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
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
25 use C4::Search::PazPar2;
27 use C4::Dates qw(format_date);
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 $debug && warn "$_ Dump($operand)";
654 if ( my ($matched) = ($operand =~
655 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
657 $operand =~ s/\Q$matched\E/ /gi;
658 push @stopwords_removed, $_;
662 return ( $operand, \@stopwords_removed );
666 sub _detect_truncation {
667 my ( $operand, $index ) = @_;
668 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
671 my @wordlist = split( /\s/, $operand );
672 foreach my $word (@wordlist) {
673 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
674 push @rightlefttruncated, $word;
676 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
677 push @lefttruncated, $word;
679 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
680 push @righttruncated, $word;
682 elsif ( index( $word, "*" ) < 0 ) {
683 push @nontruncated, $word;
686 push @regexpr, $word;
690 \@nontruncated, \@righttruncated, \@lefttruncated,
691 \@rightlefttruncated, \@regexpr
696 sub _build_stemmed_operand {
697 my ($operand,$lang) = @_;
698 require Lingua::Stem::Snowball ;
701 # If operand contains a digit, it is almost certainly an identifier, and should
702 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
703 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
704 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
705 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
706 return $operand if $operand =~ /\d/;
708 # FIXME: the locale should be set based on the user's language and/or search choice
710 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
711 encoding => "UTF-8" );
713 my @words = split( / /, $operand );
714 my @stems = $stemmer->stem(\@words);
715 for my $stem (@stems) {
716 $stemmed_operand .= "$stem";
717 $stemmed_operand .= "?"
718 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
719 $stemmed_operand .= " ";
721 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
722 return $stemmed_operand;
726 sub _build_weighted_query {
728 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
729 # pretty well but could work much better if we had a smarter query parser
730 my ( $operand, $stemmed_operand, $index ) = @_;
731 my $stemming = C4::Context->preference("QueryStemming") || 0;
732 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
733 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
735 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
737 # Keyword, or, no index specified
738 if ( ( $index eq 'kw' ) || ( !$index ) ) {
740 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
741 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
742 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
743 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
744 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
745 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
746 if $fuzzy_enabled; # add fuzzy, word list
747 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
748 if ( $stemming and $stemmed_operand )
749 ; # add stemming, right truncation
750 $weighted_query .= " or wrdl,r9=\"$operand\"";
752 # embedded sorting: 0 a-z; 1 z-a
753 # $weighted_query .= ") or (sort1,aut=1";
756 # Barcode searches should skip this process
757 elsif ( $index eq 'bc' ) {
758 $weighted_query .= "bc=\"$operand\"";
761 # Authority-number searches should skip this process
762 elsif ( $index eq 'an' ) {
763 $weighted_query .= "an=\"$operand\"";
766 # If the index already has more than one qualifier, wrap the operand
767 # in quotes and pass it back (assumption is that the user knows what they
768 # are doing and won't appreciate us mucking up their query
769 elsif ( $index =~ ',' ) {
770 $weighted_query .= " $index=\"$operand\"";
773 #TODO: build better cases based on specific search indexes
775 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
776 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
777 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
779 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
782 $weighted_query .= "))"; # close rank specification
783 return $weighted_query;
788 Return an array with available indexes.
810 'Author-personal-bibliography',
819 'Chronological-subdivision',
829 'Conference-name-heading',
830 'Conference-name-see',
831 'Conference-name-seealso',
836 'Corporate-name-heading',
837 'Corporate-name-see',
838 'Corporate-name-seealso',
840 'date-entered-on-file',
841 'Date-of-acquisition',
842 'Date-of-publication',
843 'Dewey-classification',
849 'Geographic-subdivision',
852 'Heading-use-main-or-added-entry',
853 'Heading-use-series-added-entry ',
854 'Heading-use-subject-added-entry',
870 'Local-classification',
873 'Match-heading-see-from',
879 'Name-geographic-heading',
880 'Name-geographic-see',
881 'Name-geographic-seealso',
888 'Personal-name-heading',
890 'Personal-name-seealso',
906 'Subject-heading-thesaurus',
907 'Subject-name-personal',
908 'Subject-subdivision',
917 'Term-genre-form-heading',
918 'Term-genre-form-see',
919 'Term-genre-form-seealso',
925 'Title-uniform-heading',
927 'Title-uniform-seealso',
937 'classification-source',
939 '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] ) {
1081 # A flag to determine whether or not to add the index to the query
1084 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1085 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1088 $remove_stopwords = 0;
1090 my $operand = $operands[$i];
1091 my $index = $indexes[$i];
1093 # Add index-specific attributes
1094 # Date of Publication
1095 if ( $index eq 'yr' ) {
1096 $index .= ",st-numeric";
1098 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1101 # Date of Acquisition
1102 elsif ( $index eq 'acqdate' ) {
1103 $index .= ",st-date-normalized";
1105 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1107 # ISBN,ISSN,Standard Number, don't need special treatment
1108 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1111 $stemming, $auto_truncation,
1112 $weight_fields, $fuzzy_enabled,
1114 ) = ( 0, 0, 0, 0, 0 );
1122 # Set default structure attribute (word list)
1124 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1125 $struct_attr = ",wrdl";
