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;
789 $simple_query, $query_cgi,
791 $limit_cgi, $limit_desc,
792 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
794 Build queries and limits in CCL, CGI, Human,
795 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
797 See verbose embedded documentation.
803 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
805 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
808 my @operators = $operators ? @$operators : ();
809 my @indexes = $indexes ? @$indexes : ();
810 my @operands = $operands ? @$operands : ();
811 my @limits = $limits ? @$limits : ();
812 my @sort_by = $sort_by ? @$sort_by : ();
814 my $stemming = C4::Context->preference("QueryStemming") || 0;
815 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
816 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
817 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
818 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
820 # no stemming/weight/fuzzy in NoZebra
821 if ( C4::Context->preference("NoZebra") ) {
827 my $query = $operands[0];
828 my $simple_query = $operands[0];
830 # initialize the variables we're passing back
839 my $stopwords_removed; # flag to determine if stopwords have been removed
841 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
843 if ( $query =~ /^ccl=/ ) {
844 return ( undef, $', $', "q=ccl=$'", $', '', '', '', '', 'ccl' );
846 if ( $query =~ /^cql=/ ) {
847 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
849 if ( $query =~ /^pqf=/ ) {
850 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
853 # pass nested queries directly
854 # FIXME: need better handling of some of these variables in this case
855 # Nested queries aren't handled well and this implementation is flawed and causes users to be
856 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
857 # if ( $query =~ /(\(|\))/ ) {
859 # undef, $query, $simple_query, $query_cgi,
860 # $query, $limit, $limit_cgi, $limit_desc,
861 # $stopwords_removed, 'ccl'
865 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
866 # query operands and indexes and add stemming, truncation, field weighting, etc.
867 # Once we do so, we'll end up with a value in $query, just like if we had an
868 # incoming $query from the user
871 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
873 ; # a flag used to keep track if there was a previous query
874 # if there was, we can apply the current operator
876 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
878 # COMBINE OPERANDS, INDEXES AND OPERATORS
879 if ( $operands[$i] ) {
881 # A flag to determine whether or not to add the index to the query
884 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
885 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
888 $remove_stopwords = 0;
890 my $operand = $operands[$i];
891 my $index = $indexes[$i];
893 # Add index-specific attributes
894 # Date of Publication
895 if ( $index eq 'yr' ) {
896 $index .= ",st-numeric";
898 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
901 # Date of Acquisition
902 elsif ( $index eq 'acqdate' ) {
903 $index .= ",st-date-normalized";
905 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
907 # ISBN,ISSN,Standard Number, don't need special treatment
908 elsif ( $index eq 'nb' || $index eq 'ns' ) {
911 $stemming, $auto_truncation,
912 $weight_fields, $fuzzy_enabled,
914 ) = ( 0, 0, 0, 0, 0 );
917 # Set default structure attribute (word list)
919 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
920 $struct_attr = ",wrdl";
923 # Some helpful index variants
924 my $index_plus = $index . $struct_attr . ":" if $index;
925 my $index_plus_comma = $index . $struct_attr . "," if $index;
928 if ($remove_stopwords) {
929 ( $operand, $stopwords_removed ) =
930 _remove_stopwords( $operand, $index );
931 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
932 warn "REMOVED STOPWORDS: @$stopwords_removed"
933 if ( $stopwords_removed && $DEBUG );
936 if ($auto_truncation){
937 $operand=~join(" ",map{ "$_*" }split (/\s+/,$operand));
941 my $truncated_operand;
942 my( $nontruncated, $righttruncated, $lefttruncated,
943 $rightlefttruncated, $regexpr
944 ) = _detect_truncation( $operand, $index );
946 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
951 scalar(@$righttruncated) + scalar(@$lefttruncated) +
952 scalar(@$rightlefttruncated) > 0 )
955 # Don't field weight or add the index to the query, we do it here
957 undef $weight_fields;
958 my $previous_truncation_operand;
959 if (scalar @$nontruncated) {
960 $truncated_operand .= "$index_plus @$nontruncated ";
961 $previous_truncation_operand = 1;
963 if (scalar @$righttruncated) {
964 $truncated_operand .= "and " if $previous_truncation_operand;
965 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
966 $previous_truncation_operand = 1;
968 if (scalar @$lefttruncated) {
969 $truncated_operand .= "and " if $previous_truncation_operand;
970 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
