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();
451 warn $results[$i-1]->record($j)->render() ;
453 # warn "RECORD $j:".$record;
454 $results_hash->{'RECORDS'}[$j] = $record;
456 # Fill the facets while we're looping, but only for the biblioserver
457 $facet_record = MARC::Record->new_from_usmarc($record)
458 if $servers[ $i - 1 ] =~ /biblioserver/;
460 #warn $servers[$i-1]."\n".$record; #.$facet_record->title();
462 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
463 ($facets->[$k]) or next;
464 my @fields = map {$facet_record->field($_)} @{$facets->[$k]->{'tags'}} ;
465 for my $field (@fields) {
466 my @subfields = $field->subfields();
467 for my $subfield (@subfields) {
468 my ( $code, $data ) = @$subfield;
469 ($code eq $facets->[$k]->{'subfield'}) or next;
470 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
473 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} =
474 $facets->[$k]->{'label_value'};
475 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} =
476 $facets->[$k]->{'expanded'};
481 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
484 # warn "connection ", $i-1, ": $size hits";
485 # warn $results[$i-1]->record(0)->render() if $size > 0;
488 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
490 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
491 keys %$facets_counter )
494 my $number_of_facets;
495 my @this_facets_array;
498 $facets_counter->{$link_value}->{$b}
499 <=> $facets_counter->{$link_value}->{$a}
500 } keys %{ $facets_counter->{$link_value} }
504 if ( ( $number_of_facets < 6 )
505 || ( $expanded_facet eq $link_value )
506 || ( $facets_info->{$link_value}->{'expanded'} ) )
509 # Sanitize the link value ), ( will cause errors with CCL,
510 my $facet_link_value = $one_facet;
511 $facet_link_value =~ s/(\(|\))/ /g;
513 # fix the length that will display in the label,
514 my $facet_label_value = $one_facet;
516 substr( $one_facet, 0, 20 ) . "..."
517 unless length($facet_label_value) <= 20;
519 # if it's a branch, label by the name, not the code,
520 if ( $link_value =~ /branch/ ) {
521 if (defined $branches
522 && ref($branches) eq "HASH"
523 && defined $branches->{$one_facet}
524 && ref ($branches->{$one_facet}) eq "HASH")
527 $branches->{$one_facet}->{'branchname'};
530 $facet_label_value = "*";
534 # but we're down with the whole label being in the link's title.
535 push @this_facets_array, {
536 facet_count => $facets_counter->{$link_value}->{$one_facet},
537 facet_label_value => $facet_label_value,
538 facet_title_value => $one_facet,
539 facet_link_value => $facet_link_value,
540 type_link_value => $link_value,
545 # handle expanded option
546 unless ( $facets_info->{$link_value}->{'expanded'} ) {
548 if ( ( $number_of_facets > 6 )
549 && ( $expanded_facet ne $link_value ) );
552 type_link_value => $link_value,
553 type_id => $link_value . "_id",
554 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
555 facets => \@this_facets_array,
556 expandable => $expandable,
557 expand => $link_value,
558 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
563 return ( undef, $results_hashref, \@facets_loop );
568 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
569 $results_per_page, $offset, $expanded_facet, $branches,
573 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
575 $paz->search($simple_query);
576 sleep 1; # FIXME: WHY?
579 my $results_hashref = {};
580 my $stats = XMLin($paz->stat);
581 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
583 # for a grouped search result, the number of hits
584 # is the number of groups returned; 'bib_hits' will have
585 # the total number of bibs.
586 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
587 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
589 HIT: foreach my $hit (@{ $results->{'hit'} }) {
590 my $recid = $hit->{recid}->[0];
592 my $work_title = $hit->{'md-work-title'}->[0];
594 if (exists $hit->{'md-work-author'}) {
595 $work_author = $hit->{'md-work-author'}->[0];
597 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
599 my $result_group = {};
600 $result_group->{'group_label'} = $group_label;
601 $result_group->{'group_merge_key'} = $recid;
604 if (exists $hit->{count}) {
605 $count = $hit->{count}->[0];
607 $result_group->{'group_count'} = $count;
609 for (my $i = 0; $i < $count; $i++) {
610 # FIXME -- may need to worry about diacritics here
611 my $rec = $paz->record($recid, $i);
612 push @{ $result_group->{'RECORDS'} }, $rec;
615 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
618 # pass through facets
619 my $termlist_xml = $paz->termlist('author,subject');
620 my $terms = XMLin($termlist_xml, forcearray => 1);
621 my @facets_loop = ();
622 #die Dumper($results);
623 # foreach my $list (sort keys %{ $terms->{'list'} }) {
625 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
627 # facet_label_value => $facet->{'name'}->[0],
630 # push @facets_loop, ( {
631 # type_label => $list,
632 # facets => \@facets,
636 return ( undef, $results_hashref, \@facets_loop );
640 sub _remove_stopwords {
641 my ( $operand, $index ) = @_;
642 my @stopwords_removed;
644 # phrase and exact-qualified indexes shouldn't have stopwords removed
645 if ( $index !~ m/phr|ext/ ) {
647 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
648 # we use IsAlpha unicode definition, to deal correctly with diacritics.
649 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
650 # is a stopword, we'd get "çon" and wouldn't find anything...
652 foreach ( keys %{ C4::Context->stopwords } ) {
653 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
654 $debug && warn "$_ Dump($operand)";
655 if ( my ($matched) = ($operand =~
656 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
658 $operand =~ s/\Q$matched\E/ /gi;
659 push @stopwords_removed, $_;
663 return ( $operand, \@stopwords_removed );
667 sub _detect_truncation {
668 my ( $operand, $index ) = @_;
669 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
672 my @wordlist = split( /\s/, $operand );
673 foreach my $word (@wordlist) {
674 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
675 push @rightlefttruncated, $word;
677 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
678 push @lefttruncated, $word;
680 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
681 push @righttruncated, $word;
683 elsif ( index( $word, "*" ) < 0 ) {
684 push @nontruncated, $word;
687 push @regexpr, $word;
691 \@nontruncated, \@righttruncated, \@lefttruncated,
692 \@rightlefttruncated, \@regexpr
697 sub _build_stemmed_operand {
698 my ($operand,$lang) = @_;
699 require Lingua::Stem::Snowball ;
702 # If operand contains a digit, it is almost certainly an identifier, and should
703 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
704 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
705 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
706 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
707 return $operand if $operand =~ /\d/;
709 # FIXME: the locale should be set based on the user's language and/or search choice
711 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
712 encoding => "UTF-8" );
714 my @words = split( / /, $operand );
715 my @stems = $stemmer->stem(\@words);
716 for my $stem (@stems) {
717 $stemmed_operand .= "$stem";
718 $stemmed_operand .= "?"