1128 # Some helpful index variants
1129 my $index_plus = $index . $struct_attr . ":" if $index;
1130 my $index_plus_comma = $index . $struct_attr . "," if $index;
1133 if ($remove_stopwords) {
1134 ( $operand, $stopwords_removed ) =
1135 _remove_stopwords( $operand, $index );
1136 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1137 warn "REMOVED STOPWORDS: @$stopwords_removed"
1138 if ( $stopwords_removed && $DEBUG );
1141 if ($auto_truncation){
1142 unless ( $index =~ /(st-|phr|ext)/ ) {
1143 #FIXME only valid with LTR scripts
1144 $operand=join(" ",map{
1145 (index($_,"*")>0?"$_":"$_*")
1146 }split (/\s+/,$operand));
1147 warn $operand if $DEBUG;
1152 my $truncated_operand;
1153 my( $nontruncated, $righttruncated, $lefttruncated,
1154 $rightlefttruncated, $regexpr
1155 ) = _detect_truncation( $operand, $index );
1157 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1162 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1163 scalar(@$rightlefttruncated) > 0 )
1166 # Don't field weight or add the index to the query, we do it here
1168 undef $weight_fields;
1169 my $previous_truncation_operand;
1170 if (scalar @$nontruncated) {
1171 $truncated_operand .= "$index_plus @$nontruncated ";
1172 $previous_truncation_operand = 1;
1174 if (scalar @$righttruncated) {
1175 $truncated_operand .= "and " if $previous_truncation_operand;
1176 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1177 $previous_truncation_operand = 1;
1179 if (scalar @$lefttruncated) {
1180 $truncated_operand .= "and " if $previous_truncation_operand;
1181 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1182 $previous_truncation_operand = 1;
1184 if (scalar @$rightlefttruncated) {
1185 $truncated_operand .= "and " if $previous_truncation_operand;
1186 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1187 $previous_truncation_operand = 1;
1190 $operand = $truncated_operand if $truncated_operand;
1191 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1194 my $stemmed_operand;
1195 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1198 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1200 # Handle Field Weighting
1201 my $weighted_operand;
1202 if ($weight_fields) {
1203 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1204 $operand = $weighted_operand;
1208 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1210 # If there's a previous operand, we need to add an operator
1211 if ($previous_operand) {
1213 # User-specified operator
1214 if ( $operators[ $i - 1 ] ) {
1215 $query .= " $operators[$i-1] ";
1216 $query .= " $index_plus " unless $indexes_set;
1217 $query .= " $operand";
1218 $query_cgi .= "&op=$operators[$i-1]";
1219 $query_cgi .= "&idx=$index" if $index;
1220 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1222 " $operators[$i-1] $index_plus $operands[$i]";
1225 # Default operator is and
1228 $query .= "$index_plus " unless $indexes_set;
1229 $query .= "$operand";
1230 $query_cgi .= "&op=and&idx=$index" if $index;
1231 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1232 $query_desc .= " and $index_plus $operands[$i]";
1236 # There isn't a pervious operand, don't need an operator
1239 # Field-weighted queries already have indexes set
1240 $query .= " $index_plus " unless $indexes_set;
1242 $query_desc .= " $index_plus $operands[$i]";
1243 $query_cgi .= "&idx=$index" if $index;
1244 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1245 $previous_operand = 1;
1250 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1253 my $group_OR_limits;
1254 my $availability_limit;
1255 foreach my $this_limit (@limits) {
1256 # if ( $this_limit =~ /available/ ) {
1258 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1260 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1261 # $availability_limit .=
1262 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1263 # $limit_cgi .= "&limit=available";
1264 # $limit_desc .= "";
1267 # group_OR_limits, prefixed by mc-
1268 # OR every member of the group
1269 # elsif ( $this_limit =~ /mc/ ) {
1270 if ( $this_limit =~ /mc/ ) {
1271 $group_OR_limits .= " or " if $group_OR_limits;
1272 $limit_desc .= " or " if $group_OR_limits;
1273 $group_OR_limits .= "$this_limit";
1274 $limit_cgi .= "&limit=$this_limit";
1275 $limit_desc .= " $this_limit";
1278 # Regular old limits
1280 $limit .= " and " if $limit || $query;
1281 $limit .= "$this_limit";
1282 $limit_cgi .= "&limit=$this_limit";
1283 if ($this_limit =~ /^branch:(.+)/) {
1284 my $branchcode = $1;
1285 my $branchname = GetBranchName($branchcode);
1286 if (defined $branchname) {
1287 $limit_desc .= " branch:$branchname";
1289 $limit_desc .= " $this_limit";
1292 $limit_desc .= " $this_limit";
1296 if ($group_OR_limits) {
1297 $limit .= " and " if ( $query || $limit );
1298 $limit .= "($group_OR_limits)";
1300 if ($availability_limit) {
1301 $limit .= " and " if ( $query || $limit );
1302 $limit .= "($availability_limit)";
1305 # Normalize the query and limit strings
1306 # This is flawed , means we can't search anything with : in it
1307 # if user wants to do ccl or cql, start the query with that
1310 for ( $query, $query_desc, $limit, $limit_desc ) {
1311 s/ / /g; # remove extra spaces
1312 s/^ //g; # remove any beginning spaces
1313 s/ $//g; # remove any ending spaces
1314 s/==/=/g; # remove double == from query
1316 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1318 for ($query_cgi,$simple_query) {
1321 # append the limit to the query
1322 $query .= " " . $limit;
1326 warn "QUERY:" . $query;
1327 warn "QUERY CGI:" . $query_cgi;
1328 warn "QUERY DESC:" . $query_desc;
1329 warn "LIMIT:" . $limit;
1330 warn "LIMIT CGI:" . $limit_cgi;
1331 warn "LIMIT DESC:" . $limit_desc;
1332 warn "---------\nLeave buildQuery\n---------";
1335 undef, $query, $simple_query, $query_cgi,
1336 $query_desc, $limit, $limit_cgi, $limit_desc,
1337 $stopwords_removed, $query_type
1341 =head2 searchResults
1343 Format results in a form suitable for passing to the template
1347 # IMO this subroutine is pretty messy still -- it's responsible for
1348 # building the HTML output for the template
1350 my ( $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1351 my $dbh = C4::Context->dbh;
1354 #Build branchnames hash
1356 #get branch information.....