971 $previous_truncation_operand = 1;
973 if (scalar @$rightlefttruncated) {
974 $truncated_operand .= "and " if $previous_truncation_operand;
975 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
976 $previous_truncation_operand = 1;
979 $operand = $truncated_operand if $truncated_operand;
980 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
984 $stemmed_operand = _build_stemmed_operand($operand, $lang)
987 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
989 # Handle Field Weighting
990 my $weighted_operand;
991 if ($weight_fields) {
992 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
993 $operand = $weighted_operand;
997 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
999 # If there's a previous operand, we need to add an operator
1000 if ($previous_operand) {
1002 # User-specified operator
1003 if ( $operators[ $i - 1 ] ) {
1004 $query .= " $operators[$i-1] ";
1005 $query .= " $index_plus " unless $indexes_set;
1006 $query .= " $operand";
1007 $query_cgi .= "&op=$operators[$i-1]";
1008 $query_cgi .= "&idx=$index" if $index;
1009 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1011 " $operators[$i-1] $index_plus $operands[$i]";
1014 # Default operator is and
1017 $query .= "$index_plus " unless $indexes_set;
1018 $query .= "$operand";
1019 $query_cgi .= "&op=and&idx=$index" if $index;
1020 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1021 $query_desc .= " and $index_plus $operands[$i]";
1025 # There isn't a pervious operand, don't need an operator
1028 # Field-weighted queries already have indexes set
1029 $query .= " $index_plus " unless $indexes_set;
1031 $query_desc .= " $index_plus $operands[$i]";
1032 $query_cgi .= "&idx=$index" if $index;
1033 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1034 $previous_operand = 1;
1039 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1042 my $group_OR_limits;
1043 my $availability_limit;
1044 foreach my $this_limit (@limits) {
1045 # if ( $this_limit =~ /available/ ) {
1047 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1049 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1050 # $availability_limit .=
1051 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1052 # $limit_cgi .= "&limit=available";
1053 # $limit_desc .= "";
1056 # group_OR_limits, prefixed by mc-
1057 # OR every member of the group
1058 # elsif ( $this_limit =~ /mc/ ) {
1059 if ( $this_limit =~ /mc/ ) {
1060 $group_OR_limits .= " or " if $group_OR_limits;
1061 $limit_desc .= " or " if $group_OR_limits;
1062 $group_OR_limits .= "$this_limit";
1063 $limit_cgi .= "&limit=$this_limit";
1064 $limit_desc .= " $this_limit";
1067 # Regular old limits
1069 $limit .= " and " if $limit || $query;
1070 $limit .= "$this_limit";
1071 $limit_cgi .= "&limit=$this_limit";
1072 if ($this_limit =~ /^branch:(.+)/) {
1073 my $branchcode = $1;
1074 my $branchname = GetBranchName($branchcode);
1075 if (defined $branchname) {
1076 $limit_desc .= " branch:$branchname";
1078 $limit_desc .= " $this_limit";
1081 $limit_desc .= " $this_limit";
1085 if ($group_OR_limits) {
1086 $limit .= " and " if ( $query || $limit );
1087 $limit .= "($group_OR_limits)";
1089 if ($availability_limit) {
1090 $limit .= " and " if ( $query || $limit );
1091 $limit .= "($availability_limit)";
1094 # Normalize the query and limit strings
1095 # This is flawed , means we can't search anything with : in it
1096 # if user wants to do ccl or cql, start the query with that
1097 # $query =~ s/:/=/g;
1099 for ( $query, $query_desc, $limit, $limit_desc ) {
1100 s/ / /g; # remove extra spaces
1101 s/^ //g; # remove any beginning spaces
1102 s/ $//g; # remove any ending spaces
1103 s/==/=/g; # remove double == from query
1105 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1107 for ($query_cgi,$simple_query) {
1110 # append the limit to the query
1111 $query .= " " . $limit;
1115 warn "QUERY:" . $query;
1116 warn "QUERY CGI:" . $query_cgi;
1117 warn "QUERY DESC:" . $query_desc;
1118 warn "LIMIT:" . $limit;
1119 warn "LIMIT CGI:" . $limit_cgi;
1120 warn "LIMIT DESC:" . $limit_desc;
1121 warn "---------\nLeave buildQuery\n---------";
1124 undef, $query, $simple_query, $query_cgi,
1125 $query_desc, $limit, $limit_cgi, $limit_desc,
1126 $stopwords_removed, $query_type
1130 =head2 searchResults
1132 Format results in a form suitable for passing to the template
1136 # IMO this subroutine is pretty messy still -- it's responsible for
1137 # building the HTML output for the template
1139 my ( $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1140 my $dbh = C4::Context->dbh;
1143 #Build branchnames hash
1145 #get branch information.....
1147 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1149 while ( my $bdata = $bsth->fetchrow_hashref ) {
1150 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1152 # FIXME - We build an authorised values hash here, using the default framework
1153 # though it is possible to have different authvals for different fws.
