783 $str.=" ORDER BY dateaccessioned DESC";
1785 $str.=" ORDER BY title";
1788 my $qdataacquisitions=$dbh->prepare($str);
1790 my @loopacquisitions;
1791 foreach my $value(@loopcriteria){
1792 push @params,$value;
1794 $cell{"title"}=$value;
1795 $cell{"titlecode"}=$value;
1797 eval{$qdataacquisitions->execute(@params);};
1799 if ($@){ warn "recentacquisitions Error :$@";}
1802 while (my $data=$qdataacquisitions->fetchrow_hashref){
1803 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1805 $cell{"loopdata"}=\@loopdata;
1807 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1810 $qdataacquisitions->finish;
1811 return \@loopacquisitions;
1813 #----------------------------------------------------------------------
1815 # Non-Zebra GetRecords#
1816 #----------------------------------------------------------------------
1820 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1826 $query, $simple_query, $sort_by_ref, $servers_ref,
1827 $results_per_page, $offset, $expanded_facet, $branches,
1830 warn "query =$query" if $DEBUG;
1831 my $result = NZanalyse($query);
1832 warn "results =$result" if $DEBUG;
1834 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1840 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1841 the list is built from an inverted index in the nozebra SQL table
1842 note that title is here only for convenience : the sorting will be very fast when requested on title
1843 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1848 my ( $string, $server ) = @_;
1849 # warn "---------" if $DEBUG;
1850 warn " NZanalyse" if $DEBUG;
1851 # warn "---------" if $DEBUG;
1853 # $server contains biblioserver or authorities, depending on what we search on.
1854 #warn "querying : $string on $server";
1855 $server = 'biblioserver' unless $server;
1857 # if we have a ", replace the content to discard temporarily any and/or/not inside
1859 if ( $string =~ /"/ ) {
1860 $string =~ s/"(.*?)"/__X__/;
1862 warn "commacontent : $commacontent" if $DEBUG;
1865 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1866 # then, call again NZanalyse with $left and $right
1867 # (recursive until we find a leaf (=> something without and/or/not)
1868 # delete repeated operator... Would then go in infinite loop
1869 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1872 #process parenthesis before.
1873 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1876 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1878 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1880 my $leftresult = NZanalyse( $left, $server );
1882 my $rightresult = NZanalyse( $right, $server );
1884 # OK, we have the results for right and left part of the query
1885 # depending of operand, intersect, union or exclude both lists
1886 # to get a result list
1887 if ( $operator eq ' and ' ) {
1888 return NZoperatorAND($leftresult,$rightresult);
1890 elsif ( $operator eq ' or ' ) {
1892 # just merge the 2 strings
1893 return $leftresult . $rightresult;
1895 elsif ( $operator eq ' not ' ) {
1896 return NZoperatorNOT($leftresult,$rightresult);
1900 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1904 warn "string :" . $string if $DEBUG;
1908 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1911 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1913 warn "no parenthesis. left : $left operator: $operator right: $right"
1916 # it's not a leaf, we have a and/or/not
1919 # reintroduce comma content if needed
1920 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1921 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1922 warn "node : $left / $operator / $right\n" if $DEBUG;
1923 my $leftresult = NZanalyse( $left, $server );
1924 my $rightresult = NZanalyse( $right, $server );
1925 warn " leftresult : $leftresult" if $DEBUG;
1926 warn " rightresult : $rightresult" if $DEBUG;
1927 # OK, we have the results for right and left part of the query
1928 # depending of operand, intersect, union or exclude both lists
1929 # to get a result list
1930 if ( $operator eq ' and ' ) {
1931 return NZoperatorAND($leftresult,$rightresult);
1933 elsif ( $operator eq ' or ' ) {
1935 # just merge the 2 strings
1936 return $leftresult . $rightresult;
1938 elsif ( $operator eq ' not ' ) {
1939 return NZoperatorNOT($leftresult,$rightresult);
1943 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1944 die "error : operand unknown : $operator for $string";
1947 # it's a leaf, do the real SQL query and return the result
