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
21 use C4::Biblio; # GetMarcFromKohaField
22 use C4::Koha; # getFacets
24 use C4::Dates qw(format_date);
26 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
28 # set the version for version checking
31 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
36 C4::Search - Functions for searching the Koha catalog.
40 see opac/opac-search.pl or catalogue/search.pl for example of usage
44 This module provides the searching facilities for the Koha into a zebra catalog.
62 # make all your functions, whether exported or not;
64 =head2 findseealso($dbh,$fields);
66 C<$dbh> is a link to the DB handler.
69 my $dbh =C4::Context->dbh;
71 C<$fields> is a reference to the fields array
73 This function modify the @$fields array and add related fields to search on.
78 my ( $dbh, $fields ) = @_;
79 my $tagslib = GetMarcStructure( 1 );
80 for ( my $i = 0 ; $i <= $#{$fields} ; $i++ ) {
81 my ($tag) = substr( @$fields[$i], 1, 3 );
82 my ($subfield) = substr( @$fields[$i], 4, 1 );
83 @$fields[$i] .= ',' . $tagslib->{$tag}->{$subfield}->{seealso}
84 if ( $tagslib->{$tag}->{$subfield}->{seealso} );
90 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
96 my $dbh = C4::Context->dbh;
97 my $result = TransformMarcToKoha( $dbh, $record, '' );
102 my ( $biblionumber, $title );
104 # search duplicate on ISBN, easy and fast..
105 # ... normalize first
106 if ( $result->{isbn} ) {
107 $result->{isbn} =~ s/\(.*$//;
108 $result->{isbn} =~ s/\s+$//;
110 #$search->{'avoidquerylog'}=1;
111 if ( $result->{isbn} ) {
112 $query = "isbn=$result->{isbn}";
115 $result->{title} =~ s /\\//g;
116 $result->{title} =~ s /\"//g;
117 $result->{title} =~ s /\(//g;
118 $result->{title} =~ s /\)//g;
119 # remove valid operators
120 $result->{title} =~ s/(and|or|not)//g;
121 $query = "ti,ext=$result->{title}";
122 $query .= " and itemtype=$result->{itemtype}" if ($result->{itemtype});
123 if ($result->{author}){
124 $result->{author} =~ s /\\//g;
125 $result->{author} =~ s /\"//g;
126 $result->{author} =~ s /\(//g;
127 $result->{author} =~ s /\)//g;
128 # remove valid operators
129 $result->{author} =~ s/(and|or|not)//g;
130 $query .= " and au,ext=$result->{author}";
133 my ($error,$searchresults) =
134 SimpleSearch($query); # FIXME :: hardcoded !
136 foreach my $possible_duplicate_record (@$searchresults) {
138 MARC::Record->new_from_usmarc($possible_duplicate_record);
139 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
141 # FIXME :: why 2 $biblionumber ?
143 push @results, $result->{'biblionumber'};
144 push @results, $result->{'title'};
152 ($error,$results) = SimpleSearch($query,@servers);
154 this function performs a simple search on the catalog using zoom.
160 * $query could be a simple keyword or a complete CCL query wich is depending on your ccl file.
161 * @servers is optionnal. default one is read on koha-conf.xml
164 * $error is a string which containt the description error if there is one. Else it's empty.
165 * \@results is an array of marc record.
167 =item C<usage in the script:>
171 my ($error, $marcresults) = SimpleSearch($query);
173 if (defined $error) {
174 $template->param(query_error => $error);
175 warn "error: ".$error;
176 output_html_with_http_headers $input, $cookie, $template->output;
180 my $hits = scalar @$marcresults;
183 for(my $i=0;$i<$hits;$i++) {
185 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
186 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
188 #build the hash for the template.
189 $resultsloop{highlight} = ($i % 2)?(1):(0);
190 $resultsloop{title} = $biblio->{'title'};
191 $resultsloop{subtitle} = $biblio->{'subtitle'};
192 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
193 $resultsloop{author} = $biblio->{'author'};
194 $resultsloop{publishercode} = $biblio->{'publishercode'};
195 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
197 push @results, \%resultsloop;
199 $template->param(result=>\@results);
205 if (C4::Context->preference('NoZebra')) {
206 my $result = NZorder(NZanalyse($query))->{'biblioserver'};
207 my $search_result = ( $result->{hits} && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
208 return (undef,$search_result);
214 return ( "No query entered", undef ) unless $query;
216 #@servers = (C4::Context->config("biblioserver")) unless @servers;
218 ("biblioserver") unless @servers
219 ; # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
222 for ( my $i = 0 ; $i < @servers ; $i++ ) {
224 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
227 ->search( new ZOOM::Query::CCL2RPN( $query, $zconns[$i] ) );
229 # getting error message if one occured.