1155 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1157 # get notforloan authorised value list (see $shelflocations FIXME)
1158 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1160 #Build itemtype hash
1161 #find itemtype & itemtype image
1165 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1168 while ( my $bdata = $bsth->fetchrow_hashref ) {
1169 foreach (qw(description imageurl summary notforloan)) {
1170 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1174 #search item field code
1177 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1180 my ($itemtag) = $sth->fetchrow;
1182 ## find column names of items related to MARC
1183 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1185 my %subfieldstosearch;
1186 while ( ( my $column ) = $sth2->fetchrow ) {
1187 my ( $tagfield, $tagsubfield ) =
1188 &GetMarcFromKohaField( "items." . $column, "" );
1189 $subfieldstosearch{$column} = $tagsubfield;
1192 # handle which records to actually retrieve
1194 if ( $hits && $offset + $results_per_page <= $hits ) {
1195 $times = $offset + $results_per_page;
1198 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1201 my $marcflavour = C4::Context->preference("marcflavour");
1202 # We get the biblionumber position in MARC
1203 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1206 # loop through all of the records we've retrieved
1207 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1208 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1210 $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1212 $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1215 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1216 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1217 $oldbiblio->{result_number} = $i + 1;
1219 # add imageurl to itemtype if there is one
1220 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1222 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1223 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1224 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1225 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1226 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1227 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1229 # edition information, if any
1230 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1231 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1232 # Build summary if there is one (the summary is defined in the itemtypes table)
1233 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1234 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1235 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1236 my @fields = $marcrecord->fields();
1239 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1241 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1242 $tag =~ /(.{3})(.)/;
1243 if($marcrecord->field($1)){
1244 my @abc = $marcrecord->field($1)->subfield($2);
1245 $tags->{$tag} = $#abc + 1 ;
1249 # We catch how many times to repeat this line
1251 foreach my $tag (keys(%$tags)){
1252 $max = $tags->{$tag} if($tags->{$tag} > $max);
1255 # we replace, and repeat each line
1256 for (my $i = 0 ; $i < $max ; $i++){
1257 my $newline = $line;
1259 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1260 $tag =~ /(.{3})(.)/;
1262 if($marcrecord->field($1)){
1263 my @repl = $marcrecord->field($1)->subfield($2);
1264 my $subfieldvalue = $repl[$i];
1266 if (! utf8::is_utf8($subfieldvalue)) {
1267 utf8::decode($subfieldvalue);
1270 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1273 $newsummary .= "$newline\n";
1277 $newsummary =~ s/\[(.*?)]//g;
1278 $newsummary =~ s/\n/<br\/>/g;
1279 $oldbiblio->{summary} = $newsummary;
1282 # Pull out the items fields
1283 my @fields = $marcrecord->field($itemtag);
1285 # Setting item statuses for display
1286 my @available_items_loop;
1287 my @onloan_items_loop;
1288 my @other_items_loop;
1290 my $available_items;
1294 my $ordered_count = 0;
1295 my $available_count = 0;
1296 my $onloan_count = 0;
1297 my $longoverdue_count = 0;
1298 my $other_count = 0;
1299 my $wthdrawn_count = 0;
1300 my $itemlost_count = 0;
1301 my $itembinding_count = 0;
1302 my $itemdamaged_count = 0;
1303 my $item_in_transit_count = 0;
1304 my $can_place_holds = 0;
1305 my $items_count = scalar(@fields);
1307 ( C4::Context->preference('maxItemsinSearchResults') )
1308 ? C4::Context->preference('maxItemsinSearchResults') - 1
1311 # loop through every item
1312 foreach my $field (@fields) {
1315 # populate the items hash
1316 foreach my $code ( keys %subfieldstosearch ) {
1317 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1320 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1321 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1322 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1323 if ($item->{$hbranch}) {
1324 $item->{'branchname'} = $branches{$item->{$hbranch}};
1326 elsif ($item->{$otherbranch}) { # Last resort
1327 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1330 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1331 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1332 if ( $item->{onloan} ) {
1334 my $key = $prefix . $item->{onloan} . $item->{barcode};
1335 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1336 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1337 $onloan_items->{$key}->{branchname} = $item->{branchname};
1338 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1339 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1340 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1341 # if something's checked out and lost, mark it as 'long overdue'
1342 if ( $item->{itemlost} ) {
1343 $onloan_items->{$prefix}->{longoverdue}++;
1344 $longoverdue_count++;
1345 } else { # can place holds as long as item isn't lost
1346 $can_place_holds = 1;
1350 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1354 if ( $item->{notforloan} == -1 ) {
1358 # is item in transit?
1359 my $transfertwhen = '';
1360 my ($transfertfrom, $transfertto);
1362 unless ($item->{wthdrawn}
1363 || $item->{itemlost}
1365 || $item->{notforloan}
1366 || $items_count > 20) {
1368 # A couple heuristics to limit how many times
1369 # we query the database for item transfer information, sacrificing
1370 # accuracy in some cases for speed;
1372 # 1. don't query if item has one of the other statuses
1373 # 2. don't check transit status if the bib has
1374 # more than 20 items
1376 # FIXME: to avoid having the query the database like this, and to make
1377 # the in transit status count as unavailable for search limiting,
1378 # should map transit status to record indexed in Zebra.