719 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
720 $stemmed_operand .= " ";
722 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
723 return $stemmed_operand;
727 sub _build_weighted_query {
729 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
730 # pretty well but could work much better if we had a smarter query parser
731 my ( $operand, $stemmed_operand, $index ) = @_;
732 my $stemming = C4::Context->preference("QueryStemming") || 0;
733 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
734 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
736 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
738 # Keyword, or, no index specified
739 if ( ( $index eq 'kw' ) || ( !$index ) ) {
741 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
742 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
743 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
744 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
745 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
746 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
747 if $fuzzy_enabled; # add fuzzy, word list
748 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
749 if ( $stemming and $stemmed_operand )
750 ; # add stemming, right truncation
751 $weighted_query .= " or wrdl,r9=\"$operand\"";
753 # embedded sorting: 0 a-z; 1 z-a
754 # $weighted_query .= ") or (sort1,aut=1";
757 # Barcode searches should skip this process
758 elsif ( $index eq 'bc' ) {
759 $weighted_query .= "bc=\"$operand\"";
762 # Authority-number searches should skip this process
763 elsif ( $index eq 'an' ) {
764 $weighted_query .= "an=\"$operand\"";
767 # If the index already has more than one qualifier, wrap the operand
768 # in quotes and pass it back (assumption is that the user knows what they
769 # are doing and won't appreciate us mucking up their query
770 elsif ( $index =~ ',' ) {
771 $weighted_query .= " $index=\"$operand\"";
774 #TODO: build better cases based on specific search indexes
776 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
777 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
778 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
780 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
783 $weighted_query .= "))"; # close rank specification
784 return $weighted_query;
789 Return an array with available indexes.
811 'Author-personal-bibliography',
820 'Chronological-subdivision',
830 'Conference-name-heading',
831 'Conference-name-see',
832 'Conference-name-seealso',
837 'Corporate-name-heading',
838 'Corporate-name-see',
839 'Corporate-name-seealso',
841 'date-entered-on-file',
842 'Date-of-acquisition',
843 'Date-of-publication',
844 'Dewey-classification',
850 'Geographic-subdivision',
853 'Heading-use-main-or-added-entry',
854 'Heading-use-series-added-entry ',
855 'Heading-use-subject-added-entry',
871 'Local-classification',
874 'Match-heading-see-from',
880 'Name-geographic-heading',
881 'Name-geographic-see',
882 'Name-geographic-seealso',
889 'Personal-name-heading',
891 'Personal-name-seealso',
907 'Subject-heading-thesaurus',
908 'Subject-name-personal',
909 'Subject-subdivision',
918 'Term-genre-form-heading',
919 'Term-genre-form-see',
920 'Term-genre-form-seealso',
926 'Title-uniform-heading',
928 'Title-uniform-seealso',
938 'classification-source',
940 'coded-location-qualifier',
950 'Local-classification',
953 'materials-specified',
962 'replacementpricedate',
978 $simple_query, $query_cgi,
980 $limit_cgi, $limit_desc,
981 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
983 Build queries and limits in CCL, CGI, Human,
984 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
986 See verbose embedded documentation.
992 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
994 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
997 my @operators = $operators ? @$operators : ();
998 my @indexes = $indexes ? @$indexes : ();
999 my @operands = $operands ? @$operands : ();
1000 my @limits = $limits ? @$limits : ();
1001 my @sort_by = $sort_by ? @$sort_by : ();
1003 my $stemming = C4::Context->preference("QueryStemming") || 0;
1004 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1005 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1006 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1007 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1009 # no stemming/weight/fuzzy in NoZebra
1010 if ( C4::Context->preference("NoZebra") ) {
1014 $auto_truncation = 0;
1017 my $query = $operands[0];
1018 my $simple_query = $operands[0];
1020 # initialize the variables we're passing back
1029 my $stopwords_removed; # flag to determine if stopwords have been removed
1032 my $cclindexes = getIndexes();
1033 if( $query !~ /\s*ccl=/ ){
1034 for my $index (@$cclindexes){
1035 if($query =~ /($index)(,?\w)*[:=]/){
1039 $query = "ccl=$query" if($cclq);
1042 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1044 if ( $query =~ /^ccl=/ ) {
1045 return ( undef, $', $', "q=ccl=$'", $', '', '', '', '', 'ccl' );
1047 if ( $query =~ /^cql=/ ) {
1048 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1050 if ( $query =~ /^pqf=/ ) {
1051 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1054 # pass nested queries directly
1055 # FIXME: need better handling of some of these variables in this case
1056 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1057 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1058 # if ( $query =~ /(\(|\))/ ) {
1060 # undef, $query, $simple_query, $query_cgi,
1061 # $query, $limit, $limit_cgi, $limit_desc,
1062 # $stopwords_removed, 'ccl'
1066 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1067 # query operands and indexes and add stemming, truncation, field weighting, etc.