1358 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1360 while ( my $bdata = $bsth->fetchrow_hashref ) {
1361 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1363 # FIXME - We build an authorised values hash here, using the default framework
1364 # though it is possible to have different authvals for different fws.
1366 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1368 # get notforloan authorised value list (see $shelflocations FIXME)
1369 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1371 #Build itemtype hash
1372 #find itemtype & itemtype image
1376 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1379 while ( my $bdata = $bsth->fetchrow_hashref ) {
1380 foreach (qw(description imageurl summary notforloan)) {
1381 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1385 #search item field code
1388 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1391 my ($itemtag) = $sth->fetchrow;
1393 ## find column names of items related to MARC
1394 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1396 my %subfieldstosearch;
1397 while ( ( my $column ) = $sth2->fetchrow ) {
1398 my ( $tagfield, $tagsubfield ) =
1399 &GetMarcFromKohaField( "items." . $column, "" );
1400 $subfieldstosearch{$column} = $tagsubfield;
1403 # handle which records to actually retrieve
1405 if ( $hits && $offset + $results_per_page <= $hits ) {
1406 $times = $offset + $results_per_page;
1409 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1412 my $marcflavour = C4::Context->preference("marcflavour");
1413 # We get the biblionumber position in MARC
1414 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1417 # loop through all of the records we've retrieved
1418 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1419 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1421 $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1423 $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1426 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1427 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1428 $oldbiblio->{result_number} = $i + 1;
1430 # add imageurl to itemtype if there is one
1431 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1433 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1434 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1435 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1436 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1437 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1438 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1440 # edition information, if any
1441 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1442 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1443 # Build summary if there is one (the summary is defined in the itemtypes table)
1444 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1445 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1446 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1447 my @fields = $marcrecord->fields();
1450 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1452 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1453 $tag =~ /(.{3})(.)/;
1454 if($marcrecord->field($1)){
1455 my @abc = $marcrecord->field($1)->subfield($2);
1456 $tags->{$tag} = $#abc + 1 ;
1460 # We catch how many times to repeat this line
1462 foreach my $tag (keys(%$tags)){
1463 $max = $tags->{$tag} if($tags->{$tag} > $max);
1466 # we replace, and repeat each line
1467 for (my $i = 0 ; $i < $max ; $i++){
1468 my $newline = $line;
1470 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1471 $tag =~ /(.{3})(.)/;
1473 if($marcrecord->field($1)){
1474 my @repl = $marcrecord->field($1)->subfield($2);
1475 my $subfieldvalue = $repl[$i];
1477 if (! utf8::is_utf8($subfieldvalue)) {
1478 utf8::decode($subfieldvalue);
1481 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1484 $newsummary .= "$newline\n";
1488 $newsummary =~ s/\[(.*?)]//g;
1489 $newsummary =~ s/\n/<br\/>/g;
1490 $oldbiblio->{summary} = $newsummary;
1493 # Pull out the items fields
1494 my @fields = $marcrecord->field($itemtag);
1496 # Setting item statuses for display
1497 my @available_items_loop;
1498 my @onloan_items_loop;
1499 my @other_items_loop;
1501 my $available_items;
1505 my $ordered_count = 0;
1506 my $available_count = 0;
1507 my $onloan_count = 0;
1508 my $longoverdue_count = 0;
1509 my $other_count = 0;
1510 my $wthdrawn_count = 0;
1511 my $itemlost_count = 0;
1512 my $itembinding_count = 0;
1513 my $itemdamaged_count = 0;
1514 my $item_in_transit_count = 0;
1515 my $can_place_holds = 0;
1516 my $items_count = scalar(@fields);
1518 ( C4::Context->preference('maxItemsinSearchResults') )
1519 ? C4::Context->preference('maxItemsinSearchResults') - 1
1522 # loop through every item
1523 foreach my $field (@fields) {
1526 # populate the items hash
1527 foreach my $code ( keys %subfieldstosearch ) {
1528 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1531 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1532 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1533 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1534 if ($item->{$hbranch}) {
1535 $item->{'branchname'} = $branches{$item->{$hbranch}};
1537 elsif ($item->{$otherbranch}) { # Last resort
1538 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1541 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1542 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1543 if ( $item->{onloan} ) {
1545 my $key = $prefix . $item->{onloan} . $item->{barcode};
1546 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1547 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1548 $onloan_items->{$key}->{branchname} = $item->{branchname};
1549 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1550 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1551 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1552 # if something's checked out and lost, mark it as 'long overdue'
1553 if ( $item->{itemlost} ) {
1554 $onloan_items->{$prefix}->{longoverdue}++;
1555 $longoverdue_count++;
1556 } else { # can place holds as long as item isn't lost
1557 $can_place_holds = 1;
1561 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1565 if ( $item->{notforloan} == -1 ) {
1569 # is item in transit?