1380 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1383 # item is withdrawn, lost or damaged
1384 if ( $item->{wthdrawn}
1385 || $item->{itemlost}
1387 || $item->{notforloan}
1388 || ($transfertwhen ne ''))
1390 $wthdrawn_count++ if $item->{wthdrawn};
1391 $itemlost_count++ if $item->{itemlost};
1392 $itemdamaged_count++ if $item->{damaged};
1393 $item_in_transit_count++ if $transfertwhen ne '';
1394 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1397 my $key = $prefix . $item->{status};
1398 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1399 $other_items->{$key}->{$_} = $item->{$_};
1401 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1402 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1403 $other_items->{$key}->{count}++ if $item->{$hbranch};
1404 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1405 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1409 $can_place_holds = 1;
1411 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1412 foreach (qw(branchname itemcallnumber)) {
1413 $available_items->{$prefix}->{$_} = $item->{$_};
1415 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1416 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1419 } # notforloan, item level and biblioitem level
1420 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1422 ( C4::Context->preference('maxItemsinSearchResults') )
1423 ? C4::Context->preference('maxItemsinSearchResults') - 1
1425 for my $key ( sort keys %$onloan_items ) {
1426 (++$onloanitemscount > $maxitems) and last;
1427 push @onloan_items_loop, $onloan_items->{$key};
1429 for my $key ( sort keys %$other_items ) {
1430 (++$otheritemscount > $maxitems) and last;
1431 push @other_items_loop, $other_items->{$key};
1433 for my $key ( sort keys %$available_items ) {
1434 (++$availableitemscount > $maxitems) and last;
1435 push @available_items_loop, $available_items->{$key}
1438 # XSLT processing of some stuff
1439 if (C4::Context->preference("XSLTResultsDisplay") && !$scan) {
1440 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1441 $oldbiblio->{biblionumber}, $marcrecord, 'Results' );
1444 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1445 $can_place_holds = 0
1446 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1447 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1448 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1449 $oldbiblio->{items_count} = $items_count;
1450 $oldbiblio->{available_items_loop} = \@available_items_loop;
1451 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1452 $oldbiblio->{other_items_loop} = \@other_items_loop;
1453 $oldbiblio->{availablecount} = $available_count;
1454 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1455 $oldbiblio->{onloancount} = $onloan_count;
1456 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1457 $oldbiblio->{othercount} = $other_count;
1458 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1459 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1460 $oldbiblio->{itemlostcount} = $itemlost_count;
1461 $oldbiblio->{damagedcount} = $itemdamaged_count;
1462 $oldbiblio->{intransitcount} = $item_in_transit_count;
1463 $oldbiblio->{orderedcount} = $ordered_count;
1464 $oldbiblio->{isbn} =~
1465 s/-//g; # deleting - in isbn to enable amazon content
1466 push( @newresults, $oldbiblio )
1467 if(not $hidelostitems
1468 or (($items_count > $itemlost_count )
1469 && $hidelostitems));
1475 =head2 SearchAcquisitions
1476 Search for acquisitions
1479 sub SearchAcquisitions{
1480 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1482 my $dbh=C4::Context->dbh;
1483 # Variable initialization
1487 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1488 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1489 WHERE dateaccessioned BETWEEN ? AND ?
1492 my (@params,@loopcriteria);
1494 push @params, $datebegin->output("iso");
1495 push @params, $dateend->output("iso");
1497 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1498 if(C4::Context->preference("item-level_itypes")){
1499 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1501 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1503 push @params, @$itemtypes;
1506 if ($criteria =~/itemtype/){
1507 if(C4::Context->preference("item-level_itypes")){
1508 $str .= "AND items.itype=? ";
1510 $str .= "AND biblioitems.itemtype=? ";
1513 if(scalar(@$itemtypes) == 0){
1514 my $itypes = GetItemTypes();
1515 for my $key (keys %$itypes){
1516 push @$itemtypes, $key;
1520 @loopcriteria= @$itemtypes;
1521 }elsif ($criteria=~/itemcallnumber/){
1522 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1523 OR items.itemcallnumber is NULL
1524 OR items.itemcallnumber = '')";
1526 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1528 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1529 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1532 if ($orderby =~ /date_desc/){
1533 $str.=" ORDER BY dateaccessioned DESC";
1535 $str.=" ORDER BY title";
1538 my $qdataacquisitions=$dbh->prepare($str);
1540 my @loopacquisitions;
1541 foreach my $value(@loopcriteria){
1542 push @params,$value;
1544 $cell{"title"}=$value;
1545 $cell{"titlecode"}=$value;
1547 eval{$qdataacquisitions->execute(@params);};
1549 if ($@){ warn "recentacquisitions Error :$@";}
1552 while (my $data=$qdataacquisitions->fetchrow_hashref){
1553 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1555 $cell{"loopdata"}=\@loopdata;
1557 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1560 $qdataacquisitions->finish;
1561 return \@loopacquisitions;
1563 #----------------------------------------------------------------------
1565 # Non-Zebra GetRecords#
1566 #----------------------------------------------------------------------
1570 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1576 $query, $simple_query, $sort_by_ref, $servers_ref,
1577 $results_per_page, $offset, $expanded_facet, $branches,
1580 warn "query =$query" if $DEBUG;
1581 my $result = NZanalyse($query);
1582 warn "results =$result" if $DEBUG;
1584 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1590 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1591 the list is built from an inverted index in the nozebra SQL table
1592 note that title is here only for convenience : the sorting will be very fast when requested on title
1593 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1598 my ( $string, $server ) = @_;
1599 # warn "---------" if $DEBUG;
1600 warn " NZanalyse" if $DEBUG;
1601 # warn "---------" if $DEBUG;
1603 # $server contains biblioserver or authorities, depending on what we search on.