1068 # Once we do so, we'll end up with a value in $query, just like if we had an
1069 # incoming $query from the user
1072 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1073 my $previous_operand
1074 ; # a flag used to keep track if there was a previous query
1075 # if there was, we can apply the current operator
1077 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1079 # COMBINE OPERANDS, INDEXES AND OPERATORS
1080 if ( $operands[$i] ) {
1082 # A flag to determine whether or not to add the index to the query
1085 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1086 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1089 $remove_stopwords = 0;
1091 my $operand = $operands[$i];
1092 my $index = $indexes[$i];
1094 # Add index-specific attributes
1095 # Date of Publication
1096 if ( $index eq 'yr' ) {
1097 $index .= ",st-numeric";
1099 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1102 # Date of Acquisition
1103 elsif ( $index eq 'acqdate' ) {
1104 $index .= ",st-date-normalized";
1106 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1108 # ISBN,ISSN,Standard Number, don't need special treatment
1109 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1112 $stemming, $auto_truncation,
1113 $weight_fields, $fuzzy_enabled,
1115 ) = ( 0, 0, 0, 0, 0 );
1123 # Set default structure attribute (word list)
1125 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1126 $struct_attr = ",wrdl";
1129 # Some helpful index variants
1130 my $index_plus = $index . $struct_attr . ":" if $index;
1131 my $index_plus_comma = $index . $struct_attr . "," if $index;
1134 if ($remove_stopwords) {
1135 ( $operand, $stopwords_removed ) =
1136 _remove_stopwords( $operand, $index );
1137 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1138 warn "REMOVED STOPWORDS: @$stopwords_removed"
1139 if ( $stopwords_removed && $DEBUG );
1142 if ($auto_truncation){
1143 unless ( $index =~ /(st-|phr|ext)/ ) {
1144 #FIXME only valid with LTR scripts
1145 $operand=join(" ",map{
1146 (index($_,"*")>0?"$_":"$_*")
1147 }split (/\s+/,$operand));
1148 warn $operand if $DEBUG;
1153 my $truncated_operand;
1154 my( $nontruncated, $righttruncated, $lefttruncated,
1155 $rightlefttruncated, $regexpr
1156 ) = _detect_truncation( $operand, $index );
1158 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1163 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1164 scalar(@$rightlefttruncated) > 0 )
1167 # Don't field weight or add the index to the query, we do it here
1169 undef $weight_fields;
1170 my $previous_truncation_operand;
1171 if (scalar @$nontruncated) {
1172 $truncated_operand .= "$index_plus @$nontruncated ";
1173 $previous_truncation_operand = 1;
1175 if (scalar @$righttruncated) {
1176 $truncated_operand .= "and " if $previous_truncation_operand;
1177 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1178 $previous_truncation_operand = 1;
1180 if (scalar @$lefttruncated) {
1181 $truncated_operand .= "and " if $previous_truncation_operand;
1182 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1183 $previous_truncation_operand = 1;
1185 if (scalar @$rightlefttruncated) {
1186 $truncated_operand .= "and " if $previous_truncation_operand;
1187 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1188 $previous_truncation_operand = 1;
1191 $operand = $truncated_operand if $truncated_operand;
1192 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1195 my $stemmed_operand;
1196 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1199 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1201 # Handle Field Weighting
1202 my $weighted_operand;
1203 if ($weight_fields) {
1204 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1205 $operand = $weighted_operand;
1209 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1211 # If there's a previous operand, we need to add an operator
1212 if ($previous_operand) {
1214 # User-specified operator
1215 if ( $operators[ $i - 1 ] ) {
1216 $query .= " $operators[$i-1] ";
1217 $query .= " $index_plus " unless $indexes_set;
1218 $query .= " $operand";
1219 $query_cgi .= "&op=$operators[$i-1]";
1220 $query_cgi .= "&idx=$index" if $index;
1221 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1223 " $operators[$i-1] $index_plus $operands[$i]";
1226 # Default operator is and
1229 $query .= "$index_plus " unless $indexes_set;
1230 $query .= "$operand";
1231 $query_cgi .= "&op=and&idx=$index" if $index;
1232 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1233 $query_desc .= " and $index_plus $operands[$i]";
1237 # There isn't a pervious operand, don't need an operator
1240 # Field-weighted queries already have indexes set
1241 $query .= " $index_plus " unless $indexes_set;
1243 $query_desc .= " $index_plus $operands[$i]";
1244 $query_cgi .= "&idx=$index" if $index;
1245 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1246 $previous_operand = 1;
1251 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1254 my $group_OR_limits;
1255 my $availability_limit;
1256 foreach my $this_limit (@limits) {
1257 # if ( $this_limit =~ /available/ ) {
1259 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1261 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1262 # $availability_limit .=
1263 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1264 # $limit_cgi .= "&limit=available";
1265 # $limit_desc .= "";
1268 # group_OR_limits, prefixed by mc-
1269 # OR every member of the group
1270 # elsif ( $this_limit =~ /mc/ ) {
1271 if ( $this_limit =~ /mc/ ) {
1272 $group_OR_limits .= " or " if $group_OR_limits;
1273 $limit_desc .= " or " if $group_OR_limits;
1274 $group_OR_limits .= "$this_limit";
1275 $limit_cgi .= "&limit=$this_limit";
1276 $limit_desc .= " $this_limit";
1279 # Regular old limits
1281 $limit .= " and " if $limit || $query;
1282 $limit .= "$this_limit";
1283 $limit_cgi .= "&limit=$this_limit";
1284 if ($this_limit =~ /^branch:(.+)/) {
1285 my $branchcode = $1;
1286 my $branchname = GetBranchName($branchcode);
1287 if (defined $branchname) {
1288 $limit_desc .= " branch:$branchname";
1290 $limit_desc .= " $this_limit";
1293 $limit_desc .= " $this_limit";
1297 if ($group_OR_limits) {
1298 $limit .= " and " if ( $query || $limit );
1299 $limit .= "($group_OR_limits)";
1301 if ($availability_limit) {
1302 $limit .= " and " if ( $query || $limit );
1303 $limit .= "($availability_limit)";
1306 # Normalize the query and limit strings
1307 # This is flawed , means we can't search anything with : in it
1308 # if user wants to do ccl or cql, start the query with that
1311 for ( $query, $query_desc, $limit, $limit_desc ) {
1312 s/ / /g; # remove extra spaces
1313 s/^ //g; # remove any beginning spaces
1314 s/ $//g; # remove any ending spaces
1315 s/==/=/g; # remove double == from query
1317 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1319 for ($query_cgi,$simple_query) {
1322 # append the limit to the query
1323 $query .= " " . $limit;
1327 warn "QUERY:" . $query;
1328 warn "QUERY CGI:" . $query_cgi;
1329 warn "QUERY DESC:" . $query_desc;
1330 warn "LIMIT:" . $limit;
1331 warn "LIMIT CGI:" . $limit_cgi;
1332 warn "LIMIT DESC:" . $limit_desc;
1333 warn "---------\nLeave buildQuery\n---------";
1336 undef, $query, $simple_query, $query_cgi,
1337 $query_desc, $limit, $limit_cgi, $limit_desc,
1338 $stopwords_removed, $query_type
1342 =head2 searchResults
1344 Format results in a form suitable for passing to the template
1348 # IMO this subroutine is pretty messy still -- it's responsible for
1349 # building the HTML output for the template
1351 my ( $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1352 my $dbh = C4::Context->dbh;
1355 #Build branchnames hash
1357 #get branch information.....