231 $zconns[$i]->errmsg() . " ("
232 . $zconns[$i]->errcode() . ") "
233 . $zconns[$i]->addinfo() . " "
234 . $zconns[$i]->diagset();
236 return ( $error, undef ) if $zconns[$i]->errcode();
239 # caught a ZOOM::Exception
243 . $@->addinfo() . " "
246 return ( $error, undef );
251 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
252 $ev = $zconns[ $i - 1 ]->last_event();
253 if ( $ev == ZOOM::Event::ZEND ) {
254 $hits = $tmpresults[ $i - 1 ]->size();
257 for ( my $j = 0 ; $j < $hits ; $j++ ) {
258 my $record = $tmpresults[ $i - 1 ]->record($j)->raw();
259 push @results, $record;
263 return ( undef, \@results );
267 # performs the search
270 $koha_query, $simple_query, $sort_by_ref,
271 $servers_ref, $results_per_page, $offset,
272 $expanded_facet, $branches, $query_type,
275 # warn "Query : $koha_query";
276 my @servers = @$servers_ref;
277 my @sort_by = @$sort_by_ref;
279 # create the zoom connection and query object
283 my $results_hashref = ();
286 my $facets_counter = ();
287 my $facets_info = ();
288 my $facets = getFacets();
290 #### INITIALIZE SOME VARS USED CREATE THE FACETED RESULTS
291 my @facets_loop; # stores the ref to array of hashes for template
292 for ( my $i = 0 ; $i < @servers ; $i++ ) {
293 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
295 # perform the search, create the results objects
296 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
298 if ( $servers[$i] =~ /biblioserver/ ) {
299 $query_to_use = $koha_query;
302 $query_to_use = $simple_query;
305 #$query_to_use = $simple_query if $scan;
306 #warn $simple_query if ($scan && $DEBUG);
307 # check if we've got a query_type defined
311 if ( $query_type =~ /^ccl/ ) {
313 s/\:/\=/g; # change : to = last minute (FIXME)
315 # warn "CCL : $query_to_use";
318 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
321 elsif ( $query_type =~ /^cql/ ) {
323 # warn "CQL : $query_to_use";
326 new ZOOM::Query::CQL( $query_to_use, $zconns[$i] ) );
328 elsif ( $query_type =~ /^pqf/ ) {
330 # warn "PQF : $query_to_use";
333 new ZOOM::Query::PQF( $query_to_use, $zconns[$i] ) );
338 # warn "preparing to scan:$query_to_use";
341 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
345 # warn "LAST : $query_to_use";
348 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
354 warn "WARNING: query problem with $query_to_use " . $@;
357 # concatenate the sort_by limits and pass them to the results object
359 foreach my $sort (@sort_by) {
360 if ($sort eq "author_az") {
361 $sort_by.="1=1003 <i ";
363 elsif ($sort eq "author_za") {
364 $sort_by.="1=1003 >i ";
366 elsif ($sort eq "popularity_asc") {
367 $sort_by.="1=9003 <i ";
369 elsif ($sort eq "popularity_dsc") {
370 $sort_by.="1=9003 >i ";
372 elsif ($sort eq "call_number_asc") {
373 $sort_by.="1=20 <i ";
375 elsif ($sort eq "call_number_dsc") {
376 $sort_by.="1=20 >i ";
378 elsif ($sort eq "pubdate_asc") {
379 $sort_by.="1=31 <i ";
381 elsif ($sort eq "pubdate_dsc") {
382 $sort_by.="1=31 >i ";
384 elsif ($sort eq "acqdate_asc") {
385 $sort_by.="1=32 <i ";
387 elsif ($sort eq "acqdate_dsc") {
388 $sort_by.="1=32 >i ";
390 elsif ($sort eq "title_az") {
393 elsif ($sort eq "title_za") {
398 if ( $results[$i]->sort( "yaz", $sort_by ) < 0) {
399 warn "WARNING sort $sort_by failed";
403 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
404 my $ev = $zconns[ $i - 1 ]->last_event();
405 if ( $ev == ZOOM::Event::ZEND ) {
406 my $size = $results[ $i - 1 ]->size();
409 #$results_hash->{'server'} = $servers[$i-1];
410 # loop through the results
411 $results_hash->{'hits'} = $size;
413 if ( $offset + $results_per_page <= $size ) {
414 $times = $offset + $results_per_page;
419 for ( my $j = $offset ; $j < $times ; $j++ )
420 { #(($offset+$count<=$size) ? ($offset+$count):$size) ; $j++){
424 ## This is just an index scan
426 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');
434 # the minimal record in author/title (depending on MARC flavour)
435 if ( C4::Context->preference("marcflavour") eq
438 $tmptitle = MARC::Field->new(
445 $tmptitle = MARC::Field->new('245', ' ', ' ',a => $term,);
446 $tmpauthor = MARC::Field->new('100', ' ', ' ',a => $occ,);
448 $tmprecord->append_fields($tmptitle);
449 $tmprecord->append_fields($tmpauthor);
450 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
453 $record = $results[ $i - 1 ]->record($j)->raw();
455 #warn "RECORD $j:".$record;
456 $results_hash->{'RECORDS'}[$j] =
457 $record; # making a reference to a hash
458 # Fill the facets while we're looping
459 $facet_record = MARC::Record->new_from_usmarc($record);
461 #warn $servers[$i-1].$facet_record->title();
462 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
463 if ( $facets->[$k] ) {
465 for my $tag ( @{ $facets->[$k]->{'tags'} } ) {
466 push @fields, $facet_record->field($tag);
468 for my $field (@fields) {
469 my @subfields = $field->subfields();
470 for my $subfield (@subfields) {
471 my ( $code, $data ) = @$subfield;
473 $facets->[$k]->{'subfield'} )
475 $facets_counter->{ $facets->[$k]
476 ->{'link_value'} }->{$data}++;
480 $facets_info->{ $facets->[$k]->{'link_value'} }
482 $facets->[$k]->{'label_value'};
483 $facets_info->{ $facets->[$k]->{'link_value'} }
484 ->{'expanded'} = $facets->[$k]->{'expanded'};
489 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
492 #print "connection ", $i-1, ": $size hits";
493 #print $results[$i-1]->record(0)->render() if $size > 0;
496 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
497 keys %$facets_counter
501 my $number_of_facets;
502 my @this_facets_array;
505 $facets_counter->{$link_value}
506 ->{$b} <=> $facets_counter->{$link_value}->{$a}
507 } keys %{ $facets_counter->{$link_value} }
511 if ( ( $number_of_facets < 6 )
512 || ( $expanded_facet eq $link_value )
513 || ( $facets_info->{$link_value}->{'expanded'} ) )
516 # sanitize the link value ), ( will cause errors with CCL
517 my $facet_link_value = $one_facet;
518 $facet_link_value =~ s/(\(|\))/ /g;
520 # fix the length that will display in the label
521 my $facet_label_value = $one_facet;
522 $facet_label_value = substr( $one_facet, 0, 20 ) . "..."