1950 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1951 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1952 #remove trailing blank at the beginning
1954 warn "leaf:$string" if $DEBUG;
1956 # parse the string in in operator/operand/value again
1960 if ($string =~ /(.*)(>=|<=)(.*)/) {
1967 # warn "handling leaf... left:$left operator:$operator right:$right"
1969 unless ($operator) {
1970 if ($string =~ /(.*)(>|<|=)(.*)/) {
1975 "handling unless (operator)... left:$left operator:$operator right:$right"
1983 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1986 # automatic replace for short operators
1987 $left = 'title' if $left =~ '^ti$';
1988 $left = 'author' if $left =~ '^au$';
1989 $left = 'publisher' if $left =~ '^pb$';
1990 $left = 'subject' if $left =~ '^su$';
1991 $left = 'koha-Auth-Number' if $left =~ '^an$';
1992 $left = 'keyword' if $left =~ '^kw$';
1993 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1994 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1995 my $dbh = C4::Context->dbh;
1996 if ( $operator && $left ne 'keyword' ) {
1997 #do a specific search
1998 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1999 my $sth = $dbh->prepare(
2000 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2002 warn "$left / $operator / $right\n" if $DEBUG;
2004 # split each word, query the DB and build the biblionumbers result
2005 #sanitizing leftpart
2006 $left =~ s/^\s+|\s+$//;
2007 foreach ( split / /, $right ) {
2009 $_ =~ s/^\s+|\s+$//;
2011 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2012 $sth->execute( $server, $left, $_ )
2013 or warn "execute failed: $!";
2014 while ( my ( $line, $value ) = $sth->fetchrow ) {
2016 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2017 # otherwise, fill the result
2018 $biblionumbers .= $line
2019 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2020 warn "result : $value "
2021 . ( $right =~ /\d/ ) . "=="
2022 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2025 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2027 warn "NZAND" if $DEBUG;
2028 $results = NZoperatorAND($biblionumbers,$results);
2030 $results = $biblionumbers;
2035 #do a complete search (all indexes), if index='kw' do complete search too.
2036 my $sth = $dbh->prepare(
2037 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2040 # split each word, query the DB and build the biblionumbers result
2041 foreach ( split / /, $string ) {
2042 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2043 warn "search on all indexes on $_" if $DEBUG;
2046 $sth->execute( $server, $_ );
2047 while ( my $line = $sth->fetchrow ) {
2048 $biblionumbers .= $line;
2051 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2053 $results = NZoperatorAND($biblionumbers,$results);
2056 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2057 $results = $biblionumbers;
2061 warn "return : $results for LEAF : $string" if $DEBUG;
2064 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2068 my ($rightresult, $leftresult)=@_;
2070 my @leftresult = split /;/, $leftresult;
2071 warn " @leftresult / $rightresult \n" if $DEBUG;
2073 # my @rightresult = split /;/,$leftresult;
2076 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2077 # the result is stored twice, to have the same weight for AND than OR.
2078 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2079 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2080 foreach (@leftresult) {
2083 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2084 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2085 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2087 "$value-$countvalue;$value-$countvalue;";
2090 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2091 return $finalresult;
2095 my ($rightresult, $leftresult)=@_;
2096 return $rightresult.$leftresult;
2100 my ($leftresult, $rightresult)=@_;
2102 my @leftresult = split /;/, $leftresult;
2104 # my @rightresult = split /;/,$leftresult;
2106 foreach (@leftresult) {
2108 $value=$1 if $value=~m/(.*)-\d+$/;
2109 unless ($rightresult =~ "$value-") {
2110 $finalresult .= "$_;";
2113 return $finalresult;
2118 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2125 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2126 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2128 # order title asc by default
2129 # $ordering = '1=36 <i' unless $ordering;
2130 $results_per_page = 20 unless $results_per_page;
2131 $offset = 0 unless $offset;
2132 my $dbh = C4::Context->dbh;
