3 # This file is part of Koha.
5 # Koha is free software; you can redistribute it and/or modify it under the
6 # terms of the GNU General Public License as published by the Free Software
7 # Foundation; either version 2 of the License, or (at your option) any later
10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
14 # You should have received a copy of the GNU General Public License along with
15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
16 # Suite 330, Boston, MA 02111-1307 USA
19 #use warnings; FIXME - Bug 2505
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
25 use C4::Search::PazPar2;
27 use C4::Dates qw(format_date);
28 use C4::Members qw(GetHideLostItemsPreference);
31 use C4::Reserves; # CheckReserves
35 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
37 # set the version for version checking
40 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
45 C4::Search - Functions for searching the Koha catalog.
49 See opac/opac-search.pl or catalogue/search.pl for example of usage
53 This module provides searching functions for Koha's bibliographic databases
71 #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)
73 # make all your functions, whether exported or not;
77 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
79 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
85 my $dbh = C4::Context->dbh;
86 my $result = TransformMarcToKoha( $dbh, $record, '' );
91 my ( $biblionumber, $title );
93 # search duplicate on ISBN, easy and fast..
95 if ( $result->{isbn} ) {
96 $result->{isbn} =~ s/\(.*$//;
97 $result->{isbn} =~ s/\s+$//;
98 $query = "isbn=$result->{isbn}";
101 $result->{title} =~ s /\\//g;
102 $result->{title} =~ s /\"//g;
103 $result->{title} =~ s /\(//g;
104 $result->{title} =~ s /\)//g;
106 # FIXME: instead of removing operators, could just do
107 # quotes around the value
108 $result->{title} =~ s/(and|or|not)//g;
109 $query = "ti,ext=$result->{title}";
110 $query .= " and itemtype=$result->{itemtype}"
111 if ( $result->{itemtype} );
112 if ( $result->{author} ) {
113 $result->{author} =~ s /\\//g;
114 $result->{author} =~ s /\"//g;
115 $result->{author} =~ s /\(//g;
116 $result->{author} =~ s /\)//g;
118 # remove valid operators
119 $result->{author} =~ s/(and|or|not)//g;
120 $query .= " and au,ext=$result->{author}";
124 # FIXME: add error handling
125 my ( $error, $searchresults ) = SimpleSearch($query); # FIXME :: hardcoded !
127 foreach my $possible_duplicate_record (@$searchresults) {
129 MARC::Record->new_from_usmarc($possible_duplicate_record);
130 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
132 # FIXME :: why 2 $biblionumber ?
134 push @results, $result->{'biblionumber'};
135 push @results, $result->{'title'};
143 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
145 This function provides a simple search API on the bibliographic catalog
151 * $query can be a simple keyword or a complete CCL query
152 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
153 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
154 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
159 * $error is a empty unless an error is detected
160 * \@results is an array of records.
161 * $total_hits is the number of hits that would have been returned with no limit
163 =item C<usage in the script:>
167 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
169 if (defined $error) {
170 $template->param(query_error => $error);
171 warn "error: ".$error;
172 output_html_with_http_headers $input, $cookie, $template->output;
176 my $hits = scalar @$marcresults;
179 for my $i (0..$hits) {
181 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
182 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
184 #build the hash for the template.
185 $resultsloop{title} = $biblio->{'title'};
186 $resultsloop{subtitle} = $biblio->{'subtitle'};
187 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
188 $resultsloop{author} = $biblio->{'author'};
189 $resultsloop{publishercode} = $biblio->{'publishercode'};
190 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
192 push @results, \%resultsloop;
195 $template->param(result=>\@results);
200 my ( $query, $offset, $max_results, $servers ) = @_;
202 if ( C4::Context->preference('NoZebra') ) {
203 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
206 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
207 return ( undef, $search_result, scalar($result->{hits}) );
210 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
211 my @servers = defined ( $servers ) ? @$servers : ( "biblioserver" );
217 return ( "No query entered", undef, undef ) unless $query;
219 # Initialize & Search Zebra
220 for ( my $i = 0 ; $i < @servers ; $i++ ) {
222 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
223 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
224 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
228 $zconns[$i]->errmsg() . " ("
229 . $zconns[$i]->errcode() . ") "
230 . $zconns[$i]->addinfo() . " "
231 . $zconns[$i]->diagset();
233 return ( $error, undef, undef ) if $zconns[$i]->errcode();
237 # caught a ZOOM::Exception
241 . $@->addinfo() . " "
244 return ( $error, undef, undef );
247 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
248 my $event = $zconns[ $i - 1 ]->last_event();
249 if ( $event == ZOOM::Event::ZEND ) {
251 my $first_record = defined( $offset ) ? $offset+1 : 1;
252 my $hits = $tmpresults[ $i - 1 ]->size();
253 $total_hits += $hits;
254 my $last_record = $hits;
255 if ( defined $max_results && $offset + $max_results < $hits ) {
256 $last_record = $offset + $max_results;
259 for my $j ( $first_record..$last_record ) {
260 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
261 push @results, $record;
266 foreach my $result (@tmpresults) {
269 foreach my $zoom_query (@zoom_queries) {
270 $zoom_query->destroy();
273 return ( undef, \@results, $total_hits );