1604 #warn "querying : $string on $server";
1605 $server = 'biblioserver' unless $server;
1607 # if we have a ", replace the content to discard temporarily any and/or/not inside
1609 if ( $string =~ /"/ ) {
1610 $string =~ s/"(.*?)"/__X__/;
1612 warn "commacontent : $commacontent" if $DEBUG;
1615 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1616 # then, call again NZanalyse with $left and $right
1617 # (recursive until we find a leaf (=> something without and/or/not)
1618 # delete repeated operator... Would then go in infinite loop
1619 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1622 #process parenthesis before.
1623 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1626 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1628 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1630 my $leftresult = NZanalyse( $left, $server );
1632 my $rightresult = NZanalyse( $right, $server );
1634 # OK, we have the results for right and left part of the query
1635 # depending of operand, intersect, union or exclude both lists
1636 # to get a result list
1637 if ( $operator eq ' and ' ) {
1638 return NZoperatorAND($leftresult,$rightresult);
1640 elsif ( $operator eq ' or ' ) {
1642 # just merge the 2 strings
1643 return $leftresult . $rightresult;
1645 elsif ( $operator eq ' not ' ) {
1646 return NZoperatorNOT($leftresult,$rightresult);
1650 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1654 warn "string :" . $string if $DEBUG;
1658 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1661 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1663 warn "no parenthesis. left : $left operator: $operator right: $right"
1666 # it's not a leaf, we have a and/or/not
1669 # reintroduce comma content if needed
1670 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1671 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1672 warn "node : $left / $operator / $right\n" if $DEBUG;
1673 my $leftresult = NZanalyse( $left, $server );
1674 my $rightresult = NZanalyse( $right, $server );
1675 warn " leftresult : $leftresult" if $DEBUG;
1676 warn " rightresult : $rightresult" if $DEBUG;
1677 # OK, we have the results for right and left part of the query
1678 # depending of operand, intersect, union or exclude both lists
1679 # to get a result list
1680 if ( $operator eq ' and ' ) {
1682 return NZoperatorAND($leftresult,$rightresult);
1684 elsif ( $operator eq ' or ' ) {
1686 # just merge the 2 strings
1687 return $leftresult . $rightresult;
1689 elsif ( $operator eq ' not ' ) {
1690 return NZoperatorNOT($leftresult,$rightresult);
1694 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1695 die "error : operand unknown : $operator for $string";
1698 # it's a leaf, do the real SQL query and return the result
1701 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1702 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1703 #remove trailing blank at the beginning
1705 warn "leaf:$string" if $DEBUG;
1707 # parse the string in in operator/operand/value again
1711 if ($string =~ /(.*)(>=|<=)(.*)/) {
1718 # warn "handling leaf... left:$left operator:$operator right:$right"
1720 unless ($operator) {
1721 if ($string =~ /(.*)(>|<|=)(.*)/) {
1726 "handling unless (operator)... left:$left operator:$operator right:$right"
1734 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1737 # automatic replace for short operators
1738 $left = 'title' if $left =~ '^ti$';
1739 $left = 'author' if $left =~ '^au$';
1740 $left = 'publisher' if $left =~ '^pb$';
1741 $left = 'subject' if $left =~ '^su$';
1742 $left = 'koha-Auth-Number' if $left =~ '^an$';
1743 $left = 'keyword' if $left =~ '^kw$';
1744 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1745 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1746 my $dbh = C4::Context->dbh;
1747 if ( $operator && $left ne 'keyword' ) {
1748 #do a specific search
1749 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1750 my $sth = $dbh->prepare(
1751 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1753 warn "$left / $operator / $right\n" if $DEBUG;
1755 # split each word, query the DB and build the biblionumbers result
1756 #sanitizing leftpart
1757 $left =~ s/^\s+|\s+$//;
1758 foreach ( split / /, $right ) {
1760 $_ =~ s/^\s+|\s+$//;
1762 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1763 $sth->execute( $server, $left, $_ )
1764 or warn "execute failed: $!";
1765 while ( my ( $line, $value ) = $sth->fetchrow ) {
1767 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1768 # otherwise, fill the result
1769 $biblionumbers .= $line
1770 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
1771 warn "result : $value "
1772 . ( $right =~ /\d/ ) . "=="
1773 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
1776 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1778 warn "NZAND" if $DEBUG;
1779 $results = NZoperatorAND($biblionumbers,$results);
1781 $results = $biblionumbers;
1786 #do a complete search (all indexes), if index='kw' do complete search too.