2135 # order by POPULARITY
2137 if ( $ordering =~ /popularity/ ) {
2141 # popularity is not in MARC record, it's builded from a specific query
2143 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2144 foreach ( split /;/, $biblionumbers ) {
2145 my ( $biblionumber, $title ) = split /,/, $_;
2146 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2147 $sth->execute($biblionumber);
2148 my $popularity = $sth->fetchrow || 0;
2150 # hint : the key is popularity.title because we can have
2151 # many results with the same popularity. In this case, sub-ordering is done by title
2152 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2153 # (un-frequent, I agree, but we won't forget anything that way ;-)
2154 $popularity{ sprintf( "%10d", $popularity ) . $title
2155 . $biblionumber } = $biblionumber;
2158 # sort the hash and return the same structure as GetRecords (Zebra querying)
2161 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2162 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2163 $result_hash->{'RECORDS'}[ $numbers++ ] =
2164 $result{ $popularity{$key} }->as_usmarc();
2167 else { # sort popularity ASC
2168 foreach my $key ( sort ( keys %popularity ) ) {
2169 $result_hash->{'RECORDS'}[ $numbers++ ] =
2170 $result{ $popularity{$key} }->as_usmarc();
2173 my $finalresult = ();
2174 $result_hash->{'hits'} = $numbers;
2175 $finalresult->{'biblioserver'} = $result_hash;
2176 return $finalresult;
2182 elsif ( $ordering =~ /author/ ) {
2184 foreach ( split /;/, $biblionumbers ) {
2185 my ( $biblionumber, $title ) = split /,/, $_;
2186 my $record = GetMarcBiblio($biblionumber);
2188 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2189 $author = $record->subfield( '200', 'f' );
2190 $author = $record->subfield( '700', 'a' ) unless $author;
2193 $author = $record->subfield( '100', 'a' );
2196 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2197 # and we don't want to get only 1 result for each of them !!!
2198 $result{ $author . $biblionumber } = $record;
2201 # sort the hash and return the same structure as GetRecords (Zebra querying)
2204 if ( $ordering eq 'author_za' ) { # sort by author desc
2205 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2206 $result_hash->{'RECORDS'}[ $numbers++ ] =
2207 $result{$key}->as_usmarc();
2210 else { # sort by author ASC
2211 foreach my $key ( sort ( keys %result ) ) {
2212 $result_hash->{'RECORDS'}[ $numbers++ ] =
2213 $result{$key}->as_usmarc();
2216 my $finalresult = ();
2217 $result_hash->{'hits'} = $numbers;
2218 $finalresult->{'biblioserver'} = $result_hash;
2219 return $finalresult;
2222 # ORDER BY callnumber
2225 elsif ( $ordering =~ /callnumber/ ) {
2227 foreach ( split /;/, $biblionumbers ) {
2228 my ( $biblionumber, $title ) = split /,/, $_;
2229 my $record = GetMarcBiblio($biblionumber);
2231 my $frameworkcode = GetFrameworkCode($biblionumber);
2232 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2233 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2234 unless $callnumber_tag;
2235 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2236 $callnumber = $record->subfield( '200', 'f' );
2238 $callnumber = $record->subfield( '100', 'a' );
2241 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2242 # and we don't want to get only 1 result for each of them !!!
2243 $result{ $callnumber . $biblionumber } = $record;
2246 # sort the hash and return the same structure as GetRecords (Zebra querying)
2249 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2250 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2251 $result_hash->{'RECORDS'}[ $numbers++ ] =
2252 $result{$key}->as_usmarc();
2255 else { # sort by title ASC
2256 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2257 $result_hash->{'RECORDS'}[ $numbers++ ] =
2258 $result{$key}->as_usmarc();
2261 my $finalresult = ();
2262 $result_hash->{'hits'} = $numbers;
2263 $finalresult->{'biblioserver'} = $result_hash;
2264 return $finalresult;
2266 elsif ( $ordering =~ /pubdate/ ) { #pub year
2268 foreach ( split /;/, $biblionumbers ) {
2269 my ( $biblionumber, $title ) = split /,/, $_;
2270 my $record = GetMarcBiblio($biblionumber);
2271 my ( $publicationyear_tag, $publicationyear_subfield ) =
2272 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2273 my $publicationyear =
2274 $record->subfield( $publicationyear_tag,
2275 $publicationyear_subfield );
2277 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2278 # and we don't want to get only 1 result for each of them !!!