523 unless length($facet_label_value) <= 20;
525 # well, if it's a branch, label by the name, not the code
526 if ( $link_value =~ /branch/ ) {
528 $branches->{$one_facet}->{'branchname'};
531 # but we're down with the whole label being in the link's title
532 my $facet_title_value = $one_facet;
534 push @this_facets_array,
538 $facets_counter->{$link_value}->{$one_facet},
539 facet_label_value => $facet_label_value,
540 facet_title_value => $facet_title_value,
541 facet_link_value => $facet_link_value,
542 type_link_value => $link_value,
547 unless ( $facets_info->{$link_value}->{'expanded'} ) {
549 if ( ( $number_of_facets > 6 )
550 && ( $expanded_facet ne $link_value ) );
555 type_link_value => $link_value,
556 type_id => $link_value . "_id",
558 $facets_info->{$link_value}->{'label_value'},
559 facets => \@this_facets_array,
560 expandable => $expandable,
561 expand => $link_value,
567 return ( undef, $results_hashref, \@facets_loop );
571 sub _remove_stopwords {
572 my ($operand,$index) = @_;
573 my @stopwords_removed;
574 # phrase and exact-qualified indexes shouldn't have stopwords removed
575 if ($index!~m/phr|ext/){
576 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
577 # we use IsAlpha unicode definition, to deal correctly with diacritics.
578 # otherwise, a French word like "leçon" woudl be split into "le" "çon", le
579 # is an empty word, we'd get "çon" and wouldn't find anything...
580 foreach (keys %{C4::Context->stopwords}) {
581 next if ($_ =~/(and|or|not)/); # don't remove operators
582 if ($operand =~ /(\P{IsAlpha}$_\P{IsAlpha}|^$_\P{IsAlpha}|\P{IsAlpha}$_$)/) {
583 $operand=~ s/\P{IsAlpha}$_\P{IsAlpha}/ /gi;
584 $operand=~ s/^$_\P{IsAlpha}/ /gi;
585 $operand=~ s/\P{IsAlpha}$_$/ /gi;
586 push @stopwords_removed, $_;
590 return ($operand, \@stopwords_removed);
594 sub _detect_truncation {
595 my ($operand,$index) = @_;
596 my (@nontruncated,@righttruncated,@lefttruncated,@rightlefttruncated,@regexpr);
598 my @wordlist= split (/\s/,$operand);
599 foreach my $word (@wordlist){
600 if ($word=~s/^\*([^\*]+)\*$/$1/){
601 push @rightlefttruncated,$word;
603 elsif($word=~s/^\*([^\*]+)$/$1/){
604 push @lefttruncated,$word;
606 elsif ($word=~s/^([^\*]+)\*$/$1/){
607 push @righttruncated,$word;
609 elsif (index($word,"*")<0){
610 push @nontruncated,$word;
616 return (\@nontruncated,\@righttruncated,\@lefttruncated,\@rightlefttruncated,\@regexpr);
619 sub _build_stemmed_operand {
622 # FIXME: the locale should be set based on the user's language and/or search choice
623 my $stemmer = Lingua::Stem->new( -locale => 'EN-US' );
624 # FIXME: these should be stored in the db so the librarian can modify the behavior
625 $stemmer->add_exceptions(
633 my @words = split( / /, $operand );
634 my $stems = $stemmer->stem(@words);
635 for my $stem (@$stems) {
636 $stemmed_operand .= "$stem";
637 $stemmed_operand .= "?" unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
638 $stemmed_operand .= " ";
640 #warn "STEMMED OPERAND: $stemmed_operand";
641 return $stemmed_operand;
644 sub _build_weighted_query {
645 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
646 # pretty well but will work much better when we have an actual query parser
647 my ($operand,$stemmed_operand,$index) = @_;
648 my $stemming = C4::Context->preference("QueryStemming") || 0;
649 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
650 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
652 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
654 # Keyword, or, no index specified
655 if ( ( $index eq 'kw' ) || ( !$index ) ) {
656 $weighted_query .= "Title-cover,ext,r1=\"$operand\""; # exact title-cover
657 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
658 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
659 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
660 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
661 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\"" if $fuzzy_enabled; # add fuzzy, word list
662 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\"" if ($stemming and $stemmed_operand); # add stemming, right truncation
663 $weighted_query .= " or wrdl,r9=\"$operand\"";
665 # embedded sorting: 0 a-z; 1 z-a
666 # $weighted_query .= ") or (sort1,aut=1";
668 elsif ( $index eq 'bc' ) {
669 $weighted_query .= "bc=\"$operand\"";
671 # if the index already has more than one qualifier, just wrap the operand
672 # in quotes and pass it back
673 elsif ($index =~ ',') {
674 $weighted_query .=" $index=\"$operand\"";
676 #TODO: build better cases based on specific search indexes