279 ( undef, $results_hashref, \@facets_loop ) = getRecords (
281 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
282 $results_per_page, $offset, $expanded_facet, $branches,
286 The all singing, all dancing, multi-server, asynchronous, scanning,
287 searching, record nabbing, facet-building
289 See verbse embedded documentation.
295 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
296 $results_per_page, $offset, $expanded_facet, $branches,
300 my @servers = @$servers_ref;
301 my @sort_by = @$sort_by_ref;
303 # Initialize variables for the ZOOM connection and results object
307 my $results_hashref = ();
309 # Initialize variables for the faceted results objects
310 my $facets_counter = ();
311 my $facets_info = ();
312 my $facets = getFacets();
313 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
315 my @facets_loop; # stores the ref to array of hashes for template facets loop
317 ### LOOP THROUGH THE SERVERS
318 for ( my $i = 0 ; $i < @servers ; $i++ ) {
319 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
321 # perform the search, create the results objects
322 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
323 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
325 #$query_to_use = $simple_query if $scan;
326 warn $simple_query if ( $scan and $DEBUG );
328 # Check if we've got a query_type defined, if so, use it
331 if ($query_type =~ /^ccl/) {
332 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
333 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
334 } elsif ($query_type =~ /^cql/) {
335 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
336 } elsif ($query_type =~ /^pqf/) {
337 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
339 warn "Unknown query_type '$query_type'. Results undetermined.";
342 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
344 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
348 warn "WARNING: query problem with $query_to_use " . $@;
351 # Concatenate the sort_by limits and pass them to the results object
352 # Note: sort will override rank
354 foreach my $sort (@sort_by) {
355 if ( $sort eq "author_az" ) {
356 $sort_by .= "1=1003 <i ";
358 elsif ( $sort eq "author_za" ) {
359 $sort_by .= "1=1003 >i ";
361 elsif ( $sort eq "popularity_asc" ) {
362 $sort_by .= "1=9003 <i ";
364 elsif ( $sort eq "popularity_dsc" ) {
365 $sort_by .= "1=9003 >i ";
367 elsif ( $sort eq "call_number_asc" ) {
368 $sort_by .= "1=8007 <i ";
370 elsif ( $sort eq "call_number_dsc" ) {
371 $sort_by .= "1=8007 >i ";
373 elsif ( $sort eq "pubdate_asc" ) {
374 $sort_by .= "1=31 <i ";
376 elsif ( $sort eq "pubdate_dsc" ) {
377 $sort_by .= "1=31 >i ";
379 elsif ( $sort eq "acqdate_asc" ) {
380 $sort_by .= "1=32 <i ";
382 elsif ( $sort eq "acqdate_dsc" ) {
383 $sort_by .= "1=32 >i ";
385 elsif ( $sort eq "title_az" ) {
386 $sort_by .= "1=4 <i ";
388 elsif ( $sort eq "title_za" ) {
389 $sort_by .= "1=4 >i ";
392 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
395 if ($sort_by && !$scan) {
396 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
397 warn "WARNING sort $sort_by failed";
400 } # finished looping through servers
402 # The big moment: asynchronously retrieve results from all servers
403 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
404 my $ev = $zconns[ $i - 1 ]->last_event();
405 if ( $ev == ZOOM::Event::ZEND ) {
406 next unless $results[ $i - 1 ];
407 my $size = $results[ $i - 1 ]->size();
411 # loop through the results
412 $results_hash->{'hits'} = $size;
414 if ( $offset + $results_per_page <= $size ) {
415 $times = $offset + $results_per_page;
420 for ( my $j = $offset ; $j < $times ; $j++ ) {
424 ## Check if it's an index scan
426 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
428 # here we create a minimal MARC record and hand it off to the
429 # template just like a normal result ... perhaps not ideal, but
431 my $tmprecord = MARC::Record->new();
432 $tmprecord->encoding('UTF-8');
436 # the minimal record in author/title (depending on MARC flavour)
437 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
438 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
439 $tmprecord->append_fields($tmptitle);
441 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
442 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
443 $tmprecord->append_fields($tmptitle);
444 $tmprecord->append_fields($tmpauthor);
446 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
451 $record = $results[ $i - 1 ]->record($j)->raw();
453 # warn "RECORD $j:".$record;
454 $results_hash->{'RECORDS'}[$j] = $record;
458 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
460 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
461 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
463 my $jmax = $size>$facets_maxrecs? $facets_maxrecs: $size;
465 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
466 ($facets->[$k]) or next;
467 my @fcodes = @{$facets->[$k]->{'tags'}};
468 my $sfcode = $facets->[$k]->{'subfield'};
470 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
471 my $render_record = $results[ $i - 1 ]->record($j)->render();
474 foreach my $fcode (@fcodes) {
477 my $field_pattern = '\n'.$fcode.' ([^\n]+)';
478 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
480 foreach my $field_token (@field_tokens) {
481 my $subfield_pattern = '\$'.$sfcode.' ([^\$]+)';
482 my @subfield_values = ( $field_token =~ /$subfield_pattern/g );
484 foreach my $subfield_value (@subfield_values) {
486 my $data = $subfield_value;
487 $data =~ s/^\s+//; # trim left
488 $data =~ s/\s+$//; # trim right
490 unless ( $data ~~ @used_datas ) {
491 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
492 push @used_datas, $data;
499 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} = $facets->[$k]->{'label_value'};
500 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} = $facets->[$k]->{'expanded'};
506 # warn "connection ", $i-1, ": $size hits";
507 # warn $results[$i-1]->record(0)->render() if $size > 0;
510 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
512 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
513 keys %$facets_counter )
516 my $number_of_facets;
517 my @this_facets_array;
520 $facets_counter->{$link_value}->{$b}
521 <=> $facets_counter->{$link_value}->{$a}
522 } keys %{ $facets_counter->{$link_value} }
526 if ( ( $number_of_facets < 6 )
527 || ( $expanded_facet eq $link_value )
528 || ( $facets_info->{$link_value}->{'expanded'} ) )
531 # Sanitize the link value ), ( will cause errors with CCL,
532 my $facet_link_value = $one_facet;
533 $facet_link_value =~ s/(\(|\))/ /g;
535 # fix the length that will display in the label,
536 my $facet_label_value = $one_facet;
537 my $facet_max_length =
538 C4::Context->preference('FacetLabelTruncationLength') || 20;
540 substr( $one_facet, 0, $facet_max_length ) . "..."
541 if length($facet_label_value) > $facet_max_length;
543 # if it's a branch, label by the name, not the code,
544 if ( $link_value =~ /branch/ ) {
545 if (defined $branches
546 && ref($branches) eq "HASH"
547 && defined $branches->{$one_facet}
548 && ref ($branches->{$one_facet}) eq "HASH")
551 $branches->{$one_facet}->{'branchname'};
554 $facet_label_value = "*";
558 # but we're down with the whole label being in the link's title.
559 push @this_facets_array, {
560 facet_count => $facets_counter->{$link_value}->{$one_facet},
561 facet_label_value => $facet_label_value,
562 facet_title_value => $one_facet,
563 facet_link_value => $facet_link_value,
564 type_link_value => $link_value,
569 # handle expanded option
570 unless ( $facets_info->{$link_value}->{'expanded'} ) {
572 if ( ( $number_of_facets > 6 )
573 && ( $expanded_facet ne $link_value ) );
576 type_link_value => $link_value,
577 type_id => $link_value . "_id",
578 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
579 facets => \@this_facets_array,
580 expandable => $expandable,
581 expand => $link_value,
582 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
587 return ( undef, $results_hashref, \@facets_loop );
592 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
593 $results_per_page, $offset, $expanded_facet, $branches,
597 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
599 $paz->search($simple_query);
600 sleep 1; # FIXME: WHY?
603 my $results_hashref = {};
604 my $stats = XMLin($paz->stat);
605 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
607 # for a grouped search result, the number of hits
608 # is the number of groups returned; 'bib_hits' will have
609 # the total number of bibs.
610 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
611 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
613 HIT: foreach my $hit (@{ $results->{'hit'} }) {
614 my $recid = $hit->{recid}->[0];
616 my $work_title = $hit->{'md-work-title'}->[0];
618 if (exists $hit->{'md-work-author'}) {
619 $work_author = $hit->{'md-work-author'}->[0];
621 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
623 my $result_group = {};
624 $result_group->{'group_label'} = $group_label;
625 $result_group->{'group_merge_key'} = $recid;
628 if (exists $hit->{count}) {
629 $count = $hit->{count}->[0];
631 $result_group->{'group_count'} = $count;
633 for (my $i = 0; $i < $count; $i++) {
634 # FIXME -- may need to worry about diacritics here
635 my $rec = $paz->record($recid, $i);
636 push @{ $result_group->{'RECORDS'} }, $rec;
639 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
642 # pass through facets
643 my $termlist_xml = $paz->termlist('author,subject');
644 my $terms = XMLin($termlist_xml, forcearray => 1);
645 my @facets_loop = ();
646 #die Dumper($results);
647 # foreach my $list (sort keys %{ $terms->{'list'} }) {
649 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
651 # facet_label_value => $facet->{'name'}->[0],
654 # push @facets_loop, ( {
655 # type_label => $list,
656 # facets => \@facets,
660 return ( undef, $results_hashref, \@facets_loop );
664 sub _remove_stopwords {
665 my ( $operand, $index ) = @_;
666 my @stopwords_removed;
668 # phrase and exact-qualified indexes shouldn't have stopwords removed
669 if ( $index !~ m/phr|ext/ ) {
671 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
672 # we use IsAlpha unicode definition, to deal correctly with diacritics.
673 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
674 # is a stopword, we'd get "çon" and wouldn't find anything...