1359 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1361 while ( my $bdata = $bsth->fetchrow_hashref ) {
1362 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1364 # FIXME - We build an authorised values hash here, using the default framework
1365 # though it is possible to have different authvals for different fws.
1367 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1369 # get notforloan authorised value list (see $shelflocations FIXME)
1370 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1372 #Build itemtype hash
1373 #find itemtype & itemtype image
1377 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1380 while ( my $bdata = $bsth->fetchrow_hashref ) {
1381 foreach (qw(description imageurl summary notforloan)) {
1382 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1386 #search item field code
1389 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1392 my ($itemtag) = $sth->fetchrow;
1394 ## find column names of items related to MARC
1395 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1397 my %subfieldstosearch;
1398 while ( ( my $column ) = $sth2->fetchrow ) {
1399 my ( $tagfield, $tagsubfield ) =
1400 &GetMarcFromKohaField( "items." . $column, "" );
1401 $subfieldstosearch{$column} = $tagsubfield;
1404 # handle which records to actually retrieve
1406 if ( $hits && $offset + $results_per_page <= $hits ) {
1407 $times = $offset + $results_per_page;
1410 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1413 my $marcflavour = C4::Context->preference("marcflavour");
1414 # We get the biblionumber position in MARC
1415 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1418 # loop through all of the records we've retrieved
1419 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1420 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1422 $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1424 $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1427 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1428 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1429 $oldbiblio->{result_number} = $i + 1;
1431 # add imageurl to itemtype if there is one
1432 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1434 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1435 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1436 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1437 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1438 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1439 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1441 # edition information, if any
1442 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1443 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1444 # Build summary if there is one (the summary is defined in the itemtypes table)
1445 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1446 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1447 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1448 my @fields = $marcrecord->fields();
1451 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1453 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1454 $tag =~ /(.{3})(.)/;
1455 if($marcrecord->field($1)){
1456 my @abc = $marcrecord->field($1)->subfield($2);
1457 $tags->{$tag} = $#abc + 1 ;
1461 # We catch how many times to repeat this line
1463 foreach my $tag (keys(%$tags)){
1464 $max = $tags->{$tag} if($tags->{$tag} > $max);
1467 # we replace, and repeat each line
1468 for (my $i = 0 ; $i < $max ; $i++){
1469 my $newline = $line;
1471 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1472 $tag =~ /(.{3})(.)/;
1474 if($marcrecord->field($1)){
1475 my @repl = $marcrecord->field($1)->subfield($2);
1476 my $subfieldvalue = $repl[$i];
1478 if (! utf8::is_utf8($subfieldvalue)) {
1479 utf8::decode($subfieldvalue);
1482 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1485 $newsummary .= "$newline\n";
1489 $newsummary =~ s/\[(.*?)]//g;
1490 $newsummary =~ s/\n/<br\/>/g;
1491 $oldbiblio->{summary} = $newsummary;
1494 # Pull out the items fields
1495 my @fields = $marcrecord->field($itemtag);
1497 # Setting item statuses for display
1498 my @available_items_loop;
1499 my @onloan_items_loop;
1500 my @other_items_loop;
1502 my $available_items;
1506 my $ordered_count = 0;
1507 my $available_count = 0;
1508 my $onloan_count = 0;
1509 my $longoverdue_count = 0;
1510 my $other_count = 0;
1511 my $wthdrawn_count = 0;
1512 my $itemlost_count = 0;
1513 my $itembinding_count = 0;
1514 my $itemdamaged_count = 0;
1515 my $item_in_transit_count = 0;
1516 my $can_place_holds = 0;
1517 my $items_count = scalar(@fields);
1519 ( C4::Context->preference('maxItemsinSearchResults') )
1520 ? C4::Context->preference('maxItemsinSearchResults') - 1
1523 # loop through every item
1524 foreach my $field (@fields) {
1527 # populate the items hash
1528 foreach my $code ( keys %subfieldstosearch ) {
1529 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1532 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1533 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1534 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1535 if ($item->{$hbranch}) {
1536 $item->{'branchname'} = $branches{$item->{$hbranch}};
1538 elsif ($item->{$otherbranch}) { # Last resort
1539 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1542 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1543 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1544 if ( $item->{onloan} ) {
1546 my $key = $prefix . $item->{onloan} . $item->{barcode};
1547 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1548 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1549 $onloan_items->{$key}->{branchname} = $item->{branchname};
1550 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1551 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1552 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1553 # if something's checked out and lost, mark it as 'long overdue'
1554 if ( $item->{itemlost} ) {
1555 $onloan_items->{$prefix}->{longoverdue}++;
1556 $longoverdue_count++;
1557 } else { # can place holds as long as item isn't lost
1558 $can_place_holds = 1;
1562 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1566 if ( $item->{notforloan} == -1 ) {
1570 # is item in transit?
1571 my $transfertwhen = '';
1572 my ($transfertfrom, $transfertto);
1574 unless ($item->{wthdrawn}
1575 || $item->{itemlost}
1577 || $item->{notforloan}
1578 || $items_count > 20) {
1580 # A couple heuristics to limit how many times
1581 # we query the database for item transfer information, sacrificing
1582 # accuracy in some cases for speed;
1584 # 1. don't query if item has one of the other statuses
1585 # 2. don't check transit status if the bib has
1586 # more than 20 items
1588 # FIXME: to avoid having the query the database like this, and to make
1589 # the in transit status count as unavailable for search limiting,
1590 # should map transit status to record indexed in Zebra.