1570 my $transfertwhen = '';
1571 my ($transfertfrom, $transfertto);
1573 unless ($item->{wthdrawn}
1574 || $item->{itemlost}
1576 || $item->{notforloan}
1577 || $items_count > 20) {
1579 # A couple heuristics to limit how many times
1580 # we query the database for item transfer information, sacrificing
1581 # accuracy in some cases for speed;
1583 # 1. don't query if item has one of the other statuses
1584 # 2. don't check transit status if the bib has
1585 # more than 20 items
1587 # FIXME: to avoid having the query the database like this, and to make
1588 # the in transit status count as unavailable for search limiting,
1589 # should map transit status to record indexed in Zebra.
1591 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1594 # item is withdrawn, lost or damaged
1595 if ( $item->{wthdrawn}
1596 || $item->{itemlost}
1598 || $item->{notforloan}
1599 || ($transfertwhen ne ''))
1601 $wthdrawn_count++ if $item->{wthdrawn};
1602 $itemlost_count++ if $item->{itemlost};
1603 $itemdamaged_count++ if $item->{damaged};
1604 $item_in_transit_count++ if $transfertwhen ne '';
1605 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1608 my $key = $prefix . $item->{status};
1609 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1610 $other_items->{$key}->{$_} = $item->{$_};
1612 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1613 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1614 $other_items->{$key}->{count}++ if $item->{$hbranch};
1615 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1616 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1620 $can_place_holds = 1;
1622 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1623 foreach (qw(branchname itemcallnumber)) {
1624 $available_items->{$prefix}->{$_} = $item->{$_};
1626 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1627 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1630 } # notforloan, item level and biblioitem level
1631 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1633 ( C4::Context->preference('maxItemsinSearchResults') )
1634 ? C4::Context->preference('maxItemsinSearchResults') - 1
1636 for my $key ( sort keys %$onloan_items ) {
1637 (++$onloanitemscount > $maxitems) and last;
1638 push @onloan_items_loop, $onloan_items->{$key};
1640 for my $key ( sort keys %$other_items ) {
1641 (++$otheritemscount > $maxitems) and last;
1642 push @other_items_loop, $other_items->{$key};
1644 for my $key ( sort keys %$available_items ) {
1645 (++$availableitemscount > $maxitems) and last;
1646 push @available_items_loop, $available_items->{$key}
1649 # XSLT processing of some stuff
1652 SetUTF8Flag($marcrecord);
1653 $debug && warn $marcrecord->as_formatted;
1654 if (C4::Context->preference("XSLTResultsDisplay") && !$scan) {
1655 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1656 $oldbiblio->{biblionumber}, $marcrecord, 'Results' );
1659 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1660 $can_place_holds = 0
1661 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1662 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1663 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1664 $oldbiblio->{items_count} = $items_count;
1665 $oldbiblio->{available_items_loop} = \@available_items_loop;
1666 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1667 $oldbiblio->{other_items_loop} = \@other_items_loop;
1668 $oldbiblio->{availablecount} = $available_count;
1669 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1670 $oldbiblio->{onloancount} = $onloan_count;
1671 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1672 $oldbiblio->{othercount} = $other_count;
1673 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1674 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1675 $oldbiblio->{itemlostcount} = $itemlost_count;
1676 $oldbiblio->{damagedcount} = $itemdamaged_count;
1677 $oldbiblio->{intransitcount} = $item_in_transit_count;
1678 $oldbiblio->{orderedcount} = $ordered_count;
1679 $oldbiblio->{isbn} =~
1680 s/-//g; # deleting - in isbn to enable amazon content
1681 push( @newresults, $oldbiblio )
1682 if(not $hidelostitems
1683 or (($items_count > $itemlost_count )
1684 && $hidelostitems));
1690 =head2 SearchAcquisitions
1691 Search for acquisitions
1694 sub SearchAcquisitions{
1695 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1697 my $dbh=C4::Context->dbh;
1698 # Variable initialization
1702 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1703 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1704 WHERE dateaccessioned BETWEEN ? AND ?