676 foreach ( keys %{ C4::Context->stopwords } ) {
677 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
678 if ( my ($matched) = ($operand =~
679 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
681 $operand =~ s/\Q$matched\E/ /gi;
682 push @stopwords_removed, $_;
686 return ( $operand, \@stopwords_removed );
690 sub _detect_truncation {
691 my ( $operand, $index ) = @_;
692 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
695 my @wordlist = split( /\s/, $operand );
696 foreach my $word (@wordlist) {
697 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
698 push @rightlefttruncated, $word;
700 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
701 push @lefttruncated, $word;
703 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
704 push @righttruncated, $word;
706 elsif ( index( $word, "*" ) < 0 ) {
707 push @nontruncated, $word;
710 push @regexpr, $word;
714 \@nontruncated, \@righttruncated, \@lefttruncated,
715 \@rightlefttruncated, \@regexpr
720 sub _build_stemmed_operand {
721 my ($operand,$lang) = @_;
722 require Lingua::Stem::Snowball ;
725 # If operand contains a digit, it is almost certainly an identifier, and should
726 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
727 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
728 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
729 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
730 return $operand if $operand =~ /\d/;
732 # FIXME: the locale should be set based on the user's language and/or search choice
734 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
735 encoding => "UTF-8" );
737 my @words = split( / /, $operand );
738 my @stems = $stemmer->stem(\@words);
739 for my $stem (@stems) {
740 $stemmed_operand .= "$stem";
741 $stemmed_operand .= "?"
742 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
743 $stemmed_operand .= " ";
745 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
746 return $stemmed_operand;
750 sub _build_weighted_query {
752 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
753 # pretty well but could work much better if we had a smarter query parser
754 my ( $operand, $stemmed_operand, $index ) = @_;
755 my $stemming = C4::Context->preference("QueryStemming") || 0;
756 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
757 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
759 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
761 # Keyword, or, no index specified
762 if ( ( $index eq 'kw' ) || ( !$index ) ) {
764 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
765 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
766 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
767 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
768 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
769 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
770 if $fuzzy_enabled; # add fuzzy, word list
771 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
772 if ( $stemming and $stemmed_operand )
773 ; # add stemming, right truncation
774 $weighted_query .= " or wrdl,r9=\"$operand\"";
776 # embedded sorting: 0 a-z; 1 z-a
777 # $weighted_query .= ") or (sort1,aut=1";
780 # Barcode searches should skip this process
781 elsif ( $index eq 'bc' ) {
782 $weighted_query .= "bc=\"$operand\"";
785 # Authority-number searches should skip this process
786 elsif ( $index eq 'an' ) {
787 $weighted_query .= "an=\"$operand\"";
790 # If the index already has more than one qualifier, wrap the operand
791 # in quotes and pass it back (assumption is that the user knows what they
792 # are doing and won't appreciate us mucking up their query
793 elsif ( $index =~ ',' ) {
794 $weighted_query .= " $index=\"$operand\"";
797 #TODO: build better cases based on specific search indexes
799 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
800 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
801 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
803 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
806 $weighted_query .= "))"; # close rank specification
807 return $weighted_query;
812 Return an array with available indexes.
834 'Author-personal-bibliography',
843 'Chronological-subdivision',
853 'Conference-name-heading',
854 'Conference-name-see',
855 'Conference-name-seealso',
860 'Corporate-name-heading',
861 'Corporate-name-see',
862 'Corporate-name-seealso',
864 'date-entered-on-file',
865 'Date-of-acquisition',
866 'Date-of-publication',
867 'Dewey-classification',
873 'Geographic-subdivision',
876 'Heading-use-main-or-added-entry',
877 'Heading-use-series-added-entry ',
878 'Heading-use-subject-added-entry',
896 'Local-classification',
899 'Match-heading-see-from',
906 'Name-geographic-heading',
907 'Name-geographic-see',
908 'Name-geographic-seealso',
916 'Personal-name-heading',
918 'Personal-name-seealso',
925 'Record-control-number',
936 'Subject-heading-thesaurus',
937 'Subject-name-personal',
938 'Subject-subdivision',
947 'Term-genre-form-heading',
948 'Term-genre-form-see',
949 'Term-genre-form-seealso',
955 'Title-uniform-heading',
957 'Title-uniform-seealso',
967 'classification-source',
969 'coded-location-qualifier',
980 'Local-classification',
983 'materials-specified',
992 'replacementpricedate',
1010 $simple_query, $query_cgi,
1011 $query_desc, $limit,
1012 $limit_cgi, $limit_desc,
1013 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1015 Build queries and limits in CCL, CGI, Human,
1016 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1018 See verbose embedded documentation.
1024 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1026 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1029 my @operators = $operators ? @$operators : ();
1030 my @indexes = $indexes ? @$indexes : ();
1031 my @operands = $operands ? @$operands : ();
1032 my @limits = $limits ? @$limits : ();
1033 my @sort_by = $sort_by ? @$sort_by : ();
1035 my $stemming = C4::Context->preference("QueryStemming") || 0;
1036 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1037 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1038 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1039 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1041 # no stemming/weight/fuzzy in NoZebra
1042 if ( C4::Context->preference("NoZebra") ) {
1046 $auto_truncation = 0;
1049 my $query = $operands[0];
1050 my $simple_query = $operands[0];
1052 # initialize the variables we're passing back
1061 my $stopwords_removed; # flag to determine if stopwords have been removed
1064 my $cclindexes = getIndexes();
1065 if( $query !~ /\s*ccl=/ ){
1066 for my $index (@$cclindexes){
1067 if($query =~ /($index)(,?\w)*[:=]/){
1071 $query = "ccl=$query" if($cclq);
1074 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1076 if ( $query =~ /^ccl=/ ) {
1078 # This is needed otherwise ccl= and &limit won't work together, and
1079 # this happens when selecting a subject on the opac-detail page
1081 $q .= ' and '.join(' and ', @limits);
1083 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1085 if ( $query =~ /^cql=/ ) {
1086 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1088 if ( $query =~ /^pqf=/ ) {
1089 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1092 # pass nested queries directly
1093 # FIXME: need better handling of some of these variables in this case
1094 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1095 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1096 # if ( $query =~ /(\(|\))/ ) {
1098 # undef, $query, $simple_query, $query_cgi,
1099 # $query, $limit, $limit_cgi, $limit_desc,
1100 # $stopwords_removed, 'ccl'
1104 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1105 # query operands and indexes and add stemming, truncation, field weighting, etc.