1592 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1595 # item is withdrawn, lost or damaged
1596 if ( $item->{wthdrawn}
1597 || $item->{itemlost}
1599 || $item->{notforloan}
1600 || ($transfertwhen ne ''))
1602 $wthdrawn_count++ if $item->{wthdrawn};
1603 $itemlost_count++ if $item->{itemlost};
1604 $itemdamaged_count++ if $item->{damaged};
1605 $item_in_transit_count++ if $transfertwhen ne '';
1606 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1609 my $key = $prefix . $item->{status};
1610 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1611 $other_items->{$key}->{$_} = $item->{$_};
1613 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1614 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1615 $other_items->{$key}->{count}++ if $item->{$hbranch};
1616 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1617 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1621 $can_place_holds = 1;
1623 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1624 foreach (qw(branchname itemcallnumber)) {
1625 $available_items->{$prefix}->{$_} = $item->{$_};
1627 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1628 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1631 } # notforloan, item level and biblioitem level
1632 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1634 ( C4::Context->preference('maxItemsinSearchResults') )
1635 ? C4::Context->preference('maxItemsinSearchResults') - 1
1637 for my $key ( sort keys %$onloan_items ) {
1638 (++$onloanitemscount > $maxitems) and last;
1639 push @onloan_items_loop, $onloan_items->{$key};
1641 for my $key ( sort keys %$other_items ) {
1642 (++$otheritemscount > $maxitems) and last;
1643 push @other_items_loop, $other_items->{$key};
1645 for my $key ( sort keys %$available_items ) {
1646 (++$availableitemscount > $maxitems) and last;
1647 push @available_items_loop, $available_items->{$key}
1650 # XSLT processing of some stuff
1653 SetUTF8Flag($marcrecord);
1654 $debug && warn $marcrecord->as_formatted;
1655 if (C4::Context->preference("XSLTResultsDisplay") && !$scan) {
1656 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1657 $oldbiblio->{biblionumber}, $marcrecord, 'Results' );
1660 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1661 $can_place_holds = 0
1662 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1663 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1664 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1665 $oldbiblio->{items_count} = $items_count;
1666 $oldbiblio->{available_items_loop} = \@available_items_loop;
1667 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1668 $oldbiblio->{other_items_loop} = \@other_items_loop;
1669 $oldbiblio->{availablecount} = $available_count;
1670 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1671 $oldbiblio->{onloancount} = $onloan_count;
1672 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1673 $oldbiblio->{othercount} = $other_count;
1674 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1675 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1676 $oldbiblio->{itemlostcount} = $itemlost_count;
1677 $oldbiblio->{damagedcount} = $itemdamaged_count;
1678 $oldbiblio->{intransitcount} = $item_in_transit_count;
1679 $oldbiblio->{orderedcount} = $ordered_count;
1680 $oldbiblio->{isbn} =~
1681 s/-//g; # deleting - in isbn to enable amazon content
1682 push( @newresults, $oldbiblio )
1683 if(not $hidelostitems
1684 or (($items_count > $itemlost_count )
1685 && $hidelostitems));
1691 =head2 SearchAcquisitions
1692 Search for acquisitions
1695 sub SearchAcquisitions{
1696 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1698 my $dbh=C4::Context->dbh;
1699 # Variable initialization
1703 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1704 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1705 WHERE dateaccessioned BETWEEN ? AND ?
1708 my (@params,@loopcriteria);
1710 push @params, $datebegin->output("iso");
1711 push @params, $dateend->output("iso");
1713 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1714 if(C4::Context->preference("item-level_itypes")){
1715 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1717 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1719 push @params, @$itemtypes;
1722 if ($criteria =~/itemtype/){
1723 if(C4::Context->preference("item-level_itypes")){
1724 $str .= "AND items.itype=? ";
1726 $str .= "AND biblioitems.itemtype=? ";
1729 if(scalar(@$itemtypes) == 0){
1730 my $itypes = GetItemTypes();
1731 for my $key (keys %$itypes){
1732 push @$itemtypes, $key;
1736 @loopcriteria= @$itemtypes;
1737 }elsif ($criteria=~/itemcallnumber/){
1738 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1739 OR items.itemcallnumber is NULL
1740 OR items.itemcallnumber = '')";
1742 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1744 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1745 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1748 if ($orderby =~ /date_desc/){
1749 $str.=" ORDER BY dateaccessioned DESC";
1751 $str.=" ORDER BY title";
1754 my $qdataacquisitions=$dbh->prepare($str);
1756 my @loopacquisitions;
1757 foreach my $value(@loopcriteria){
1758 push @params,$value;
1760 $cell{"title"}=$value;
1761 $cell{"titlecode"}=$value;
1763 eval{$qdataacquisitions->execute(@params);};
1765 if ($@){ warn "recentacquisitions Error :$@";}
1768 while (my $data=$qdataacquisitions->fetchrow_hashref){
1769 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1771 $cell{"loopdata"}=\@loopdata;
1773 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1776 $qdataacquisitions->finish;
1777 return \@loopacquisitions;
1779 #----------------------------------------------------------------------
1781 # Non-Zebra GetRecords#
1782 #----------------------------------------------------------------------
1786 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1792 $query, $simple_query, $sort_by_ref, $servers_ref,
1793 $results_per_page, $offset, $expanded_facet, $branches,
1796 warn "query =$query" if $DEBUG;
1797 my $result = NZanalyse($query);
1798 warn "results =$result" if $DEBUG;
1800 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1806 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1807 the list is built from an inverted index in the nozebra SQL table
1808 note that title is here only for convenience : the sorting will be very fast when requested on title
1809 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1814 my ( $string, $server ) = @_;
1815 # warn "---------" if $DEBUG;
1816 warn " NZanalyse" if $DEBUG;
1817 # warn "---------" if $DEBUG;
1819 # $server contains biblioserver or authorities, depending on what we search on.
1820 #warn "querying : $string on $server";
1821 $server = 'biblioserver' unless $server;
1823 # if we have a ", replace the content to discard temporarily any and/or/not inside
1825 if ( $string =~ /"/ ) {
1826 $string =~ s/"(.*?)"/__X__/;
1828 warn "commacontent : $commacontent" if $DEBUG;
1831 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1832 # then, call again NZanalyse with $left and $right
1833 # (recursive until we find a leaf (=> something without and/or/not)
1834 # delete repeated operator... Would then go in infinite loop
1835 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1838 #process parenthesis before.