1707 my (@params,@loopcriteria);
1709 push @params, $datebegin->output("iso");
1710 push @params, $dateend->output("iso");
1712 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1713 if(C4::Context->preference("item-level_itypes")){
1714 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1716 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1718 push @params, @$itemtypes;
1721 if ($criteria =~/itemtype/){
1722 if(C4::Context->preference("item-level_itypes")){
1723 $str .= "AND items.itype=? ";
1725 $str .= "AND biblioitems.itemtype=? ";
1728 if(scalar(@$itemtypes) == 0){
1729 my $itypes = GetItemTypes();
1730 for my $key (keys %$itypes){
1731 push @$itemtypes, $key;
1735 @loopcriteria= @$itemtypes;
1736 }elsif ($criteria=~/itemcallnumber/){
1737 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1738 OR items.itemcallnumber is NULL
1739 OR items.itemcallnumber = '')";
1741 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1743 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1744 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1747 if ($orderby =~ /date_desc/){
1748 $str.=" ORDER BY dateaccessioned DESC";
1750 $str.=" ORDER BY title";
1753 my $qdataacquisitions=$dbh->prepare($str);
1755 my @loopacquisitions;
1756 foreach my $value(@loopcriteria){
1757 push @params,$value;
1759 $cell{"title"}=$value;
1760 $cell{"titlecode"}=$value;
1762 eval{$qdataacquisitions->execute(@params);};
1764 if ($@){ warn "recentacquisitions Error :$@";}
1767 while (my $data=$qdataacquisitions->fetchrow_hashref){
1768 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1770 $cell{"loopdata"}=\@loopdata;
1772 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1775 $qdataacquisitions->finish;
1776 return \@loopacquisitions;
1778 #----------------------------------------------------------------------
1780 # Non-Zebra GetRecords#
1781 #----------------------------------------------------------------------
1785 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1791 $query, $simple_query, $sort_by_ref, $servers_ref,
1792 $results_per_page, $offset, $expanded_facet, $branches,
1795 warn "query =$query" if $DEBUG;
1796 my $result = NZanalyse($query);
1797 warn "results =$result" if $DEBUG;
1799 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1805 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1806 the list is built from an inverted index in the nozebra SQL table
1807 note that title is here only for convenience : the sorting will be very fast when requested on title
1808 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1813 my ( $string, $server ) = @_;
1814 # warn "---------" if $DEBUG;
1815 warn " NZanalyse" if $DEBUG;
1816 # warn "---------" if $DEBUG;
1818 # $server contains biblioserver or authorities, depending on what we search on.
1819 #warn "querying : $string on $server";
1820 $server = 'biblioserver' unless $server;
1822 # if we have a ", replace the content to discard temporarily any and/or/not inside
1824 if ( $string =~ /"/ ) {
1825 $string =~ s/"(.*?)"/__X__/;
1827 warn "commacontent : $commacontent" if $DEBUG;
1830 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1831 # then, call again NZanalyse with $left and $right
1832 # (recursive until we find a leaf (=> something without and/or/not)
1833 # delete repeated operator... Would then go in infinite loop
1834 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1837 #process parenthesis before.
1838 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1841 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1843 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1845 my $leftresult = NZanalyse( $left, $server );
1847 my $rightresult = NZanalyse( $right, $server );
1849 # OK, we have the results for right and left part of the query
1850 # depending of operand, intersect, union or exclude both lists
1851 # to get a result list
1852 if ( $operator eq ' and ' ) {
1853 return NZoperatorAND($leftresult,$rightresult);
1855 elsif ( $operator eq ' or ' ) {
1857 # just merge the 2 strings
1858 return $leftresult . $rightresult;
1860 elsif ( $operator eq ' not ' ) {
1861 return NZoperatorNOT($leftresult,$rightresult);
1865 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1869 warn "string :" . $string if $DEBUG;
1873 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1876 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1878 warn "no parenthesis. left : $left operator: $operator right: $right"
1881 # it's not a leaf, we have a and/or/not
1884 # reintroduce comma content if needed
1885 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1886 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1887 warn "node : $left / $operator / $right\n" if $DEBUG;
1888 my $leftresult = NZanalyse( $left, $server );
1889 my $rightresult = NZanalyse( $right, $server );
1890 warn " leftresult : $leftresult" if $DEBUG;
1891 warn " rightresult : $rightresult" if $DEBUG;
1892 # OK, we have the results for right and left part of the query
1893 # depending of operand, intersect, union or exclude both lists
1894 # to get a result list
1895 if ( $operator eq ' and ' ) {
1897 return NZoperatorAND($leftresult,$rightresult);
1899 elsif ( $operator eq ' or ' ) {
1901 # just merge the 2 strings
1902 return $leftresult . $rightresult;
1904 elsif ( $operator eq ' not ' ) {
1905 return NZoperatorNOT($leftresult,$rightresult);
1909 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1910 die "error : operand unknown : $operator for $string";
1913 # it's a leaf, do the real SQL query and return the result
1916 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1917 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1918 #remove trailing blank at the beginning
1920 warn "leaf:$string" if $DEBUG;
1922 # parse the string in in operator/operand/value again
1926 if ($string =~ /(.*)(>=|<=)(.*)/) {
1933 # warn "handling leaf... left:$left operator:$operator right:$right"
1935 unless ($operator) {
1936 if ($string =~ /(.*)(>|<|=)(.*)/) {
1941 "handling unless (operator)... left:$left operator:$operator right:$right"
1949 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1952 # automatic replace for short operators
1953 $left = 'title' if $left =~ '^ti$';
1954 $left = 'author' if $left =~ '^au$';
1955 $left = 'publisher' if $left =~ '^pb$';
1956 $left = 'subject' if $left =~ '^su$';
1957 $left = 'koha-Auth-Number' if $left =~ '^an$';
1958 $left = 'keyword' if $left =~ '^kw$';
1959 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1960 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1961 my $dbh = C4::Context->dbh;
1962 if ( $operator && $left ne 'keyword' ) {
1963 #do a specific search
1964 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1965 my $sth = $dbh->prepare(
1966 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1968 warn "$left / $operator / $right\n" if $DEBUG;
1970 # split each word, query the DB and build the biblionumbers result
1971 #sanitizing leftpart
1972 $left =~ s/^\s+|\s+$//;
1973 foreach ( split / /, $right ) {
1975 $_ =~ s/^\s+|\s+$//;
1977 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1978 $sth->execute( $server, $left, $_ )
1979 or warn "execute failed: $!";
1980 while ( my ( $line, $value ) = $sth->fetchrow ) {
1982 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1983 # otherwise, fill the result
1984 $biblionumbers .= $line
1985 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
1986 warn "result : $value "
1987 . ( $right =~ /\d/ ) . "=="
1988 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
1991 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1993 warn "NZAND" if $DEBUG;
1994 $results = NZoperatorAND($biblionumbers,$results);
1996 $results = $biblionumbers;
2001 #do a complete search (all indexes), if index='kw' do complete search too.