1106 # Once we do so, we'll end up with a value in $query, just like if we had an
1107 # incoming $query from the user
1110 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1111 my $previous_operand
1112 ; # a flag used to keep track if there was a previous query
1113 # if there was, we can apply the current operator
1115 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1117 # COMBINE OPERANDS, INDEXES AND OPERATORS
1118 if ( $operands[$i] ) {
1119 $operands[$i]=~s/^\s+//;
1121 # A flag to determine whether or not to add the index to the query
1124 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1125 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1128 $remove_stopwords = 0;
1130 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1132 my $operand = $operands[$i];
1133 my $index = $indexes[$i];
1135 # Add index-specific attributes
1136 # Date of Publication
1137 if ( $index eq 'yr' ) {
1138 $index .= ",st-numeric";
1140 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1143 # Date of Acquisition
1144 elsif ( $index eq 'acqdate' ) {
1145 $index .= ",st-date-normalized";
1147 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1149 # ISBN,ISSN,Standard Number, don't need special treatment
1150 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1152 $stemming, $auto_truncation,
1153 $weight_fields, $fuzzy_enabled,
1155 ) = ( 0, 0, 0, 0, 0 );
1163 # Set default structure attribute (word list)
1164 my $struct_attr = q{};
1165 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1166 $struct_attr = ",wrdl";
1169 # Some helpful index variants
1170 my $index_plus = $index . $struct_attr . ':';
1171 my $index_plus_comma = $index . $struct_attr . ',';
1174 if ($remove_stopwords) {
1175 ( $operand, $stopwords_removed ) =
1176 _remove_stopwords( $operand, $index );
1177 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1178 warn "REMOVED STOPWORDS: @$stopwords_removed"
1179 if ( $stopwords_removed && $DEBUG );
1182 if ($auto_truncation){
1183 unless ( $index =~ /(st-|phr|ext)/ ) {
1184 #FIXME only valid with LTR scripts
1185 $operand=join(" ",map{
1186 (index($_,"*")>0?"$_":"$_*")
1187 }split (/\s+/,$operand));
1188 warn $operand if $DEBUG;
1193 my $truncated_operand;
1194 my( $nontruncated, $righttruncated, $lefttruncated,
1195 $rightlefttruncated, $regexpr
1196 ) = _detect_truncation( $operand, $index );
1198 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1203 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1204 scalar(@$rightlefttruncated) > 0 )
1207 # Don't field weight or add the index to the query, we do it here
1209 undef $weight_fields;
1210 my $previous_truncation_operand;
1211 if (scalar @$nontruncated) {
1212 $truncated_operand .= "$index_plus @$nontruncated ";
1213 $previous_truncation_operand = 1;
1215 if (scalar @$righttruncated) {
1216 $truncated_operand .= "and " if $previous_truncation_operand;
1217 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1218 $previous_truncation_operand = 1;
1220 if (scalar @$lefttruncated) {
1221 $truncated_operand .= "and " if $previous_truncation_operand;
1222 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1223 $previous_truncation_operand = 1;
1225 if (scalar @$rightlefttruncated) {
1226 $truncated_operand .= "and " if $previous_truncation_operand;
1227 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1228 $previous_truncation_operand = 1;
1231 $operand = $truncated_operand if $truncated_operand;
1232 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1235 my $stemmed_operand;
1236 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1239 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1241 # Handle Field Weighting
1242 my $weighted_operand;
1243 if ($weight_fields) {
1244 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1245 $operand = $weighted_operand;
1249 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1251 # If there's a previous operand, we need to add an operator
1252 if ($previous_operand) {
1254 # User-specified operator
1255 if ( $operators[ $i - 1 ] ) {
1256 $query .= " $operators[$i-1] ";
1257 $query .= " $index_plus " unless $indexes_set;
1258 $query .= " $operand";
1259 $query_cgi .= "&op=$operators[$i-1]";
1260 $query_cgi .= "&idx=$index" if $index;
1261 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1263 " $operators[$i-1] $index_plus $operands[$i]";
1266 # Default operator is and
1269 $query .= "$index_plus " unless $indexes_set;
1270 $query .= "$operand";
1271 $query_cgi .= "&op=and&idx=$index" if $index;
1272 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1273 $query_desc .= " and $index_plus $operands[$i]";
1277 # There isn't a pervious operand, don't need an operator
1280 # Field-weighted queries already have indexes set
1281 $query .= " $index_plus " unless $indexes_set;
1283 $query_desc .= " $index_plus $operands[$i]";
1284 $query_cgi .= "&idx=$index" if $index;
1285 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1286 $previous_operand = 1;
1291 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1294 my $group_OR_limits;
1295 my $availability_limit;
1296 foreach my $this_limit (@limits) {
1297 if ( $this_limit =~ /available/ ) {
1299 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1301 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1302 $availability_limit .=
1303 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1304 $limit_cgi .= "&limit=available";
1308 # group_OR_limits, prefixed by mc-
1309 # OR every member of the group
1310 elsif ( $this_limit =~ /mc/ ) {
1312 if ( $this_limit =~ /mc-ccode:/ ) {
1313 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1314 $this_limit =~ tr/"//d;
1315 my ($k,$v) = split(/:/, $this_limit,2);
1316 $this_limit = $k.":\"".$v."\"";
1319 $group_OR_limits .= " or " if $group_OR_limits;
1320 $limit_desc .= " or " if $group_OR_limits;
1321 $group_OR_limits .= "$this_limit";
1322 $limit_cgi .= "&limit=$this_limit";
1323 $limit_desc .= " $this_limit";
1326 # Regular old limits
1328 $limit .= " and " if $limit || $query;
1329 $limit .= "$this_limit";
1330 $limit_cgi .= "&limit=$this_limit";
1331 if ($this_limit =~ /^branch:(.+)/) {
1332 my $branchcode = $1;
1333 my $branchname = GetBranchName($branchcode);
1334 if (defined $branchname) {
1335 $limit_desc .= " branch:$branchname";
1337 $limit_desc .= " $this_limit";
1340 $limit_desc .= " $this_limit";
1344 if ($group_OR_limits) {
1345 $limit .= " and " if ( $query || $limit );
1346 $limit .= "($group_OR_limits)";
1348 if ($availability_limit) {
1349 $limit .= " and " if ( $query || $limit );
1350 $limit .= "($availability_limit)";
1353 # Normalize the query and limit strings
1354 # This is flawed , means we can't search anything with : in it
1355 # if user wants to do ccl or cql, start the query with that
1356 # $query =~ s/:/=/g;
1357 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1358 $query =~ s/(?<=(wrdl)):/=/g;
1359 $query =~ s/(?<=(trn|phr)):/=/g;
1361 for ( $query, $query_desc, $limit, $limit_desc ) {
1362 s/ +/ /g; # remove extra spaces
1363 s/^ //g; # remove any beginning spaces
1364 s/ $//g; # remove any ending spaces
1365 s/==/=/g; # remove double == from query
1367 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1369 for ($query_cgi,$simple_query) {
1372 # append the limit to the query
1373 $query .= " " . $limit;
1377 warn "QUERY:" . $query;
1378 warn "QUERY CGI:" . $query_cgi;
1379 warn "QUERY DESC:" . $query_desc;
1380 warn "LIMIT:" . $limit;
1381 warn "LIMIT CGI:" . $limit_cgi;
1382 warn "LIMIT DESC:" . $limit_desc;
1383 warn "---------\nLeave buildQuery\n---------";
1386 undef, $query, $simple_query, $query_cgi,
1387 $query_desc, $limit, $limit_cgi, $limit_desc,
1388 $stopwords_removed, $query_type
1392 =head2 searchResults
1394 my @search_results = searchResults($search_context, $searchdesc, $hits,
1395 $results_per_page, $offset, $scan,
1396 @marcresults, $hidelostitems);
1398 Format results in a form suitable for passing to the template
1402 # IMO this subroutine is pretty messy still -- it's responsible for
1403 # building the HTML output for the template
1405 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1406 my $dbh = C4::Context->dbh;
1409 $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1411 #Build branchnames hash
1413 #get branch information.....
1415 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1417 while ( my $bdata = $bsth->fetchrow_hashref ) {
1418 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1420 # FIXME - We build an authorised values hash here, using the default framework
1421 # though it is possible to have different authvals for different fws.