678 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
679 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
680 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
681 $weighted_query .= " or $index,rt,wrdl,r3=\"$operand\""; # word list index
683 $weighted_query .= "))"; # close rank specification
684 return $weighted_query;
687 # build the query itself
689 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan) = @_;
691 warn "---------" if $DEBUG;
692 warn "Enter buildQuery" if $DEBUG;
693 warn "---------" if $DEBUG;
695 my @operators = @$operators if $operators;
696 my @indexes = @$indexes if $indexes;
697 my @operands = @$operands if $operands;
698 my @limits = @$limits if $limits;
699 my @sort_by = @$sort_by if $sort_by;
701 my $stemming = C4::Context->preference("QueryStemming") || 0;
702 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
703 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
704 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
705 # no stemming/weight/fuzzy in NoZebra
706 if (C4::Context->preference("NoZebra")) {
711 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
713 my $query = $operands[0];
714 my $simple_query = $operands[0];
723 my $stopwords_removed;
725 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
727 if ( $query =~ /^ccl=/ ) {
728 return ( undef, $', $', $', $', '', '', '', '', 'ccl' );
730 if ( $query =~ /^cql=/ ) {
731 return ( undef, $', $', $', $', '', '', '', '', 'cql' );
733 if ( $query =~ /^pqf=/ ) {
734 return ( undef, $', $', $', $', '', '', '', '', 'pqf' );
737 # pass nested queries directly
738 if ( $query =~ /(\(|\))/ ) {
739 return ( undef, $query, $simple_query, $query_cgi, $query, $limit, $limit_cgi, $limit_desc, $stopwords_removed, 'ccl' );
742 # form-based queries are limited to non-nested at a specific depth, so we can easily
743 # modify the incoming query operands and indexes to do stemming and field weighting
744 # Once we do so, we'll end up with a value in $query, just like if we had an
745 # incoming $query from the user
747 $query = ""; # clear it out so we can populate properly with field-weighted stemmed query
748 my $previous_operand; # a flag used to keep track if there was a previous query
749 # if there was, we can apply the current operator
751 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
753 # COMBINE OPERANDS, INDEXES AND OPERATORS
754 if ( $operands[$i] ) {
756 # a flag to determine whether or not to add the index to the query
759 # if the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
760 if ($operands[$i] =~ /(:|=)/ || $scan) {
763 $remove_stopwords = 0;
765 my $operand = $operands[$i];
766 my $index = $indexes[$i];
768 # add some attributes for certain index types
769 # Date of Publication
770 if ($index eq 'yr') {
771 $index .=",st-numeric";
773 ($stemming,$auto_truncation,$weight_fields, $fuzzy_enabled, $remove_stopwords) = (0,0,0,0,0);
775 # Date of Acquisition
776 elsif ($index eq 'acqdate') {
777 $index.=",st-date-normalized";
779 ($stemming,$auto_truncation,$weight_fields, $fuzzy_enabled, $remove_stopwords) = (0,0,0,0,0);
783 # set default structure attribute (word list)
785 unless (!$index || $index =~ /(st-|phr|ext|wrdl)/) {
786 $struct_attr = ",wrdl";
788 # some helpful index modifs
789 my $index_plus = $index.$struct_attr.":" if $index;
790 my $index_plus_comma=$index.$struct_attr."," if $index;
793 if ($remove_stopwords) {
794 ($operand, $stopwords_removed) = _remove_stopwords($operand,$index);
795 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
796 warn "REMOVED STOPWORDS: @$stopwords_removed" if ($stopwords_removed && $DEBUG);
800 my ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr);
801 my $truncated_operand;
802 ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr) = _detect_truncation($operand,$index);
803 warn "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<" if $DEBUG;
806 if (scalar(@$righttruncated)+scalar(@$lefttruncated)+scalar(@$rightlefttruncated)>0){
807 # don't field weight or add the index to the query, we do it here
809 undef $weight_fields;
810 my $previous_truncation_operand;
811 if (scalar(@$nontruncated)>0) {
812 $truncated_operand.= "$index_plus @$nontruncated ";
813 $previous_truncation_operand = 1;
815 if (scalar(@$righttruncated)>0){
816 $truncated_operand .= "and " if $previous_truncation_operand;
817 $truncated_operand .= "$index_plus_comma"."rtrn:@$righttruncated ";
818 $previous_truncation_operand = 1;
820 if (scalar(@$lefttruncated)>0){
821 $truncated_operand .= "and " if $previous_truncation_operand;