1787 my $sth = $dbh->prepare(
1788 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
1791 # split each word, query the DB and build the biblionumbers result
1792 foreach ( split / /, $string ) {
1793 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
1794 warn "search on all indexes on $_" if $DEBUG;
1797 $sth->execute( $server, $_ );
1798 while ( my $line = $sth->fetchrow ) {
1799 $biblionumbers .= $line;
1802 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1804 $results = NZoperatorAND($biblionumbers,$results);
1807 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
1808 $results = $biblionumbers;
1812 warn "return : $results for LEAF : $string" if $DEBUG;
1815 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
1819 my ($rightresult, $leftresult)=@_;
1821 my @leftresult = split /;/, $leftresult;
1822 warn " @leftresult / $rightresult \n" if $DEBUG;
1824 # my @rightresult = split /;/,$leftresult;
1827 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1828 # the result is stored twice, to have the same weight for AND than OR.
1829 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1830 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1831 foreach (@leftresult) {
1834 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
1835 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
1836 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
1838 "$value-$countvalue;$value-$countvalue;";
1841 warn "NZAND DONE : $finalresult \n" if $DEBUG;
1842 return $finalresult;
1846 my ($rightresult, $leftresult)=@_;
1847 return $rightresult.$leftresult;
1851 my ($leftresult, $rightresult)=@_;
1853 my @leftresult = split /;/, $leftresult;
1855 # my @rightresult = split /;/,$leftresult;
1857 foreach (@leftresult) {
1859 $value=$1 if $value=~m/(.*)-\d+$/;
1860 unless ($rightresult =~ "$value-") {
1861 $finalresult .= "$_;";
1864 return $finalresult;
1869 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
1876 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
1877 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
1879 # order title asc by default
1880 # $ordering = '1=36 <i' unless $ordering;
1881 $results_per_page = 20 unless $results_per_page;
1882 $offset = 0 unless $offset;
1883 my $dbh = C4::Context->dbh;
1886 # order by POPULARITY
1888 if ( $ordering =~ /popularity/ ) {
1892 # popularity is not in MARC record, it's builded from a specific query
1894 $dbh->prepare("select sum(issues) from items where biblionumber=?");
1895 foreach ( split /;/, $biblionumbers ) {
1896 my ( $biblionumber, $title ) = split /,/, $_;
1897 $result{$biblionumber} = GetMarcBiblio($biblionumber);
1898 $sth->execute($biblionumber);
1899 my $popularity = $sth->fetchrow || 0;
1901 # hint : the key is popularity.title because we can have
1902 # many results with the same popularity. In this case, sub-ordering is done by title
1903 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1904 # (un-frequent, I agree, but we won't forget anything that way ;-)
1905 $popularity{ sprintf( "%10d", $popularity ) . $title
1906 . $biblionumber } = $biblionumber;
1909 # sort the hash and return the same structure as GetRecords (Zebra querying)
1912 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
1913 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
1914 $result_hash->{'RECORDS'}[ $numbers++ ] =
1915 $result{ $popularity{$key} }->as_usmarc();
1918 else { # sort popularity ASC
1919 foreach my $key ( sort ( keys %popularity ) ) {
1920 $result_hash->{'RECORDS'}[ $numbers++ ] =
1921 $result{ $popularity{$key} }->as_usmarc();
1924 my $finalresult = ();
1925 $result_hash->{'hits'} = $numbers;
1926 $finalresult->{'biblioserver'} = $result_hash;
1927 return $finalresult;
1933 elsif ( $ordering =~ /author/ ) {
1935 foreach ( split /;/, $biblionumbers ) {
1936 my ( $biblionumber, $title ) = split /,/, $_;
1937 my $record = GetMarcBiblio($biblionumber);
1939 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
1940 $author = $record->subfield( '200', 'f' );
1941 $author = $record->subfield( '700', 'a' ) unless $author;
1944 $author = $record->subfield( '100', 'a' );
1947 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1948 # and we don't want to get only 1 result for each of them !!!
1949 $result{ $author . $biblionumber } = $record;
1952 # sort the hash and return the same structure as GetRecords (Zebra querying)
1955 if ( $ordering eq 'author_za' ) { # sort by author desc
1956 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
1957 $result_hash->{'RECORDS'}[ $numbers++ ] =
1958 $result{$key}->as_usmarc();
1961 else { # sort by author ASC
1962 foreach my $key ( sort ( keys %result ) ) {
1963 $result_hash->{'RECORDS'}[ $numbers++ ] =
1964 $result{$key}->as_usmarc();
1967 my $finalresult = ();
1968 $result_hash->{'hits'} = $numbers;
1969 $finalresult->{'biblioserver'} = $result_hash;
1970 return $finalresult;
1973 # ORDER BY callnumber
1976 elsif ( $ordering =~ /callnumber/ ) {
1978 foreach ( split /;/, $biblionumbers ) {
1979 my ( $biblionumber, $title ) = split /,/, $_;
1980 my $record = GetMarcBiblio($biblionumber);
1982 my $frameworkcode = GetFrameworkCode($biblionumber);
1983 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
1984 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
1985 unless $callnumber_tag;
1986 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
1987 $callnumber = $record->subfield( '200', 'f' );
1989 $callnumber = $record->subfield( '100', 'a' );
1992 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1993 # and we don't want to get only 1 result for each of them !!!