1423 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1425 # get notforloan authorised value list (see $shelflocations FIXME)
1426 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1428 #Build itemtype hash
1429 #find itemtype & itemtype image
1433 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1436 while ( my $bdata = $bsth->fetchrow_hashref ) {
1437 foreach (qw(description imageurl summary notforloan)) {
1438 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1442 #search item field code
1445 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1448 my ($itemtag) = $sth->fetchrow;
1450 ## find column names of items related to MARC
1451 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1453 my %subfieldstosearch;
1454 while ( ( my $column ) = $sth2->fetchrow ) {
1455 my ( $tagfield, $tagsubfield ) =
1456 &GetMarcFromKohaField( "items." . $column, "" );
1457 $subfieldstosearch{$column} = $tagsubfield;
1460 # handle which records to actually retrieve
1462 if ( $hits && $offset + $results_per_page <= $hits ) {
1463 $times = $offset + $results_per_page;
1466 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1469 my $marcflavour = C4::Context->preference("marcflavour");
1470 # We get the biblionumber position in MARC
1471 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1474 # loop through all of the records we've retrieved
1475 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1476 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1480 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1481 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1482 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1483 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1484 $oldbiblio->{result_number} = $i + 1;
1486 # add imageurl to itemtype if there is one
1487 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1489 $oldbiblio->{'authorised_value_images'} = ($search_context eq 'opac' && C4::Context->preference('AuthorisedValueImages')) || ($search_context eq 'intranet' && C4::Context->preference('StaffAuthorisedValueImages')) ? C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) ) : [];
1490 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1491 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1492 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1493 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1494 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1496 # edition information, if any
1497 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1498 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1499 # Build summary if there is one (the summary is defined in the itemtypes table)
1500 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1501 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1502 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1503 my @fields = $marcrecord->fields();
1506 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1508 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1509 $tag =~ /(.{3})(.)/;
1510 if($marcrecord->field($1)){
1511 my @abc = $marcrecord->field($1)->subfield($2);
1512 $tags->{$tag} = $#abc + 1 ;
1516 # We catch how many times to repeat this line
1518 foreach my $tag (keys(%$tags)){
1519 $max = $tags->{$tag} if($tags->{$tag} > $max);
1522 # we replace, and repeat each line
1523 for (my $i = 0 ; $i < $max ; $i++){
1524 my $newline = $line;
1526 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1527 $tag =~ /(.{3})(.)/;
1529 if($marcrecord->field($1)){
1530 my @repl = $marcrecord->field($1)->subfield($2);
1531 my $subfieldvalue = $repl[$i];
1533 if (! utf8::is_utf8($subfieldvalue)) {
1534 utf8::decode($subfieldvalue);
1537 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1540 $newsummary .= "$newline\n";
1544 $newsummary =~ s/\[(.*?)]//g;
1545 $newsummary =~ s/\n/<br\/>/g;
1546 $oldbiblio->{summary} = $newsummary;
1549 # Pull out the items fields
1550 my @fields = $marcrecord->field($itemtag);
1552 # Setting item statuses for display
1553 my @available_items_loop;
1554 my @onloan_items_loop;
1555 my @other_items_loop;
1557 my $available_items;
1561 my $ordered_count = 0;
1562 my $available_count = 0;
1563 my $onloan_count = 0;
1564 my $longoverdue_count = 0;
1565 my $other_count = 0;
1566 my $wthdrawn_count = 0;
1567 my $itemlost_count = 0;
1568 my $itembinding_count = 0;
1569 my $itemdamaged_count = 0;
1570 my $item_in_transit_count = 0;
1571 my $can_place_holds = 0;
1572 my $item_onhold_count = 0;
1573 my $items_count = scalar(@fields);
1575 ( C4::Context->preference('maxItemsinSearchResults') )
1576 ? C4::Context->preference('maxItemsinSearchResults') - 1
1579 # loop through every item
1580 foreach my $field (@fields) {
1583 # populate the items hash
1584 foreach my $code ( keys %subfieldstosearch ) {
1585 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1588 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1589 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1590 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1591 if ($item->{$hbranch}) {
1592 $item->{'branchname'} = $branches{$item->{$hbranch}};
1594 elsif ($item->{$otherbranch}) { # Last resort
1595 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1598 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1599 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1600 my $userenv = C4::Context->userenv;
1601 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1603 my $key = $prefix . $item->{onloan} . $item->{barcode};
1604 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1605 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1606 $onloan_items->{$key}->{branchname} = $item->{branchname};
1607 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1608 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1609 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1610 # if something's checked out and lost, mark it as 'long overdue'
1611 if ( $item->{itemlost} ) {
1612 $onloan_items->{$prefix}->{longoverdue}++;
1613 $longoverdue_count++;
1614 } else { # can place holds as long as item isn't lost
1615 $can_place_holds = 1;
1619 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1623 if ( $item->{notforloan} == -1 ) {
1627 # is item in transit?
1628 my $transfertwhen = '';
1629 my ($transfertfrom, $transfertto);
1631 # is item on the reserve shelf?
1632 my $reservestatus = 0;
1635 unless ($item->{wthdrawn}
1636 || $item->{itemlost}
1638 || $item->{notforloan}
1639 || $items_count > 20) {
1641 # A couple heuristics to limit how many times
1642 # we query the database for item transfer information, sacrificing
1643 # accuracy in some cases for speed;
1645 # 1. don't query if item has one of the other statuses
1646 # 2. don't check transit status if the bib has
1647 # more than 20 items
1649 # FIXME: to avoid having the query the database like this, and to make
1650 # the in transit status count as unavailable for search limiting,
1651 # should map transit status to record indexed in Zebra.
1653 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1654 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1657 # item is withdrawn, lost or damaged
1658 if ( $item->{wthdrawn}
1659 || $item->{itemlost}
1661 || $item->{notforloan} > 0
1662 || $reservestatus eq 'Waiting'
1663 || ($transfertwhen ne ''))
1665 $wthdrawn_count++ if $item->{wthdrawn};
1666 $itemlost_count++ if $item->{itemlost};
1667 $itemdamaged_count++ if $item->{damaged};
1668 $item_in_transit_count++ if $transfertwhen ne '';
1669 $item_onhold_count++ if $reservestatus eq 'Waiting';
1670 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1673 my $key = $prefix . $item->{status};
1674 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1675 $other_items->{$key}->{$_} = $item->{$_};
1677 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1678 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1679 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1680 $other_items->{$key}->{count}++ if $item->{$hbranch};
1681 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1682 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1686 $can_place_holds = 1;
1688 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1689 foreach (qw(branchname itemcallnumber)) {
1690 $available_items->{$prefix}->{$_} = $item->{$_};
1692 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1693 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1696 } # notforloan, item level and biblioitem level
1697 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1699 ( C4::Context->preference('maxItemsinSearchResults') )
1700 ? C4::Context->preference('maxItemsinSearchResults') - 1
1702 for my $key ( sort keys %$onloan_items ) {
1703 (++$onloanitemscount > $maxitems) and last;
1704 push @onloan_items_loop, $onloan_items->{$key};
1706 for my $key ( sort keys %$other_items ) {
1707 (++$otheritemscount > $maxitems) and last;
1708 push @other_items_loop, $other_items->{$key};
1710 for my $key ( sort keys %$available_items ) {
1711 (++$availableitemscount > $maxitems) and last;
1712 push @available_items_loop, $available_items->{$key}
1715 # XSLT processing of some stuff
1717 SetUTF8Flag($marcrecord);
1718 $debug && warn $marcrecord->as_formatted;
1719 if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1720 # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1721 # is not implemented yet
1722 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1723 $search_context, 1);
1724 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1728 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1729 if (!C4::Context->preference("item-level_itypes")) {
1730 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1731 $can_place_holds = 0;
1734 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1735 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1736 $oldbiblio->{items_count} = $items_count;
1737 $oldbiblio->{available_items_loop} = \@available_items_loop;
1738 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1739 $oldbiblio->{other_items_loop} = \@other_items_loop;
1740 $oldbiblio->{availablecount} = $available_count;
1741 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1742 $oldbiblio->{onloancount} = $onloan_count;
1743 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1744 $oldbiblio->{othercount} = $other_count;
1745 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1746 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1747 $oldbiblio->{itemlostcount} = $itemlost_count;
1748 $oldbiblio->{damagedcount} = $itemdamaged_count;
1749 $oldbiblio->{intransitcount} = $item_in_transit_count;
1750 $oldbiblio->{onholdcount} = $item_onhold_count;
1751 $oldbiblio->{orderedcount} = $ordered_count;
1752 $oldbiblio->{isbn} =~
1753 s/-//g; # deleting - in isbn to enable amazon content
1755 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1756 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1757 my $subfields = substr $fieldspec, 3;
1758 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1759 my @alternateholdingsinfo = ();
1760 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1761 my $alternateholdingscount = 0;
1763 for my $field (@holdingsfields) {
1764 my %holding = ( holding => '' );
1765 my $havesubfield = 0;
1766 for my $subfield ($field->subfields()) {
1767 if ((index $subfields, $$subfield[0]) >= 0) {
1768 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1769 $holding{'holding'} .= $$subfield[1];
1773 if ($havesubfield) {
1774 push(@alternateholdingsinfo, \%holding);
1775 $alternateholdingscount++;
1779 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1780 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1783 push( @newresults, $oldbiblio )
1784 if(not $hidelostitems
1785 or (($items_count > $itemlost_count )
1786 && $hidelostitems));
1792 =head2 SearchAcquisitions
1793 Search for acquisitions
1796 sub SearchAcquisitions{
1797 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1799 my $dbh=C4::Context->dbh;
1800 # Variable initialization
1804 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1805 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1806 WHERE dateaccessioned BETWEEN ? AND ?