822 $truncated_operand .= "$index_plus_comma"."ltrn:@$lefttruncated ";
823 $previous_truncation_operand = 1;
825 if (scalar(@$rightlefttruncated)>0){
826 $truncated_operand .= "and " if $previous_truncation_operand;
827 $truncated_operand .= "$index_plus_comma"."rltrn:@$rightlefttruncated ";
828 $previous_truncation_operand = 1;
831 $operand = $truncated_operand if $truncated_operand;
832 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
836 $stemmed_operand = _build_stemmed_operand($operand) if $stemming;
837 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
839 # Handle Field Weighting
840 my $weighted_operand;
841 $weighted_operand = _build_weighted_query($operand,$stemmed_operand,$index) if $weight_fields;
842 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
843 $operand = $weighted_operand if $weight_fields;
844 $indexes_set = 1 if $weight_fields;
846 # If there's a previous operand, we need to add an operator
847 if ($previous_operand) {
849 # user-specified operator
850 if ( $operators[$i-1] ) {
851 $query .= " $operators[$i-1] ";
852 $query .= " $index_plus " unless $indexes_set;
853 $query .= " $operand";
854 $query_cgi .="&op=$operators[$i-1]";
855 $query_cgi .="&idx=$index" if $index;
856 $query_cgi .="&q=$operands[$i]" if $operands[$i];
857 $query_desc .=" $operators[$i-1] $index_plus $operands[$i]";
860 # the default operator is and
863 $query .= "$index_plus " unless $indexes_set;
864 $query .= "$operand";
865 $query_cgi .="&op=and&idx=$index" if $index;
866 $query_cgi .="&q=$operands[$i]" if $operands[$i];
867 $query_desc .= " and $index_plus $operands[$i]";
871 # there isn't a pervious operand, don't need an operator
873 # field-weighted queries already have indexes set
874 $query .=" $index_plus " unless $indexes_set;
876 $query_desc .= " $index_plus $operands[$i]";
877 $query_cgi.="&idx=$index" if $index;
878 $query_cgi.="&q=$operands[$i]" if $operands[$i];
880 $previous_operand = 1;
885 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
890 my $availability_limit;
891 foreach my $this_limit (@limits) {
892 if ( $this_limit =~ /available/ ) {
893 # available is defined as (items.notloan is NULL) and (items.itemlost > 0 or NULL) (last clause handles NULL values for lost in zebra)
894 # all records not indexed in the onloan register and allrecords not indexed in the lost register, or where the value of lost is equal to or less than 0
895 $availability_limit .="( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and ((lost,st-numeric <= 0) or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
896 $limit_cgi .= "&limit=available";
900 # these are treated as OR
901 elsif ( $this_limit =~ /mc/ ) {
902 $group_OR_limits .= " or " if $group_OR_limits;
903 $limit_desc .=" or " if $group_OR_limits;
904 $group_OR_limits .= "$this_limit";
905 $limit_cgi .="&limit=$this_limit";
906 $limit_desc .= " $this_limit";
910 $limit .= " and " if $limit || $query;
911 $limit .= "$this_limit";
912 $limit_cgi .="&limit=$this_limit";
913 $limit_desc .=" $this_limit";
916 if ($group_OR_limits) {
917 $limit.=" and " if ($query || $limit );
918 $limit.="($group_OR_limits)";
920 if ($availability_limit) {
921 $limit.=" and " if ($query || $limit );
922 $limit.="($availability_limit)";
924 # normalize the strings
927 for ($query, $query_desc, $limit, $limit_desc) {
928 $_ =~ s/ / /g; # remove extra spaces
929 $_ =~ s/^ //g; # remove any beginning spaces
930 $_ =~ s/ $//g; # remove any ending spaces
931 $_ =~ s/==/=/g; # remove double == from query
934 $query_cgi =~ s/^&//;
936 # append the limit to the query
937 $query ? $query .= " ".$limit : $query = $limit;
938 warn "query=$query and limit=$limit" if $DEBUG;
940 warn "QUERY:".$query if $DEBUG;
941 warn "QUERY CGI:".$query_cgi if $DEBUG;
942 warn "QUERY DESC:".$query_desc if $DEBUG;
943 warn "LIMIT:".$limit if $DEBUG;
944 warn "LIMIT CGI:".$limit_cgi if $DEBUG;
945 warn "LIMIT DESC:".$limit_desc if $DEBUG;
946 warn "---------" if $DEBUG;
947 warn "Leave buildQuery" if $DEBUG;
948 warn "---------" if $DEBUG;
949 return ( undef, $query,$simple_query,$query_cgi,$query_desc,$limit,$limit_cgi,$limit_desc,$stopwords_removed,$query_type );
952 # IMO this subroutine is pretty messy still -- it's responsible for
953 # building the HTML output for the template
955 my ( $searchdesc, $hits, $results_per_page, $offset, @marcresults ) = @_;
957 my $dbh = C4::Context->dbh;
961 my $span_terms_hashref;
962 for my $span_term ( split( / /, $searchdesc ) ) {
963 $span_term =~ s/(.*=|\)|\(|\+|\.)//g;
964 $span_terms_hashref->{$span_term}++;
967 #Build brancnames hash
969 #get branch information.....