2279 $result{ $publicationyear . $biblionumber } = $record;
2282 # sort the hash and return the same structure as GetRecords (Zebra querying)
2285 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2286 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2287 $result_hash->{'RECORDS'}[ $numbers++ ] =
2288 $result{$key}->as_usmarc();
2291 else { # sort by pub year ASC
2292 foreach my $key ( sort ( keys %result ) ) {
2293 $result_hash->{'RECORDS'}[ $numbers++ ] =
2294 $result{$key}->as_usmarc();
2297 my $finalresult = ();
2298 $result_hash->{'hits'} = $numbers;
2299 $finalresult->{'biblioserver'} = $result_hash;
2300 return $finalresult;
2306 elsif ( $ordering =~ /title/ ) {
2308 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2310 foreach ( split /;/, $biblionumbers ) {
2311 my ( $biblionumber, $title ) = split /,/, $_;
2313 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2314 # and we don't want to get only 1 result for each of them !!!
2315 # hint & speed improvement : we can order without reading the record
2316 # so order, and read records only for the requested page !
2317 $result{ $title . $biblionumber } = $biblionumber;
2320 # sort the hash and return the same structure as GetRecords (Zebra querying)
2323 if ( $ordering eq 'title_az' ) { # sort by title desc
2324 foreach my $key ( sort ( keys %result ) ) {
2325 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2328 else { # sort by title ASC
2329 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2330 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2334 # limit the $results_per_page to result size if it's more
2335 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2337 # for the requested page, replace biblionumber by the complete record
2338 # speed improvement : avoid reading too much things
2340 my $counter = $offset ;
2341 $counter <= $offset + $results_per_page ;
2345 $result_hash->{'RECORDS'}[$counter] =
2346 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2348 my $finalresult = ();
2349 $result_hash->{'hits'} = $numbers;
2350 $finalresult->{'biblioserver'} = $result_hash;
2351 return $finalresult;
2358 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2361 foreach ( split /;/, $biblionumbers ) {
2362 my ( $biblionumber, $title ) = split /,/, $_;
2363 $title =~ /(.*)-(\d)/;
2368 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2369 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2370 # biblio N has ranking = 6
2371 $count_ranking{$biblionumber} += $ranking;
2374 # build the result by "inverting" the count_ranking hash
2375 # 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
2377 foreach ( keys %count_ranking ) {
2378 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2381 # sort the hash and return the same structure as GetRecords (Zebra querying)
2384 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2385 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2388 # limit the $results_per_page to result size if it's more
2389 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2391 # for the requested page, replace biblionumber by the complete record
2392 # speed improvement : avoid reading too much things
2394 my $counter = $offset ;
2395 $counter <= $offset + $results_per_page ;
2399 $result_hash->{'RECORDS'}[$counter] =
2400 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2401 if $result_hash->{'RECORDS'}[$counter];
2403 my $finalresult = ();
2404 $result_hash->{'hits'} = $numbers;
2405 $finalresult->{'biblioserver'} = $result_hash;
2406 return $finalresult;
2410 =head2 enabled_staff_search_views
2412 %hash = enabled_staff_search_views()
2414 This function returns a hash that contains three flags obtained from the system
2415 preferences, used to determine whether a particular staff search results view
2420 =item C<Output arg:>
2422 * $hash{can_view_MARC} is true only if the MARC view is enabled
2423 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2424 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2426 =item C<usage in the script:>
2430 $template->param ( C4::Search::enabled_staff_search_views );
2434 sub enabled_staff_search_views
2437 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2438 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2439 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2443 sub AddSearchHistory{
2444 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2445 my $dbh = C4::Context->dbh;
2447 # Add the request the user just made
2448 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2449 my $sth = $dbh->prepare($sql);
2450 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2451 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2454 sub GetSearchHistory{
2455 my ($borrowernumber,$session)=@_;
2456 my $dbh = C4::Context->dbh;
2458 # Add the request the user just made
2459 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2460 my $sth = $dbh->prepare($query);
2461 $sth->execute($borrowernumber, $session);
2462 return $sth->fetchall_hashref({});
2465 =head2 z3950_search_args
2467 $arrayref = z3950_search_args($matchpoints)
2469 This function returns an array reference that contains the search parameters to be
2470 passed to the Z39.50 search script (z3950_search.pl). The array elements
2471 are hash refs whose keys are name, value and encvalue, and whose values are the
2472 name of a search parameter, the value of that search parameter and the URL encoded
2473 value of that parameter.