1809 my (@params,@loopcriteria);
1811 push @params, $datebegin->output("iso");
1812 push @params, $dateend->output("iso");
1814 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1815 if(C4::Context->preference("item-level_itypes")){
1816 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1818 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1820 push @params, @$itemtypes;
1823 if ($criteria =~/itemtype/){
1824 if(C4::Context->preference("item-level_itypes")){
1825 $str .= "AND items.itype=? ";
1827 $str .= "AND biblioitems.itemtype=? ";
1830 if(scalar(@$itemtypes) == 0){
1831 my $itypes = GetItemTypes();
1832 for my $key (keys %$itypes){
1833 push @$itemtypes, $key;
1837 @loopcriteria= @$itemtypes;
1838 }elsif ($criteria=~/itemcallnumber/){
1839 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1840 OR items.itemcallnumber is NULL
1841 OR items.itemcallnumber = '')";
1843 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1845 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1846 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1849 if ($orderby =~ /date_desc/){
1850 $str.=" ORDER BY dateaccessioned DESC";
1852 $str.=" ORDER BY title";
1855 my $qdataacquisitions=$dbh->prepare($str);
1857 my @loopacquisitions;
1858 foreach my $value(@loopcriteria){
1859 push @params,$value;
1861 $cell{"title"}=$value;
1862 $cell{"titlecode"}=$value;
1864 eval{$qdataacquisitions->execute(@params);};
1866 if ($@){ warn "recentacquisitions Error :$@";}
1869 while (my $data=$qdataacquisitions->fetchrow_hashref){
1870 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1872 $cell{"loopdata"}=\@loopdata;
1874 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1877 $qdataacquisitions->finish;
1878 return \@loopacquisitions;
1880 #----------------------------------------------------------------------
1882 # Non-Zebra GetRecords#
1883 #----------------------------------------------------------------------
1887 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1893 $query, $simple_query, $sort_by_ref, $servers_ref,
1894 $results_per_page, $offset, $expanded_facet, $branches,
1897 warn "query =$query" if $DEBUG;
1898 my $result = NZanalyse($query);
1899 warn "results =$result" if $DEBUG;
1901 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1907 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1908 the list is built from an inverted index in the nozebra SQL table
1909 note that title is here only for convenience : the sorting will be very fast when requested on title
1910 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1915 my ( $string, $server ) = @_;
1916 # warn "---------" if $DEBUG;
1917 warn " NZanalyse" if $DEBUG;
1918 # warn "---------" if $DEBUG;
1920 # $server contains biblioserver or authorities, depending on what we search on.
1921 #warn "querying : $string on $server";
1922 $server = 'biblioserver' unless $server;
1924 # if we have a ", replace the content to discard temporarily any and/or/not inside
1926 if ( $string =~ /"/ ) {
1927 $string =~ s/"(.*?)"/__X__/;
1929 warn "commacontent : $commacontent" if $DEBUG;
1932 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1933 # then, call again NZanalyse with $left and $right
1934 # (recursive until we find a leaf (=> something without and/or/not)
1935 # delete repeated operator... Would then go in infinite loop
1936 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1939 #process parenthesis before.
1940 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1943 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1945 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1947 my $leftresult = NZanalyse( $left, $server );
1949 my $rightresult = NZanalyse( $right, $server );
1951 # OK, we have the results for right and left part of the query
1952 # depending of operand, intersect, union or exclude both lists
1953 # to get a result list
1954 if ( $operator eq ' and ' ) {
1955 return NZoperatorAND($leftresult,$rightresult);
1957 elsif ( $operator eq ' or ' ) {
1959 # just merge the 2 strings
1960 return $leftresult . $rightresult;
1962 elsif ( $operator eq ' not ' ) {
1963 return NZoperatorNOT($leftresult,$rightresult);
1967 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1971 warn "string :" . $string if $DEBUG;
1975 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1978 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1980 warn "no parenthesis. left : $left operator: $operator right: $right"
1983 # it's not a leaf, we have a and/or/not
1986 # reintroduce comma content if needed
1987 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1988 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1989 warn "node : $left / $operator / $right\n" if $DEBUG;
1990 my $leftresult = NZanalyse( $left, $server );
1991 my $rightresult = NZanalyse( $right, $server );
1992 warn " leftresult : $leftresult" if $DEBUG;
1993 warn " rightresult : $rightresult" if $DEBUG;
1994 # OK, we have the results for right and left part of the query
1995 # depending of operand, intersect, union or exclude both lists
1996 # to get a result list
1997 if ( $operator eq ' and ' ) {
1998 return NZoperatorAND($leftresult,$rightresult);
2000 elsif ( $operator eq ' or ' ) {
2002 # just merge the 2 strings
2003 return $leftresult . $rightresult;
2005 elsif ( $operator eq ' not ' ) {
2006 return NZoperatorNOT($leftresult,$rightresult);
2010 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2011 die "error : operand unknown : $operator for $string";
2014 # it's a leaf, do the real SQL query and return the result
2017 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2018 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2019 #remove trailing blank at the beginning
2021 warn "leaf:$string" if $DEBUG;
2023 # parse the string in in operator/operand/value again
2027 if ($string =~ /(.*)(>=|<=)(.*)/) {
2034 # warn "handling leaf... left:$left operator:$operator right:$right"
2036 unless ($operator) {
2037 if ($string =~ /(.*)(>|<|=)(.*)/) {
2042 "handling unless (operator)... left:$left operator:$operator right:$right"
2050 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2053 # automatic replace for short operators
2054 $left = 'title' if $left =~ '^ti$';
2055 $left = 'author' if $left =~ '^au$';
2056 $left = 'publisher' if $left =~ '^pb$';
2057 $left = 'subject' if $left =~ '^su$';
2058 $left = 'koha-Auth-Number' if $left =~ '^an$';
2059 $left = 'keyword' if $left =~ '^kw$';
2060 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2061 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2062 my $dbh = C4::Context->dbh;
2063 if ( $operator && $left ne 'keyword' ) {
2064 #do a specific search
2065 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2066 my $sth = $dbh->prepare(
2067 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2069 warn "$left / $operator / $right\n" if $DEBUG;
2071 # split each word, query the DB and build the biblionumbers result
2072 #sanitizing leftpart
2073 $left =~ s/^\s+|\s+$//;
2074 foreach ( split / /, $right ) {
2076 $_ =~ s/^\s+|\s+$//;
2078 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2079 $sth->execute( $server, $left, $_ )
2080 or warn "execute failed: $!";
2081 while ( my ( $line, $value ) = $sth->fetchrow ) {
2083 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2084 # otherwise, fill the result
2085 $biblionumbers .= $line
2086 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2087 warn "result : $value "
2088 . ( $right =~ /\d/ ) . "=="
2089 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2092 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2094 warn "NZAND" if $DEBUG;
2095 $results = NZoperatorAND($biblionumbers,$results);
2097 $results = $biblionumbers;
2102 #do a complete search (all indexes), if index='kw' do complete search too.