972 $dbh->prepare("SELECT branchcode,branchname FROM branches")
973 ; # FIXME : use C4::Koha::GetBranches
975 while ( my $bdata = $bsth->fetchrow_hashref ) {
976 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
979 my $lsch = $dbh->prepare("SELECT authorised_value,lib FROM authorised_values WHERE category = 'SHELF_LOC'");
981 while (my $ldata = $lsch->fetchrow_hashref ) {
982 $locations{ $ldata->{'authorised_value'} } = $ldata->{'lib'};
986 #find itemtype & itemtype image
989 $dbh->prepare("SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes");
991 while ( my $bdata = $bsth->fetchrow_hashref ) {
992 $itemtypes{ $bdata->{'itemtype'} }->{description} =
993 $bdata->{'description'};
994 $itemtypes{ $bdata->{'itemtype'} }->{imageurl} = $bdata->{'imageurl'};
995 $itemtypes{ $bdata->{'itemtype'} }->{summary} = $bdata->{'summary'};
996 $itemtypes{ $bdata->{'itemtype'} }->{notforloan} = $bdata->{'notforloan'};
999 #search item field code
1002 "select tagfield from marc_subfield_structure where kohafield like 'items.itemnumber'"
1005 my ($itemtag) = $sth->fetchrow;
1007 ## find column names of items related to MARC
1008 my $sth2 = $dbh->prepare("SHOW COLUMNS from items");
1010 my %subfieldstosearch;
1011 while ( ( my $column ) = $sth2->fetchrow ) {
1012 my ( $tagfield, $tagsubfield ) =
1013 &GetMarcFromKohaField( "items." . $column, "" );
1014 $subfieldstosearch{$column} = $tagsubfield;
1018 if ( $hits && $offset + $results_per_page <= $hits ) {
1019 $times = $offset + $results_per_page;
1025 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1027 $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1028 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, '' );
1029 $oldbiblio->{result_number} = $i+1;
1030 # add image url if there is one
1031 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} =~ /^http:/ ) {
1032 $oldbiblio->{imageurl} =
1033 $itemtypes{ $oldbiblio->{itemtype} }->{imageurl};
1034 $oldbiblio->{description} =
1035 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1038 $oldbiblio->{imageurl} =
1039 getitemtypeimagesrc() . "/"
1040 . $itemtypes{ $oldbiblio->{itemtype} }->{imageurl}
1041 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1042 $oldbiblio->{description} =
1043 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1046 # build summary if there is one (the summary is defined in itemtypes table
1048 if ($itemtypes{ $oldbiblio->{itemtype} }->{summary}) {
1049 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1050 my @fields = $marcrecord->fields();
1051 foreach my $field (@fields) {
1052 my $tag = $field->tag();
1053 my $tagvalue = $field->as_string();
1054 $summary =~ s/\[(.?.?.?.?)$tag\*(.*?)]/$1$tagvalue$2\[$1$tag$2]/g;
1056 my @subf = $field->subfields;
1057 for my $i (0..$#subf) {
1058 my $subfieldcode = $subf[$i][0];
1059 my $subfieldvalue = $subf[$i][1];
1060 my $tagsubf = $tag.$subfieldcode;
1061 $summary =~ s/\[(.?.?.?.?)$tagsubf(.*?)]/$1$subfieldvalue$2\[$1$tagsubf$2]/g;
1065 $summary =~ s/\[(.*?)]//g;
1066 $summary =~ s/\n/<br>/g;
1067 $oldbiblio->{summary} = $summary;
1069 # add spans to search term in results for search term highlighting
1070 # save a native author, for the <a href=search.lq=<!--tmpl_var name="author"-->> link
1071 my $searchhighlightblob;
1072 for my $highlight_field ($marcrecord->fields) {
1073 next if $highlight_field->tag() =~ /(^00)/; # skip fixed fields
1075 my $field = $highlight_field->as_string();
1076 for my $term ( keys %$span_terms_hashref ) {
1077 if (($field =~ /$term/i) && (length($term) > 3)) {
1078 $field =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1082 $searchhighlightblob .= $field." ... " if $match;
1084 $oldbiblio->{'searchhighlightblob'} = $searchhighlightblob;
1086 $oldbiblio->{'author_nospan'} = $oldbiblio->{'author'};
1087 for my $term ( keys %$span_terms_hashref ) {
1088 my $old_term = $term;
1089 if ( length($term) > 3 ) {
1090 $term =~ s/(.*=|\)|\(|\+|\.|\?|\[|\]|\\|\*)//g;
1091 $oldbiblio->{'title'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1092 $oldbiblio->{'subtitle'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1093 $oldbiblio->{'author'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1094 $oldbiblio->{'publishercode'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1095 $oldbiblio->{'place'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1096 $oldbiblio->{'pages'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1097 $oldbiblio->{'notes'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1098 $oldbiblio->{'size'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1103 $toggle = "#ffffcc";
1108 $oldbiblio->{'toggle'} = $toggle;
1109 my @fields = $marcrecord->field($itemtag);
1111 # Setting item statuses for display
1112 my @available_items_loop;
1113 my @onloan_items_loop;
1114 my @other_items_loop;
1116 my $available_items;
1120 my $ordered_count = 0;
1121 my $available_count = 0;
1122 my $onloan_count = 0;
1123 my $longoverdue_count = 0;
1124 my $other_count = 0;
1125 my $wthdrawn_count = 0;
1126 my $itemlost_count = 0;
1127 my $itembinding_count = 0;
1128 my $itemdamaged_count = 0;
1129 my $can_place_holds = 0;
1130 my $items_count=scalar(@fields);
1132 my $maxitems = (C4::Context->preference('maxItemsinSearchResults')) ? C4::Context->preference('maxItemsinSearchResults')- 1 : 1;
1133 foreach my $field (@fields) {
1137 # populate the items hash
1138 foreach my $code ( keys %subfieldstosearch ) {
1139 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1142 # set item's branch name, use homebranch first, fall back to holdingbranch
1143 if ($item->{'homebranch'}) {
1144 $item->{'branchname'} = $branches{$item->{homebranch}};