1994 $result{ $callnumber . $biblionumber } = $record;
1997 # sort the hash and return the same structure as GetRecords (Zebra querying)
2000 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2001 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2002 $result_hash->{'RECORDS'}[ $numbers++ ] =
2003 $result{$key}->as_usmarc();
2006 else { # sort by title ASC
2007 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2008 $result_hash->{'RECORDS'}[ $numbers++ ] =
2009 $result{$key}->as_usmarc();
2012 my $finalresult = ();
2013 $result_hash->{'hits'} = $numbers;
2014 $finalresult->{'biblioserver'} = $result_hash;
2015 return $finalresult;
2017 elsif ( $ordering =~ /pubdate/ ) { #pub year
2019 foreach ( split /;/, $biblionumbers ) {
2020 my ( $biblionumber, $title ) = split /,/, $_;
2021 my $record = GetMarcBiblio($biblionumber);
2022 my ( $publicationyear_tag, $publicationyear_subfield ) =
2023 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2024 my $publicationyear =
2025 $record->subfield( $publicationyear_tag,
2026 $publicationyear_subfield );
2028 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2029 # and we don't want to get only 1 result for each of them !!!
2030 $result{ $publicationyear . $biblionumber } = $record;
2033 # sort the hash and return the same structure as GetRecords (Zebra querying)
2036 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2037 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2038 $result_hash->{'RECORDS'}[ $numbers++ ] =
2039 $result{$key}->as_usmarc();
2042 else { # sort by pub year ASC
2043 foreach my $key ( sort ( keys %result ) ) {
2044 $result_hash->{'RECORDS'}[ $numbers++ ] =
2045 $result{$key}->as_usmarc();
2048 my $finalresult = ();
2049 $result_hash->{'hits'} = $numbers;
2050 $finalresult->{'biblioserver'} = $result_hash;
2051 return $finalresult;
2057 elsif ( $ordering =~ /title/ ) {
2059 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2061 foreach ( split /;/, $biblionumbers ) {
2062 my ( $biblionumber, $title ) = split /,/, $_;
2064 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2065 # and we don't want to get only 1 result for each of them !!!
2066 # hint & speed improvement : we can order without reading the record
2067 # so order, and read records only for the requested page !
2068 $result{ $title . $biblionumber } = $biblionumber;
2071 # sort the hash and return the same structure as GetRecords (Zebra querying)
2074 if ( $ordering eq 'title_az' ) { # sort by title desc
2075 foreach my $key ( sort ( keys %result ) ) {
2076 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2079 else { # sort by title ASC
2080 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2081 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2085 # limit the $results_per_page to result size if it's more
2086 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2088 # for the requested page, replace biblionumber by the complete record
2089 # speed improvement : avoid reading too much things
2091 my $counter = $offset ;
2092 $counter <= $offset + $results_per_page ;
2096 $result_hash->{'RECORDS'}[$counter] =
2097 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2099 my $finalresult = ();
2100 $result_hash->{'hits'} = $numbers;
2101 $finalresult->{'biblioserver'} = $result_hash;
2102 return $finalresult;
2109 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2112 foreach ( split /;/, $biblionumbers ) {
2113 my ( $biblionumber, $title ) = split /,/, $_;
2114 $title =~ /(.*)-(\d)/;
2119 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2120 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2121 # biblio N has ranking = 6
2122 $count_ranking{$biblionumber} += $ranking;
2125 # build the result by "inverting" the count_ranking hash
2126 # 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
2128 foreach ( keys %count_ranking ) {
2129 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2132 # sort the hash and return the same structure as GetRecords (Zebra querying)
2135 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2136 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2139 # limit the $results_per_page to result size if it's more
2140 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2142 # for the requested page, replace biblionumber by the complete record
2143 # speed improvement : avoid reading too much things
2145 my $counter = $offset ;
2146 $counter <= $offset + $results_per_page ;
2150 $result_hash->{'RECORDS'}[$counter] =
2151 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2152 if $result_hash->{'RECORDS'}[$counter];
2154 my $finalresult = ();
2155 $result_hash->{'hits'} = $numbers;
2156 $finalresult->{'biblioserver'} = $result_hash;
2157 return $finalresult;
2161 =head2 enabled_staff_search_views
2163 %hash = enabled_staff_search_views()
2165 This function returns a hash that contains three flags obtained from the system
2166 preferences, used to determine whether a particular staff search results view
2171 =item C<Output arg:>
2173 * $hash{can_view_MARC} is true only if the MARC view is enabled
2174 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2175 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2177 =item C<usage in the script:>
2181 $template->param ( C4::Search::enabled_staff_search_views );
2185 sub enabled_staff_search_views
2188 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2189 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2190 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2194 sub AddSearchHistory{
2195 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2196 my $dbh = C4::Context->dbh;
2198 # Add the request the user just made
2199 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2200 my $sth = $dbh->prepare($sql);
2201 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2202 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2205 sub GetSearchHistory{
2206 my ($borrowernumber,$session)=@_;
2207 my $dbh = C4::Context->dbh;
2209 # Add the request the user just made
2210 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2211 my $sth = $dbh->prepare($query);
2212 $sth->execute($borrowernumber, $session);
2213 return $sth->fetchall_hashref({});
2216 =head2 z3950_search_args
2218 $arrayref = z3950_search_args($matchpoints)
2220 This function returns an array reference that contains the search parameters to be
2221 passed to the Z39.50 search script (z3950_search.pl). The array elements
2222 are hash refs whose keys are name, value and encvalue, and whose values are the
2223 name of a search parameter, the value of that search parameter and the URL encoded
2224 value of that parameter.