2103 my $sth = $dbh->prepare(
2104 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2107 # split each word, query the DB and build the biblionumbers result
2108 foreach ( split / /, $string ) {
2109 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2110 warn "search on all indexes on $_" if $DEBUG;
2113 $sth->execute( $server, $_ );
2114 while ( my $line = $sth->fetchrow ) {
2115 $biblionumbers .= $line;
2118 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2120 $results = NZoperatorAND($biblionumbers,$results);
2123 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2124 $results = $biblionumbers;
2128 warn "return : $results for LEAF : $string" if $DEBUG;
2131 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2135 my ($rightresult, $leftresult)=@_;
2137 my @leftresult = split /;/, $leftresult;
2138 warn " @leftresult / $rightresult \n" if $DEBUG;
2140 # my @rightresult = split /;/,$leftresult;
2143 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2144 # the result is stored twice, to have the same weight for AND than OR.
2145 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2146 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2147 foreach (@leftresult) {
2150 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2151 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2152 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2154 "$value-$countvalue;$value-$countvalue;";
2157 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2158 return $finalresult;
2162 my ($rightresult, $leftresult)=@_;
2163 return $rightresult.$leftresult;
2167 my ($leftresult, $rightresult)=@_;
2169 my @leftresult = split /;/, $leftresult;
2171 # my @rightresult = split /;/,$leftresult;
2173 foreach (@leftresult) {
2175 $value=$1 if $value=~m/(.*)-\d+$/;
2176 unless ($rightresult =~ "$value-") {
2177 $finalresult .= "$_;";
2180 return $finalresult;
2185 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2192 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2193 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2195 # order title asc by default
2196 # $ordering = '1=36 <i' unless $ordering;
2197 $results_per_page = 20 unless $results_per_page;
2198 $offset = 0 unless $offset;
2199 my $dbh = C4::Context->dbh;
2202 # order by POPULARITY
2204 if ( $ordering =~ /popularity/ ) {
2208 # popularity is not in MARC record, it's builded from a specific query
2210 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2211 foreach ( split /;/, $biblionumbers ) {
2212 my ( $biblionumber, $title ) = split /,/, $_;
2213 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2214 $sth->execute($biblionumber);
2215 my $popularity = $sth->fetchrow || 0;
2217 # hint : the key is popularity.title because we can have
2218 # many results with the same popularity. In this case, sub-ordering is done by title
2219 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2220 # (un-frequent, I agree, but we won't forget anything that way ;-)
2221 $popularity{ sprintf( "%10d", $popularity ) . $title
2222 . $biblionumber } = $biblionumber;
2225 # sort the hash and return the same structure as GetRecords (Zebra querying)
2228 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2229 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2230 $result_hash->{'RECORDS'}[ $numbers++ ] =
2231 $result{ $popularity{$key} }->as_usmarc();
2234 else { # sort popularity ASC
2235 foreach my $key ( sort ( keys %popularity ) ) {
2236 $result_hash->{'RECORDS'}[ $numbers++ ] =
2237 $result{ $popularity{$key} }->as_usmarc();
2240 my $finalresult = ();
2241 $result_hash->{'hits'} = $numbers;
2242 $finalresult->{'biblioserver'} = $result_hash;
2243 return $finalresult;
2249 elsif ( $ordering =~ /author/ ) {
2251 foreach ( split /;/, $biblionumbers ) {
2252 my ( $biblionumber, $title ) = split /,/, $_;
2253 my $record = GetMarcBiblio($biblionumber);
2255 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2256 $author = $record->subfield( '200', 'f' );
2257 $author = $record->subfield( '700', 'a' ) unless $author;
2260 $author = $record->subfield( '100', 'a' );
2263 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2264 # and we don't want to get only 1 result for each of them !!!
2265 $result{ $author . $biblionumber } = $record;
2268 # sort the hash and return the same structure as GetRecords (Zebra querying)
2271 if ( $ordering eq 'author_za' ) { # sort by author desc
2272 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2273 $result_hash->{'RECORDS'}[ $numbers++ ] =
2274 $result{$key}->as_usmarc();
2277 else { # sort by author ASC
2278 foreach my $key ( sort ( keys %result ) ) {
2279 $result_hash->{'RECORDS'}[ $numbers++ ] =
2280 $result{$key}->as_usmarc();
2283 my $finalresult = ();
2284 $result_hash->{'hits'} = $numbers;
2285 $finalresult->{'biblioserver'} = $result_hash;
2286 return $finalresult;
2289 # ORDER BY callnumber
2292 elsif ( $ordering =~ /callnumber/ ) {
2294 foreach ( split /;/, $biblionumbers ) {
2295 my ( $biblionumber, $title ) = split /,/, $_;
2296 my $record = GetMarcBiblio($biblionumber);
2298 my $frameworkcode = GetFrameworkCode($biblionumber);
2299 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2300 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2301 unless $callnumber_tag;
2302 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2303 $callnumber = $record->subfield( '200', 'f' );
2305 $callnumber = $record->subfield( '100', 'a' );
2308 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2309 # and we don't want to get only 1 result for each of them !!!
2310 $result{ $callnumber . $biblionumber } = $record;
2313 # sort the hash and return the same structure as GetRecords (Zebra querying)
2316 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2317 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2318 $result_hash->{'RECORDS'}[ $numbers++ ] =
2319 $result{$key}->as_usmarc();
2322 else { # sort by title ASC
2323 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2324 $result_hash->{'RECORDS'}[ $numbers++ ] =
2325 $result{$key}->as_usmarc();
2328 my $finalresult = ();
2329 $result_hash->{'hits'} = $numbers;
2330 $finalresult->{'biblioserver'} = $result_hash;
2331 return $finalresult;
2333 elsif ( $ordering =~ /pubdate/ ) { #pub year
2335 foreach ( split /;/, $biblionumbers ) {
2336 my ( $biblionumber, $title ) = split /,/, $_;
2337 my $record = GetMarcBiblio($biblionumber);
2338 my ( $publicationyear_tag, $publicationyear_subfield ) =
2339 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2340 my $publicationyear =
2341 $record->subfield( $publicationyear_tag,
2342 $publicationyear_subfield );
2344 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2345 # and we don't want to get only 1 result for each of them !!!
2346 $result{ $publicationyear . $biblionumber } = $record;
2349 # sort the hash and return the same structure as GetRecords (Zebra querying)
2352 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2353 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2354 $result_hash->{'RECORDS'}[ $numbers++ ] =
2355 $result{$key}->as_usmarc();
2358 else { # sort by pub year ASC
2359 foreach my $key ( sort ( keys %result ) ) {
2360 $result_hash->{'RECORDS'}[ $numbers++ ] =
2361 $result{$key}->as_usmarc();
2364 my $finalresult = ();
2365 $result_hash->{'hits'} = $numbers;
2366 $finalresult->{'biblioserver'} = $result_hash;
2367 return $finalresult;
2373 elsif ( $ordering =~ /title/ ) {
2375 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2377 foreach ( split /;/, $biblionumbers ) {
2378 my ( $biblionumber, $title ) = split /,/, $_;
2380 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2381 # and we don't want to get only 1 result for each of them !!!