1147 elsif ($item->{'holdingbranch'}) {
1148 $item->{'branchname'} = $branches{$item->{holdingbranch}};
1150 # key for items results is built from branchcode . coded location qualifier . itemcallnumber
1151 if ($item->{onloan}) {
1153 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{due_date} = format_date($item->{onloan});
1154 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{count}++ if $item->{'homebranch'};
1155 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{branchname} = $item->{'branchname'};
1156 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{location} = $locations{$item->{location}};
1157 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{itemcallnumber} = $item->{itemcallnumber};
1159 # if something's checked out and lost, mark it as 'long overdue'
1160 if ( $item->{itemlost} ) {
1161 $onloan_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{due_date} }->{longoverdue}++;
1162 $longoverdue_count++;
1164 # can place holds as long as this item isn't lost
1166 $can_place_holds = 1;
1170 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1173 if ( $item->{notforloan} == -1) {
1177 # item is withdrawn, lost or damaged
1178 if ( $item->{wthdrawn} || $item->{itemlost} || $item->{damaged} ) {
1179 $wthdrawn_count++ if $item->{wthdrawn};
1180 $itemlost_count++ if $item->{itemlost};
1181 $itemdamaged_count++ if $item->{damaged};
1182 $item->{status} = $item->{wthdrawn}."-".$item->{itemlost}."-".$item->{damaged};
1184 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{wthdrawn} = $item->{wthdrawn};
1185 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{itemlost} = $item->{itemlost};
1186 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{damaged} = $item->{damaged};
1187 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{count}++ if $item->{'homebranch'};
1188 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{branchname} = $item->{'branchname'};
1189 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{location} = $locations{$item->{location}};
1190 $other_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'}.$item->{status} }->{itemcallnumber} = $item->{itemcallnumber};
1195 $can_place_holds = 1;
1197 $available_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'} }->{count}++ if $item->{'homebranch'};
1198 $available_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'} }->{branchname} = $item->{'branchname'};
1199 $available_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'} }->{location} = $locations{$item->{location}};
1200 $available_items->{ $item->{'homebranch'}.'--'.$item->{location}.$item->{'itemcallnumber'} }->{itemcallnumber} = $item->{itemcallnumber};
1203 } # notforloan, item level and biblioitem level
1204 my ($availableitemscount, $onloanitemscount, $otheritemscount);
1205 my $maxitems = (C4::Context->preference('maxItemsinSearchResults')) ? C4::Context->preference('maxItemsinSearchResults')- 1 : 1;
1206 for my $key ( sort keys %$onloan_items ) {
1207 $onloanitemscount++;
1208 push @onloan_items_loop, $onloan_items->{$key} unless $onloanitemscount > $maxitems;
1210 for my $key ( sort keys %$other_items ) {
1212 push @other_items_loop, $other_items->{$key} unless $otheritemscount > $maxitems;
1214 for my $key ( sort keys %$available_items ) {
1215 $availableitemscount++;
1216 push @available_items_loop, $available_items->{$key} unless $availableitemscount > $maxitems;
1218 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1219 $can_place_holds = 0 if $itemtypes{$oldbiblio->{itemtype}}->{notforloan};
1220 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1221 $oldbiblio->{itemsplural} = 1 if $items_count>1;
1222 $oldbiblio->{items_count} = $items_count;
1223 $oldbiblio->{available_items_loop} = \@available_items_loop;
1224 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1225 $oldbiblio->{other_items_loop} = \@other_items_loop;
1226 $oldbiblio->{availablecount} = $available_count;
1227 $oldbiblio->{availableplural} = 1 if $available_count>1;
1228 $oldbiblio->{onloancount} = $onloan_count;
1229 $oldbiblio->{onloanplural} = 1 if $onloan_count>1;
1230 $oldbiblio->{othercount} = $other_count;
1231 $oldbiblio->{otherplural} = 1 if $other_count>1;
1232 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1233 $oldbiblio->{itemlostcount} = $itemlost_count;
1234 $oldbiblio->{damagedcount} = $itemdamaged_count;
1235 $oldbiblio->{orderedcount} = $ordered_count;
1236 $oldbiblio->{isbn} =~ s/-//g; # deleting - in isbn to enable amazon content
1237 push( @newresults, $oldbiblio );
1244 #----------------------------------------------------------------------
1246 # Non-Zebra GetRecords#
1247 #----------------------------------------------------------------------
1251 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1255 my ($query,$simple_query,$sort_by_ref,$servers_ref,$results_per_page,$offset,$expanded_facet,$branches,$query_type,$scan) = @_;
1256 warn "query =$query" if $DEBUG;
1257 my $result = NZanalyse($query);
1258 warn "results =$result" if $DEBUG;
1259 return (undef,NZorder($result,@$sort_by_ref[0],$results_per_page,$offset),undef);
1264 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1265 the list is built from an inverted index in the nozebra SQL table
1266 note that title is here only for convenience : the sorting will be very fast when requested on title
1267 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1272 my ($string,$server) = @_;
1273 warn "---------" if $DEBUG;
1274 warn "Enter NZanalyse" if $DEBUG;
1275 warn "---------" if $DEBUG;
1277 # $server contains biblioserver or authorities, depending on what we search on.