1839 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1842 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1844 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1846 my $leftresult = NZanalyse( $left, $server );
1848 my $rightresult = NZanalyse( $right, $server );
1850 # OK, we have the results for right and left part of the query
1851 # depending of operand, intersect, union or exclude both lists
1852 # to get a result list
1853 if ( $operator eq ' and ' ) {
1854 return NZoperatorAND($leftresult,$rightresult);
1856 elsif ( $operator eq ' or ' ) {
1858 # just merge the 2 strings
1859 return $leftresult . $rightresult;
1861 elsif ( $operator eq ' not ' ) {
1862 return NZoperatorNOT($leftresult,$rightresult);
1866 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1870 warn "string :" . $string if $DEBUG;
1874 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1877 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1879 warn "no parenthesis. left : $left operator: $operator right: $right"
1882 # it's not a leaf, we have a and/or/not
1885 # reintroduce comma content if needed
1886 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1887 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1888 warn "node : $left / $operator / $right\n" if $DEBUG;
1889 my $leftresult = NZanalyse( $left, $server );
1890 my $rightresult = NZanalyse( $right, $server );
1891 warn " leftresult : $leftresult" if $DEBUG;
1892 warn " rightresult : $rightresult" if $DEBUG;
1893 # OK, we have the results for right and left part of the query
1894 # depending of operand, intersect, union or exclude both lists
1895 # to get a result list
1896 if ( $operator eq ' and ' ) {
1898 return NZoperatorAND($leftresult,$rightresult);
1900 elsif ( $operator eq ' or ' ) {
1902 # just merge the 2 strings
1903 return $leftresult . $rightresult;
1905 elsif ( $operator eq ' not ' ) {
1906 return NZoperatorNOT($leftresult,$rightresult);
1910 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1911 die "error : operand unknown : $operator for $string";
1914 # it's a leaf, do the real SQL query and return the result
1917 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1918 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1919 #remove trailing blank at the beginning
1921 warn "leaf:$string" if $DEBUG;
1923 # parse the string in in operator/operand/value again
1927 if ($string =~ /(.*)(>=|<=)(.*)/) {
1934 # warn "handling leaf... left:$left operator:$operator right:$right"
1936 unless ($operator) {
1937 if ($string =~ /(.*)(>|<|=)(.*)/) {
1942 "handling unless (operator)... left:$left operator:$operator right:$right"
1950 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1953 # automatic replace for short operators
1954 $left = 'title' if $left =~ '^ti$';
1955 $left = 'author' if $left =~ '^au$';
1956 $left = 'publisher' if $left =~ '^pb$';
1957 $left = 'subject' if $left =~ '^su$';
1958 $left = 'koha-Auth-Number' if $left =~ '^an$';
1959 $left = 'keyword' if $left =~ '^kw$';
1960 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1961 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1962 my $dbh = C4::Context->dbh;
1963 if ( $operator && $left ne 'keyword' ) {
1964 #do a specific search
1965 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1966 my $sth = $dbh->prepare(
1967 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1969 warn "$left / $operator / $right\n" if $DEBUG;
1971 # split each word, query the DB and build the biblionumbers result
1972 #sanitizing leftpart
1973 $left =~ s/^\s+|\s+$//;
1974 foreach ( split / /, $right ) {
1976 $_ =~ s/^\s+|\s+$//;
1978 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1979 $sth->execute( $server, $left, $_ )
1980 or warn "execute failed: $!";
1981 while ( my ( $line, $value ) = $sth->fetchrow ) {
1983 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1984 # otherwise, fill the result
1985 $biblionumbers .= $line
1986 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
1987 warn "result : $value "
1988 . ( $right =~ /\d/ ) . "=="
1989 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
1992 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1994 warn "NZAND" if $DEBUG;
1995 $results = NZoperatorAND($biblionumbers,$results);
1997 $results = $biblionumbers;
2002 #do a complete search (all indexes), if index='kw' do complete search too.
2003 my $sth = $dbh->prepare(
2004 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2007 # split each word, query the DB and build the biblionumbers result
2008 foreach ( split / /, $string ) {
2009 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2010 warn "search on all indexes on $_" if $DEBUG;
2013 $sth->execute( $server, $_ );
2014 while ( my $line = $sth->fetchrow ) {
2015 $biblionumbers .= $line;
2018 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2020 $results = NZoperatorAND($biblionumbers,$results);
2023 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2024 $results = $biblionumbers;
2028 warn "return : $results for LEAF : $string" if $DEBUG;
2031 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2035 my ($rightresult, $leftresult)=@_;
2037 my @leftresult = split /;/, $leftresult;
2038 warn " @leftresult / $rightresult \n" if $DEBUG;
2040 # my @rightresult = split /;/,$leftresult;
2043 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2044 # the result is stored twice, to have the same weight for AND than OR.