2382 # hint & speed improvement : we can order without reading the record
2383 # so order, and read records only for the requested page !
2384 $result{ $title . $biblionumber } = $biblionumber;
2387 # sort the hash and return the same structure as GetRecords (Zebra querying)
2390 if ( $ordering eq 'title_az' ) { # sort by title desc
2391 foreach my $key ( sort ( keys %result ) ) {
2392 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2395 else { # sort by title ASC
2396 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2397 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2401 # limit the $results_per_page to result size if it's more
2402 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2404 # for the requested page, replace biblionumber by the complete record
2405 # speed improvement : avoid reading too much things
2407 my $counter = $offset ;
2408 $counter <= $offset + $results_per_page ;
2412 $result_hash->{'RECORDS'}[$counter] =
2413 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2415 my $finalresult = ();
2416 $result_hash->{'hits'} = $numbers;
2417 $finalresult->{'biblioserver'} = $result_hash;
2418 return $finalresult;
2425 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2428 foreach ( split /;/, $biblionumbers ) {
2429 my ( $biblionumber, $title ) = split /,/, $_;
2430 $title =~ /(.*)-(\d)/;
2435 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2436 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2437 # biblio N has ranking = 6
2438 $count_ranking{$biblionumber} += $ranking;
2441 # build the result by "inverting" the count_ranking hash
2442 # 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
2444 foreach ( keys %count_ranking ) {
2445 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2448 # sort the hash and return the same structure as GetRecords (Zebra querying)
2451 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2452 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2455 # limit the $results_per_page to result size if it's more
2456 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2458 # for the requested page, replace biblionumber by the complete record
2459 # speed improvement : avoid reading too much things
2461 my $counter = $offset ;
2462 $counter <= $offset + $results_per_page ;
2466 $result_hash->{'RECORDS'}[$counter] =
2467 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2468 if $result_hash->{'RECORDS'}[$counter];
2470 my $finalresult = ();
2471 $result_hash->{'hits'} = $numbers;
2472 $finalresult->{'biblioserver'} = $result_hash;
2473 return $finalresult;
2477 =head2 enabled_staff_search_views
2479 %hash = enabled_staff_search_views()
2481 This function returns a hash that contains three flags obtained from the system
2482 preferences, used to determine whether a particular staff search results view
2487 =item C<Output arg:>
2489 * $hash{can_view_MARC} is true only if the MARC view is enabled
2490 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2491 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2493 =item C<usage in the script:>
2497 $template->param ( C4::Search::enabled_staff_search_views );
2501 sub enabled_staff_search_views
2504 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2505 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2506 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2510 sub AddSearchHistory{
2511 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2512 my $dbh = C4::Context->dbh;
2514 # Add the request the user just made
2515 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2516 my $sth = $dbh->prepare($sql);
2517 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2518 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2521 sub GetSearchHistory{
2522 my ($borrowernumber,$session)=@_;
2523 my $dbh = C4::Context->dbh;
2525 # Add the request the user just made
2526 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2527 my $sth = $dbh->prepare($query);
2528 $sth->execute($borrowernumber, $session);
2529 return $sth->fetchall_hashref({});
2532 =head2 z3950_search_args
2534 $arrayref = z3950_search_args($matchpoints)
2536 This function returns an array reference that contains the search parameters to be
2537 passed to the Z39.50 search script (z3950_search.pl). The array elements
2538 are hash refs whose keys are name, value and encvalue, and whose values are the
2539 name of a search parameter, the value of that search parameter and the URL encoded
2540 value of that parameter.
2542 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2544 The search parameter values are obtained from the bibliographic record whose
2545 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2547 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2548 a general purpose search argument. In this case, the returned array contains only
2549 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2551 If a search parameter value is undefined or empty, it is not included in the returned
2554 The returned array reference may be passed directly to the template parameters.
2558 =item C<Output arg:>
2560 * $array containing hash refs as described above
2562 =item C<usage in the script:>
2566 $data = Biblio::GetBiblioData($bibno);
2567 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2571 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2575 sub z3950_search_args {
2577 $bibrec = { title => $bibrec } if !ref $bibrec;
2579 for my $field (qw/ lccn isbn issn title author dewey subject /)
2581 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2582 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2587 =head2 BiblioAddAuthorities
2589 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2591 this function finds the authorities linked to the biblio
2592 * search in the authority DB for the same authid (in $9 of the biblio)
2593 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2594 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2595 OR adds a new authority record
2601 * $record is the MARC record in question (marc blob)
2602 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2604 =item C<Output arg:>
2606 * $countlinked is the number of authorities records that are linked to this authority
2610 * 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)
2617 sub BiblioAddAuthorities{
2618 my ( $record, $frameworkcode ) = @_;
2619 my $dbh=C4::Context->dbh;
2620 my $query=$dbh->prepare(qq|
2621 SELECT authtypecode,tagfield
2622 FROM marc_subfield_structure
2623 WHERE frameworkcode=?
2624 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2625 # SELECT authtypecode,tagfield
2626 # FROM marc_subfield_structure
2627 # WHERE frameworkcode=?
2628 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2629 $query->execute($frameworkcode);
2630 my ($countcreated,$countlinked);
2631 while (my $data=$query->fetchrow_hashref){
2632 foreach my $field ($record->field($data->{tagfield})){
2633 next if ($field->subfield('3')||$field->subfield('9'));
2634 # No authorities id in the tag.
2635 # Search if there is any authorities to link to.
2636 my $query='at='.$data->{authtypecode}.' ';
2637 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2638 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2639 # there is only 1 result
2641 warn "BIBLIOADDSAUTHORITIES: $error";
2644 if ($results && scalar(@$results)==1) {
2645 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2646 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2648 } elsif (scalar(@$results)>1) {
2649 #More than One result
2650 #This can comes out of a lack of a subfield.
2651 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2652 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2655 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2656 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2657 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2658 my $authtypedata=C4::AuthoritiesMarc::GetAuthType($data->{authtypecode});
2659 next unless $authtypedata;
2660 my $marcrecordauth=MARC::Record->new();
2661 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2662 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2663 $marcrecordauth->insert_fields_ordered($authfield);
2665 # bug 2317: ensure new authority knows it's using UTF-8; currently
2666 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2667 # automatically for UNIMARC (by not transcoding)
2668 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2669 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2670 # of change to a core API just before the 3.0 release.
2671 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2672 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2675 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2677 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2679 $field->add_subfields('9'=>$authid);
2683 return ($countlinked,$countcreated);
2686 =head2 GetDistinctValues($field);
2688 C<$field> is a reference to the fields array
2692 sub GetDistinctValues {
2693 my ($fieldname,$string)=@_;
2694 # returns a reference to a hash of references to branches...
2695 if ($fieldname=~/\./){
2696 my ($table,$column)=split /\./, $fieldname;
2697 my $dbh = C4::Context->dbh;
2698 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2699 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 ");
2701 my $elements=$sth->fetchall_arrayref({});
2706 my @servers=qw<biblioserver authorityserver>;
2707 my (@zconns,@results);
2708 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2709 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2712 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2715 # The big moment: asynchronously retrieve results from all servers
2717 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2718 my $ev = $zconns[ $i - 1 ]->last_event();
2719 if ( $ev == ZOOM::Event::ZEND ) {
2720 next unless $results[ $i - 1 ];
2721 my $size = $results[ $i - 1 ]->size();
2723 for (my $j=0;$j<$size;$j++){
2725 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2726 push @elements, \%hashscan;
2736 END { } # module clean-up code here (global destructor)
2743 Koha Development Team <http://koha-community.org/>