2475 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2477 The search parameter values are obtained from the bibliographic record whose
2478 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2480 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2481 a general purpose search argument. In this case, the returned array contains only
2482 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2484 If a search parameter value is undefined or empty, it is not included in the returned
2487 The returned array reference may be passed directly to the template parameters.
2491 =item C<Output arg:>
2493 * $array containing hash refs as described above
2495 =item C<usage in the script:>
2499 $data = Biblio::GetBiblioData($bibno);
2500 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2504 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2508 sub z3950_search_args {
2510 $bibrec = { title => $bibrec } if !ref $bibrec;
2512 for my $field (qw/ lccn isbn issn title author dewey subject /)
2514 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2515 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2520 =head2 BiblioAddAuthorities
2522 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2524 this function finds the authorities linked to the biblio
2525 * search in the authority DB for the same authid (in $9 of the biblio)
2526 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2527 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2528 OR adds a new authority record
2534 * $record is the MARC record in question (marc blob)
2535 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2537 =item C<Output arg:>
2539 * $countlinked is the number of authorities records that are linked to this authority
2543 * 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)
2550 sub BiblioAddAuthorities{
2551 my ( $record, $frameworkcode ) = @_;
2552 my $dbh=C4::Context->dbh;
2553 my $query=$dbh->prepare(qq|
2554 SELECT authtypecode,tagfield
2555 FROM marc_subfield_structure
2556 WHERE frameworkcode=?
2557 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2558 # SELECT authtypecode,tagfield
2559 # FROM marc_subfield_structure
2560 # WHERE frameworkcode=?
2561 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2562 $query->execute($frameworkcode);
2563 my ($countcreated,$countlinked);
2564 while (my $data=$query->fetchrow_hashref){
2565 foreach my $field ($record->field($data->{tagfield})){
2566 next if ($field->subfield('3')||$field->subfield('9'));
2567 # No authorities id in the tag.
2568 # Search if there is any authorities to link to.
2569 my $query='at='.$data->{authtypecode}.' ';
2570 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2571 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2572 # there is only 1 result
2574 warn "BIBLIOADDSAUTHORITIES: $error";
2577 if ($results && scalar(@$results)==1) {
2578 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2579 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2581 } elsif (scalar(@$results)>1) {
2582 #More than One result
2583 #This can comes out of a lack of a subfield.
2584 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2585 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2588 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2589 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2590 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2591 my $authtypedata=C4::AuthoritiesMarc::GetAuthType($data->{authtypecode});
2592 next unless $authtypedata;
2593 my $marcrecordauth=MARC::Record->new();
2594 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2595 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2596 $marcrecordauth->insert_fields_ordered($authfield);
2598 # bug 2317: ensure new authority knows it's using UTF-8; currently
2599 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2600 # automatically for UNIMARC (by not transcoding)
2601 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2602 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2603 # of change to a core API just before the 3.0 release.
2604 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2605 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2608 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2610 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2612 $field->add_subfields('9'=>$authid);
2616 return ($countlinked,$countcreated);
2619 =head2 GetDistinctValues($field);
2621 C<$field> is a reference to the fields array
2625 sub GetDistinctValues {
2626 my ($fieldname,$string)=@_;
2627 # returns a reference to a hash of references to branches...
2628 if ($fieldname=~/\./){
2629 my ($table,$column)=split /\./, $fieldname;
2630 my $dbh = C4::Context->dbh;
2631 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2632 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 ");
2634 my $elements=$sth->fetchall_arrayref({});
2639 my @servers=qw<biblioserver authorityserver>;
2640 my (@zconns,@results);
2641 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2642 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2645 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2648 # The big moment: asynchronously retrieve results from all servers
2650 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2651 my $ev = $zconns[ $i - 1 ]->last_event();
2652 if ( $ev == ZOOM::Event::ZEND ) {
2653 next unless $results[ $i - 1 ];
2654 my $size = $results[ $i - 1 ]->size();
2656 for (my $j=0;$j<$size;$j++){
2658 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2659 push @elements, \%hashscan;
2669 END { } # module clean-up code here (global destructor)
2676 Koha Development Team <http://koha-community.org/>