2045 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2046 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2047 foreach (@leftresult) {
2050 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2051 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2052 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2054 "$value-$countvalue;$value-$countvalue;";
2057 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2058 return $finalresult;
2062 my ($rightresult, $leftresult)=@_;
2063 return $rightresult.$leftresult;
2067 my ($leftresult, $rightresult)=@_;
2069 my @leftresult = split /;/, $leftresult;
2071 # my @rightresult = split /;/,$leftresult;
2073 foreach (@leftresult) {
2075 $value=$1 if $value=~m/(.*)-\d+$/;
2076 unless ($rightresult =~ "$value-") {
2077 $finalresult .= "$_;";
2080 return $finalresult;
2085 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2092 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2093 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2095 # order title asc by default
2096 # $ordering = '1=36 <i' unless $ordering;
2097 $results_per_page = 20 unless $results_per_page;
2098 $offset = 0 unless $offset;
2099 my $dbh = C4::Context->dbh;
2102 # order by POPULARITY
2104 if ( $ordering =~ /popularity/ ) {
2108 # popularity is not in MARC record, it's builded from a specific query
2110 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2111 foreach ( split /;/, $biblionumbers ) {
2112 my ( $biblionumber, $title ) = split /,/, $_;
2113 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2114 $sth->execute($biblionumber);
2115 my $popularity = $sth->fetchrow || 0;
2117 # hint : the key is popularity.title because we can have
2118 # many results with the same popularity. In this case, sub-ordering is done by title
2119 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2120 # (un-frequent, I agree, but we won't forget anything that way ;-)
2121 $popularity{ sprintf( "%10d", $popularity ) . $title
2122 . $biblionumber } = $biblionumber;
2125 # sort the hash and return the same structure as GetRecords (Zebra querying)
2128 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2129 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2130 $result_hash->{'RECORDS'}[ $numbers++ ] =
2131 $result{ $popularity{$key} }->as_usmarc();
2134 else { # sort popularity ASC
2135 foreach my $key ( sort ( keys %popularity ) ) {
2136 $result_hash->{'RECORDS'}[ $numbers++ ] =
2137 $result{ $popularity{$key} }->as_usmarc();
2140 my $finalresult = ();
2141 $result_hash->{'hits'} = $numbers;
2142 $finalresult->{'biblioserver'} = $result_hash;
2143 return $finalresult;
2149 elsif ( $ordering =~ /author/ ) {
2151 foreach ( split /;/, $biblionumbers ) {
2152 my ( $biblionumber, $title ) = split /,/, $_;
2153 my $record = GetMarcBiblio($biblionumber);
2155 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2156 $author = $record->subfield( '200', 'f' );
2157 $author = $record->subfield( '700', 'a' ) unless $author;
2160 $author = $record->subfield( '100', 'a' );
2163 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2164 # and we don't want to get only 1 result for each of them !!!
2165 $result{ $author . $biblionumber } = $record;
2168 # sort the hash and return the same structure as GetRecords (Zebra querying)
2171 if ( $ordering eq 'author_za' ) { # sort by author desc
2172 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2173 $result_hash->{'RECORDS'}[ $numbers++ ] =
2174 $result{$key}->as_usmarc();
2177 else { # sort by author ASC
2178 foreach my $key ( sort ( keys %result ) ) {
2179 $result_hash->{'RECORDS'}[ $numbers++ ] =
2180 $result{$key}->as_usmarc();
2183 my $finalresult = ();
2184 $result_hash->{'hits'} = $numbers;
2185 $finalresult->{'biblioserver'} = $result_hash;
2186 return $finalresult;
2189 # ORDER BY callnumber
2192 elsif ( $ordering =~ /callnumber/ ) {
2194 foreach ( split /;/, $biblionumbers ) {
2195 my ( $biblionumber, $title ) = split /,/, $_;
2196 my $record = GetMarcBiblio($biblionumber);
2198 my $frameworkcode = GetFrameworkCode($biblionumber);
2199 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2200 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2201 unless $callnumber_tag;
2202 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2203 $callnumber = $record->subfield( '200', 'f' );
2205 $callnumber = $record->subfield( '100', 'a' );
2208 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2209 # and we don't want to get only 1 result for each of them !!!
2210 $result{ $callnumber . $biblionumber } = $record;
2213 # sort the hash and return the same structure as GetRecords (Zebra querying)
2216 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2217 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2218 $result_hash->{'RECORDS'}[ $numbers++ ] =
2219 $result{$key}->as_usmarc();
2222 else { # sort by title ASC
2223 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2224 $result_hash->{'RECORDS'}[ $numbers++ ] =
2225 $result{$key}->as_usmarc();
2228 my $finalresult = ();
2229 $result_hash->{'hits'} = $numbers;
2230 $finalresult->{'biblioserver'} = $result_hash;
2231 return $finalresult;
2233 elsif ( $ordering =~ /pubdate/ ) { #pub year
2235 foreach ( split /;/, $biblionumbers ) {
2236 my ( $biblionumber, $title ) = split /,/, $_;
2237 my $record = GetMarcBiblio($biblionumber);
2238 my ( $publicationyear_tag, $publicationyear_subfield ) =
2239 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2240 my $publicationyear =
2241 $record->subfield( $publicationyear_tag,
2242 $publicationyear_subfield );
2244 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2245 # and we don't want to get only 1 result for each of them !!!
2246 $result{ $publicationyear . $biblionumber } = $record;
2249 # sort the hash and return the same structure as GetRecords (Zebra querying)
2252 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2253 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2254 $result_hash->{'RECORDS'}[ $numbers++ ] =
2255 $result{$key}->as_usmarc();
2258 else { # sort by pub year ASC
2259 foreach my $key ( sort ( keys %result ) ) {
2260 $result_hash->{'RECORDS'}[ $numbers++ ] =
2261 $result{$key}->as_usmarc();
2264 my $finalresult = ();
2265 $result_hash->{'hits'} = $numbers;
2266 $finalresult->{'biblioserver'} = $result_hash;
2267 return $finalresult;
2273 elsif ( $ordering =~ /title/ ) {
2275 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2277 foreach ( split /;/, $biblionumbers ) {
2278 my ( $biblionumber, $title ) = split /,/, $_;
2280 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2281 # and we don't want to get only 1 result for each of them !!!
2282 # hint & speed improvement : we can order without reading the record
2283 # so order, and read records only for the requested page !