2002 my $sth = $dbh->prepare(
2003 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2006 # split each word, query the DB and build the biblionumbers result
2007 foreach ( split / /, $string ) {
2008 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2009 warn "search on all indexes on $_" if $DEBUG;
2012 $sth->execute( $server, $_ );
2013 while ( my $line = $sth->fetchrow ) {
2014 $biblionumbers .= $line;
2017 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2019 $results = NZoperatorAND($biblionumbers,$results);
2022 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2023 $results = $biblionumbers;
2027 warn "return : $results for LEAF : $string" if $DEBUG;
2030 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2034 my ($rightresult, $leftresult)=@_;
2036 my @leftresult = split /;/, $leftresult;
2037 warn " @leftresult / $rightresult \n" if $DEBUG;
2039 # my @rightresult = split /;/,$leftresult;
2042 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2043 # the result is stored twice, to have the same weight for AND than OR.
2044 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2045 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2046 foreach (@leftresult) {
2049 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2050 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2051 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2053 "$value-$countvalue;$value-$countvalue;";
2056 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2057 return $finalresult;
2061 my ($rightresult, $leftresult)=@_;
2062 return $rightresult.$leftresult;
2066 my ($leftresult, $rightresult)=@_;
2068 my @leftresult = split /;/, $leftresult;
2070 # my @rightresult = split /;/,$leftresult;
2072 foreach (@leftresult) {
2074 $value=$1 if $value=~m/(.*)-\d+$/;
2075 unless ($rightresult =~ "$value-") {
2076 $finalresult .= "$_;";
2079 return $finalresult;
2084 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2091 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2092 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2094 # order title asc by default
2095 # $ordering = '1=36 <i' unless $ordering;
2096 $results_per_page = 20 unless $results_per_page;
2097 $offset = 0 unless $offset;
2098 my $dbh = C4::Context->dbh;
2101 # order by POPULARITY
2103 if ( $ordering =~ /popularity/ ) {
2107 # popularity is not in MARC record, it's builded from a specific query
2109 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2110 foreach ( split /;/, $biblionumbers ) {
2111 my ( $biblionumber, $title ) = split /,/, $_;
2112 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2113 $sth->execute($biblionumber);
2114 my $popularity = $sth->fetchrow || 0;
2116 # hint : the key is popularity.title because we can have
2117 # many results with the same popularity. In this case, sub-ordering is done by title
2118 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2119 # (un-frequent, I agree, but we won't forget anything that way ;-)
2120 $popularity{ sprintf( "%10d", $popularity ) . $title
2121 . $biblionumber } = $biblionumber;
2124 # sort the hash and return the same structure as GetRecords (Zebra querying)
2127 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2128 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2129 $result_hash->{'RECORDS'}[ $numbers++ ] =
2130 $result{ $popularity{$key} }->as_usmarc();
2133 else { # sort popularity ASC
2134 foreach my $key ( sort ( keys %popularity ) ) {
2135 $result_hash->{'RECORDS'}[ $numbers++ ] =
2136 $result{ $popularity{$key} }->as_usmarc();
2139 my $finalresult = ();
2140 $result_hash->{'hits'} = $numbers;
2141 $finalresult->{'biblioserver'} = $result_hash;
2142 return $finalresult;
2148 elsif ( $ordering =~ /author/ ) {
2150 foreach ( split /;/, $biblionumbers ) {
2151 my ( $biblionumber, $title ) = split /,/, $_;
2152 my $record = GetMarcBiblio($biblionumber);
2154 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2155 $author = $record->subfield( '200', 'f' );
2156 $author = $record->subfield( '700', 'a' ) unless $author;
2159 $author = $record->subfield( '100', 'a' );
2162 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2163 # and we don't want to get only 1 result for each of them !!!