2226 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2228 The search parameter values are obtained from the bibliographic record whose
2229 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2231 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2232 a general purpose search argument. In this case, the returned array contains only
2233 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2235 If a search parameter value is undefined or empty, it is not included in the returned
2238 The returned array reference may be passed directly to the template parameters.
2242 =item C<Output arg:>
2244 * $array containing hash refs as described above
2246 =item C<usage in the script:>
2250 $data = Biblio::GetBiblioData($bibno);
2251 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2255 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2259 sub z3950_search_args {
2261 $bibrec = { title => $bibrec } if !ref $bibrec;
2263 for my $field (qw/ lccn isbn issn title author dewey subject /)
2265 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2266 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2271 =head2 BiblioAddAuthorities
2273 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2275 this function finds the authorities linked to the biblio
2276 * search in the authority DB for the same authid (in $9 of the biblio)
2277 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2278 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2279 OR adds a new authority record
2285 * $record is the MARC record in question (marc blob)
2286 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2288 =item C<Output arg:>
2290 * $countlinked is the number of authorities records that are linked to this authority
2294 * 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)
2300 sub BiblioAddAuthorities{
2301 my ( $record, $frameworkcode ) = @_;
2302 my $dbh=C4::Context->dbh;
2303 my $query=$dbh->prepare(qq|
2304 SELECT authtypecode,tagfield
2305 FROM marc_subfield_structure
2306 WHERE frameworkcode=?
2307 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2308 # SELECT authtypecode,tagfield
2309 # FROM marc_subfield_structure
2310 # WHERE frameworkcode=?
2311 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2312 $query->execute($frameworkcode);
2313 my ($countcreated,$countlinked);
2314 while (my $data=$query->fetchrow_hashref){
2315 foreach my $field ($record->field($data->{tagfield})){
2316 next if ($field->subfield('3')||$field->subfield('9'));
2317 # No authorities id in the tag.
2318 # Search if there is any authorities to link to.
2319 my $query='at='.$data->{authtypecode}.' ';
2320 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2321 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2322 # there is only 1 result
2324 warn "BIBLIOADDSAUTHORITIES: $error";
2327 if ($results && scalar(@$results)==1) {
2328 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2329 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2331 } elsif (scalar(@$results)>1) {
2332 #More than One result
2333 #This can comes out of a lack of a subfield.
2334 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2335 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2338 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2339 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2340 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2341 my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2342 next unless $authtypedata;
2343 my $marcrecordauth=MARC::Record->new();
2344 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2345 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2346 $marcrecordauth->insert_fields_ordered($authfield);
2348 # bug 2317: ensure new authority knows it's using UTF-8; currently
2349 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2350 # automatically for UNIMARC (by not transcoding)
2351 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2352 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2353 # of change to a core API just before the 3.0 release.
2354 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2355 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2358 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2360 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2362 $field->add_subfields('9'=>$authid);
2366 return ($countlinked,$countcreated);
2369 =head2 GetDistinctValues($field);
2371 C<$field> is a reference to the fields array
2375 sub GetDistinctValues {
2376 my ($fieldname,$string)=@_;
2377 # returns a reference to a hash of references to branches...
2378 if ($fieldname=~/\./){
2379 my ($table,$column)=split /\./, $fieldname;
2380 my $dbh = C4::Context->dbh;
2381 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column ";
2382 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 ");
2384 my $elements=$sth->fetchall_arrayref({});
2389 my @servers=qw<biblioserver authorityserver>;
2390 my (@zconns,@results);
2391 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2392 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2395 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2398 # The big moment: asynchronously retrieve results from all servers
2400 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2401 my $ev = $zconns[ $i - 1 ]->last_event();
2402 if ( $ev == ZOOM::Event::ZEND ) {
2403 next unless $results[ $i - 1 ];
2404 my $size = $results[ $i - 1 ]->size();
2406 for (my $j=0;$j<$size;$j++){
2408 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2409 push @elements, \%hashscan;
2419 END { } # module clean-up code here (global destructor)
2426 Koha Developement team <info@koha.org>