2284 $result{ $title . $biblionumber } = $biblionumber;
2287 # sort the hash and return the same structure as GetRecords (Zebra querying)
2290 if ( $ordering eq 'title_az' ) { # sort by title desc
2291 foreach my $key ( sort ( keys %result ) ) {
2292 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2295 else { # sort by title ASC
2296 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2297 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2301 # limit the $results_per_page to result size if it's more
2302 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2304 # for the requested page, replace biblionumber by the complete record
2305 # speed improvement : avoid reading too much things
2307 my $counter = $offset ;
2308 $counter <= $offset + $results_per_page ;
2312 $result_hash->{'RECORDS'}[$counter] =
2313 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2315 my $finalresult = ();
2316 $result_hash->{'hits'} = $numbers;
2317 $finalresult->{'biblioserver'} = $result_hash;
2318 return $finalresult;
2325 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2328 foreach ( split /;/, $biblionumbers ) {
2329 my ( $biblionumber, $title ) = split /,/, $_;
2330 $title =~ /(.*)-(\d)/;
2335 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2336 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2337 # biblio N has ranking = 6
2338 $count_ranking{$biblionumber} += $ranking;
2341 # build the result by "inverting" the count_ranking hash
2342 # 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
2344 foreach ( keys %count_ranking ) {
2345 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2348 # sort the hash and return the same structure as GetRecords (Zebra querying)
2351 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2352 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2355 # limit the $results_per_page to result size if it's more
2356 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2358 # for the requested page, replace biblionumber by the complete record
2359 # speed improvement : avoid reading too much things
2361 my $counter = $offset ;
2362 $counter <= $offset + $results_per_page ;
2366 $result_hash->{'RECORDS'}[$counter] =
2367 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2368 if $result_hash->{'RECORDS'}[$counter];
2370 my $finalresult = ();
2371 $result_hash->{'hits'} = $numbers;
2372 $finalresult->{'biblioserver'} = $result_hash;
2373 return $finalresult;
2377 =head2 enabled_staff_search_views
2379 %hash = enabled_staff_search_views()
2381 This function returns a hash that contains three flags obtained from the system
2382 preferences, used to determine whether a particular staff search results view
2387 =item C<Output arg:>
2389 * $hash{can_view_MARC} is true only if the MARC view is enabled
2390 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2391 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2393 =item C<usage in the script:>
2397 $template->param ( C4::Search::enabled_staff_search_views );
2401 sub enabled_staff_search_views
2404 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2405 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2406 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2410 sub AddSearchHistory{
2411 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2412 my $dbh = C4::Context->dbh;
2414 # Add the request the user just made
2415 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2416 my $sth = $dbh->prepare($sql);
2417 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2418 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2421 sub GetSearchHistory{
2422 my ($borrowernumber,$session)=@_;
2423 my $dbh = C4::Context->dbh;
2425 # Add the request the user just made
2426 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2427 my $sth = $dbh->prepare($query);
2428 $sth->execute($borrowernumber, $session);
2429 return $sth->fetchall_hashref({});
2432 =head2 z3950_search_args
2434 $arrayref = z3950_search_args($matchpoints)
2436 This function returns an array reference that contains the search parameters to be
2437 passed to the Z39.50 search script (z3950_search.pl). The array elements
2438 are hash refs whose keys are name, value and encvalue, and whose values are the
2439 name of a search parameter, the value of that search parameter and the URL encoded
2440 value of that parameter.
2442 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2444 The search parameter values are obtained from the bibliographic record whose
2445 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2447 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2448 a general purpose search argument. In this case, the returned array contains only
2449 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2451 If a search parameter value is undefined or empty, it is not included in the returned
2454 The returned array reference may be passed directly to the template parameters.
2458 =item C<Output arg:>
2460 * $array containing hash refs as described above
2462 =item C<usage in the script:>
2466 $data = Biblio::GetBiblioData($bibno);
2467 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2471 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2475 sub z3950_search_args {
2477 $bibrec = { title => $bibrec } if !ref $bibrec;
2479 for my $field (qw/ lccn isbn issn title author dewey subject /)
2481 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2482 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2487 =head2 BiblioAddAuthorities
2489 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2491 this function finds the authorities linked to the biblio
2492 * search in the authority DB for the same authid (in $9 of the biblio)
2493 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2494 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2495 OR adds a new authority record
2501 * $record is the MARC record in question (marc blob)
2502 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2504 =item C<Output arg:>
2506 * $countlinked is the number of authorities records that are linked to this authority
2510 * 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)
2516 sub BiblioAddAuthorities{
2517 my ( $record, $frameworkcode ) = @_;
2518 my $dbh=C4::Context->dbh;
2519 my $query=$dbh->prepare(qq|
2520 SELECT authtypecode,tagfield
2521 FROM marc_subfield_structure
2522 WHERE frameworkcode=?
2523 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2524 # SELECT authtypecode,tagfield
2525 # FROM marc_subfield_structure
2526 # WHERE frameworkcode=?
2527 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2528 $query->execute($frameworkcode);
2529 my ($countcreated,$countlinked);
2530 while (my $data=$query->fetchrow_hashref){
2531 foreach my $field ($record->field($data->{tagfield})){
2532 next if ($field->subfield('3')||$field->subfield('9'));
2533 # No authorities id in the tag.
2534 # Search if there is any authorities to link to.
2535 my $query='at='.$data->{authtypecode}.' ';
2536 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2537 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2538 # there is only 1 result
2540 warn "BIBLIOADDSAUTHORITIES: $error";
2543 if ($results && scalar(@$results)==1) {
2544 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2545 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2547 } elsif (scalar(@$results)>1) {
2548 #More than One result
2549 #This can comes out of a lack of a subfield.
2550 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2551 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2554 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2555 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2556 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2557 my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2558 next unless $authtypedata;
2559 my $marcrecordauth=MARC::Record->new();
2560 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2561 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2562 $marcrecordauth->insert_fields_ordered($authfield);
2564 # bug 2317: ensure new authority knows it's using UTF-8; currently
2565 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2566 # automatically for UNIMARC (by not transcoding)
2567 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2568 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2569 # of change to a core API just before the 3.0 release.
2570 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2571 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2574 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2576 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2578 $field->add_subfields('9'=>$authid);
2582 return ($countlinked,$countcreated);
2585 =head2 GetDistinctValues($field);
2587 C<$field> is a reference to the fields array
2591 sub GetDistinctValues {
2592 my ($fieldname,$string)=@_;
2593 # returns a reference to a hash of references to branches...
2594 if ($fieldname=~/\./){
2595 my ($table,$column)=split /\./, $fieldname;
2596 my $dbh = C4::Context->dbh;
2597 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column ";
2598 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 ");
2600 my $elements=$sth->fetchall_arrayref({});
2605 my @servers=qw<biblioserver authorityserver>;
2606 my (@zconns,@results);
2607 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2608 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2611 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2614 # The big moment: asynchronously retrieve results from all servers
2616 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2617 my $ev = $zconns[ $i - 1 ]->last_event();
2618 if ( $ev == ZOOM::Event::ZEND ) {
2619 next unless $results[ $i - 1 ];
2620 my $size = $results[ $i - 1 ]->size();
2622 for (my $j=0;$j<$size;$j++){
2624 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2625 push @elements, \%hashscan;
2635 END { } # module clean-up code here (global destructor)
2642 Koha Developement team <info@koha.org>