2164 $result{ $author . $biblionumber } = $record;
2167 # sort the hash and return the same structure as GetRecords (Zebra querying)
2170 if ( $ordering eq 'author_za' ) { # sort by author desc
2171 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2172 $result_hash->{'RECORDS'}[ $numbers++ ] =
2173 $result{$key}->as_usmarc();
2176 else { # sort by author ASC
2177 foreach my $key ( sort ( keys %result ) ) {
2178 $result_hash->{'RECORDS'}[ $numbers++ ] =
2179 $result{$key}->as_usmarc();
2182 my $finalresult = ();
2183 $result_hash->{'hits'} = $numbers;
2184 $finalresult->{'biblioserver'} = $result_hash;
2185 return $finalresult;
2188 # ORDER BY callnumber
2191 elsif ( $ordering =~ /callnumber/ ) {
2193 foreach ( split /;/, $biblionumbers ) {
2194 my ( $biblionumber, $title ) = split /,/, $_;
2195 my $record = GetMarcBiblio($biblionumber);
2197 my $frameworkcode = GetFrameworkCode($biblionumber);
2198 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2199 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2200 unless $callnumber_tag;
2201 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2202 $callnumber = $record->subfield( '200', 'f' );
2204 $callnumber = $record->subfield( '100', 'a' );
2207 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2208 # and we don't want to get only 1 result for each of them !!!
2209 $result{ $callnumber . $biblionumber } = $record;
2212 # sort the hash and return the same structure as GetRecords (Zebra querying)
2215 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2216 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2217 $result_hash->{'RECORDS'}[ $numbers++ ] =
2218 $result{$key}->as_usmarc();
2221 else { # sort by title ASC
2222 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2223 $result_hash->{'RECORDS'}[ $numbers++ ] =
2224 $result{$key}->as_usmarc();
2227 my $finalresult = ();
2228 $result_hash->{'hits'} = $numbers;
2229 $finalresult->{'biblioserver'} = $result_hash;
2230 return $finalresult;
2232 elsif ( $ordering =~ /pubdate/ ) { #pub year
2234 foreach ( split /;/, $biblionumbers ) {
2235 my ( $biblionumber, $title ) = split /,/, $_;
2236 my $record = GetMarcBiblio($biblionumber);
2237 my ( $publicationyear_tag, $publicationyear_subfield ) =
2238 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2239 my $publicationyear =
2240 $record->subfield( $publicationyear_tag,
2241 $publicationyear_subfield );
2243 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2244 # and we don't want to get only 1 result for each of them !!!
2245 $result{ $publicationyear . $biblionumber } = $record;
2248 # sort the hash and return the same structure as GetRecords (Zebra querying)
2251 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2252 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2253 $result_hash->{'RECORDS'}[ $numbers++ ] =
2254 $result{$key}->as_usmarc();
2257 else { # sort by pub year ASC
2258 foreach my $key ( sort ( keys %result ) ) {
2259 $result_hash->{'RECORDS'}[ $numbers++ ] =
2260 $result{$key}->as_usmarc();
2263 my $finalresult = ();
2264 $result_hash->{'hits'} = $numbers;
2265 $finalresult->{'biblioserver'} = $result_hash;
2266 return $finalresult;
2272 elsif ( $ordering =~ /title/ ) {
2274 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2276 foreach ( split /;/, $biblionumbers ) {
2277 my ( $biblionumber, $title ) = split /,/, $_;
2279 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2280 # and we don't want to get only 1 result for each of them !!!
2281 # hint & speed improvement : we can order without reading the record
2282 # so order, and read records only for the requested page !
2283 $result{ $title . $biblionumber } = $biblionumber;
2286 # sort the hash and return the same structure as GetRecords (Zebra querying)
2289 if ( $ordering eq 'title_az' ) { # sort by title desc
2290 foreach my $key ( sort ( keys %result ) ) {
2291 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2294 else { # sort by title ASC
2295 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2296 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2300 # limit the $results_per_page to result size if it's more
2301 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2303 # for the requested page, replace biblionumber by the complete record
2304 # speed improvement : avoid reading too much things
2306 my $counter = $offset ;
2307 $counter <= $offset + $results_per_page ;
2311 $result_hash->{'RECORDS'}[$counter] =
2312 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2314 my $finalresult = ();
2315 $result_hash->{'hits'} = $numbers;
2316 $finalresult->{'biblioserver'} = $result_hash;
2317 return $finalresult;
2324 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2327 foreach ( split /;/, $biblionumbers ) {
2328 my ( $biblionumber, $title ) = split /,/, $_;
2329 $title =~ /(.*)-(\d)/;
2334 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2335 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2336 # biblio N has ranking = 6
2337 $count_ranking{$biblionumber} += $ranking;
2340 # build the result by "inverting" the count_ranking hash
2341 # 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
2343 foreach ( keys %count_ranking ) {
2344 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2347 # sort the hash and return the same structure as GetRecords (Zebra querying)
2350 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2351 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2354 # limit the $results_per_page to result size if it's more
2355 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2357 # for the requested page, replace biblionumber by the complete record
2358 # speed improvement : avoid reading too much things
2360 my $counter = $offset ;
2361 $counter <= $offset + $results_per_page ;
2365 $result_hash->{'RECORDS'}[$counter] =
2366 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2367 if $result_hash->{'RECORDS'}[$counter];
2369 my $finalresult = ();
2370 $result_hash->{'hits'} = $numbers;
2371 $finalresult->{'biblioserver'} = $result_hash;
2372 return $finalresult;
2376 =head2 enabled_staff_search_views
2378 %hash = enabled_staff_search_views()
2380 This function returns a hash that contains three flags obtained from the system
2381 preferences, used to determine whether a particular staff search results view
2386 =item C<Output arg:>
2388 * $hash{can_view_MARC} is true only if the MARC view is enabled
2389 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2390 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2392 =item C<usage in the script:>
2396 $template->param ( C4::Search::enabled_staff_search_views );
2400 sub enabled_staff_search_views
2403 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2404 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2405 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2409 sub AddSearchHistory{
2410 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2411 my $dbh = C4::Context->dbh;
2413 # Add the request the user just made
2414 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2415 my $sth = $dbh->prepare($sql);
2416 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2417 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2420 sub GetSearchHistory{
2421 my ($borrowernumber,$session)=@_;
2422 my $dbh = C4::Context->dbh;
2424 # Add the request the user just made
2425 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2426 my $sth = $dbh->prepare($query);
2427 $sth->execute($borrowernumber, $session);
2428 return $sth->fetchall_hashref({});
2431 =head2 z3950_search_args
2433 $arrayref = z3950_search_args($matchpoints)
2435 This function returns an array reference that contains the search parameters to be
2436 passed to the Z39.50 search script (z3950_search.pl). The array elements
2437 are hash refs whose keys are name, value and encvalue, and whose values are the
2438 name of a search parameter, the value of that search parameter and the URL encoded
2439 value of that parameter.
2441 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2443 The search parameter values are obtained from the bibliographic record whose
2444 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2446 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2447 a general purpose search argument. In this case, the returned array contains only
2448 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2450 If a search parameter value is undefined or empty, it is not included in the returned
2453 The returned array reference may be passed directly to the template parameters.
2457 =item C<Output arg:>
2459 * $array containing hash refs as described above
2461 =item C<usage in the script:>
2465 $data = Biblio::GetBiblioData($bibno);
2466 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2470 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2474 sub z3950_search_args {
2476 $bibrec = { title => $bibrec } if !ref $bibrec;
2478 for my $field (qw/ lccn isbn issn title author dewey subject /)
2480 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2481 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2486 =head2 BiblioAddAuthorities
2488 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2490 this function finds the authorities linked to the biblio
2491 * search in the authority DB for the same authid (in $9 of the biblio)
2492 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2493 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2494 OR adds a new authority record
2500 * $record is the MARC record in question (marc blob)
2501 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2503 =item C<Output arg:>
2505 * $countlinked is the number of authorities records that are linked to this authority
2509 * 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)
2515 sub BiblioAddAuthorities{
2516 my ( $record, $frameworkcode ) = @_;
2517 my $dbh=C4::Context->dbh;
2518 my $query=$dbh->prepare(qq|
2519 SELECT authtypecode,tagfield
2520 FROM marc_subfield_structure
2521 WHERE frameworkcode=?
2522 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2523 # SELECT authtypecode,tagfield
2524 # FROM marc_subfield_structure
2525 # WHERE frameworkcode=?
2526 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2527 $query->execute($frameworkcode);
2528 my ($countcreated,$countlinked);
2529 while (my $data=$query->fetchrow_hashref){
2530 foreach my $field ($record->field($data->{tagfield})){
2531 next if ($field->subfield('3')||$field->subfield('9'));
2532 # No authorities id in the tag.
2533 # Search if there is any authorities to link to.
2534 my $query='at='.$data->{authtypecode}.' ';
2535 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2536 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2537 # there is only 1 result
2539 warn "BIBLIOADDSAUTHORITIES: $error";
2542 if ($results && scalar(@$results)==1) {
2543 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2544 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2546 } elsif (scalar(@$results)>1) {
2547 #More than One result
2548 #This can comes out of a lack of a subfield.
2549 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2550 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2553 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2554 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2555 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2556 my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2557 next unless $authtypedata;
2558 my $marcrecordauth=MARC::Record->new();
2559 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2560 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2561 $marcrecordauth->insert_fields_ordered($authfield);
2563 # bug 2317: ensure new authority knows it's using UTF-8; currently
2564 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2565 # automatically for UNIMARC (by not transcoding)
2566 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2567 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2568 # of change to a core API just before the 3.0 release.
2569 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2570 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2573 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2575 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2577 $field->add_subfields('9'=>$authid);
2581 return ($countlinked,$countcreated);
2584 =head2 GetDistinctValues($field);
2586 C<$field> is a reference to the fields array
2590 sub GetDistinctValues {
2591 my ($fieldname,$string)=@_;
2592 # returns a reference to a hash of references to branches...
2593 if ($fieldname=~/\./){
2594 my ($table,$column)=split /\./, $fieldname;
2595 my $dbh = C4::Context->dbh;
2596 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column ";
2597 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 ");
2599 my $elements=$sth->fetchall_arrayref({});
2604 my @servers=qw<biblioserver authorityserver>;
2605 my (@zconns,@results);
2606 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2607 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2610 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2613 # The big moment: asynchronously retrieve results from all servers
2615 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2616 my $ev = $zconns[ $i - 1 ]->last_event();
2617 if ( $ev == ZOOM::Event::ZEND ) {
2618 next unless $results[ $i - 1 ];
2619 my $size = $results[ $i - 1 ]->size();
2621 for (my $j=0;$j<$size;$j++){
2623 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2624 push @elements, \%hashscan;
2634 END { } # module clean-up code here (global destructor)
2641 Koha Developement team <info@koha.org>