1278 #warn "querying : $string on $server";
1279 $server='biblioserver' unless $server;
1281 # if we have a ", replace the content to discard temporarily any and/or/not inside
1283 if ($string =~/"/) {
1284 $string =~ s/"(.*?)"/__X__/;
1286 warn "commacontent : $commacontent" if $DEBUG;
1288 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1289 # then, call again NZanalyse with $left and $right
1290 # (recursive until we find a leaf (=> something without and/or/not)
1291 # delete repeated operator... Would then go in infinite loop
1292 while ($string =~s/( and| or| not| AND| OR| NOT)\1/$1/g){
1294 #process parenthesis before.
1295 if ($string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/){
1298 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1299 warn "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right" if $DEBUG;
1300 my $leftresult = NZanalyse($left,$server);
1302 my $rightresult = NZanalyse($right,$server);
1303 # OK, we have the results for right and left part of the query
1304 # depending of operand, intersect, union or exclude both lists
1305 # to get a result list
1306 if ($operator eq ' and ') {
1307 my @leftresult = split /;/, $leftresult;
1308 warn " @leftresult / $rightresult \n" if $DEBUG;
1309 # my @rightresult = split /;/,$leftresult;
1311 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1312 # the result is stored twice, to have the same weight for AND than OR.
1313 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1314 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1315 foreach (@leftresult) {
1318 ($value,$countvalue)=($1,$2) if $value=~m/(.*)-(\d+)$/;
1319 if ($rightresult =~ /$value-(\d+);/) {
1320 $countvalue=($1>$countvalue?$countvalue:$1);
1321 $finalresult .= "$value-$countvalue;$value-$countvalue;";
1324 warn " $finalresult \n" if $DEBUG;
1325 return $finalresult;
1326 } elsif ($operator eq ' or ') {
1327 # just merge the 2 strings
1328 return $leftresult.$rightresult;
1329 } elsif ($operator eq ' not ') {
1330 my @leftresult = split /;/, $leftresult;
1331 # my @rightresult = split /;/,$leftresult;
1333 foreach (@leftresult) {
1335 $value=$1 if $value=~m/(.*)-\d+$/;
1336 unless ($rightresult =~ "$value-") {
1339 return $finalresult;
1341 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1347 warn "string :".$string if $DEBUG;
1348 $string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/;
1351 my $operator = lc($2); # FIXME: and/or/not are operators, not operands
1352 warn "dealing w/parenthesis. left :$left operator:$operator right:$right" if $DEBUG;
1353 # it's not a leaf, we have a and/or/not
1355 # reintroduce comma content if needed
1356 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1357 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1358 warn "node : $left / $operator / $right\n" if $DEBUG;
1359 my $leftresult = NZanalyse($left,$server);
1360 my $rightresult = NZanalyse($right,$server);
1361 # OK, we have the results for right and left part of the query
1362 # depending of operand, intersect, union or exclude both lists
1363 # to get a result list
1364 if ($operator eq ' and ') {
1365 my @leftresult = split /;/, $leftresult;
1366 # my @rightresult = split /;/,$leftresult;
1368 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1369 # the result is stored twice, to have the same weight for AND than OR.
1370 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1371 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1372 foreach (@leftresult) {
1373 if ($rightresult =~ "$_;") {
1374 $finalresult .= "$_;$_;";
1377 return $finalresult;
1378 } elsif ($operator eq ' or ') {
1379 # just merge the 2 strings
1380 return $leftresult.$rightresult;
1381 } elsif ($operator eq ' not ') {
1382 my @leftresult = split /;/, $leftresult;
1383 # my @rightresult = split /;/,$leftresult;
1385 foreach (@leftresult) {
1386 unless ($rightresult =~ "$_;") {
1387 $finalresult .= "$_;";
1390 return $finalresult;
1392 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1393 die "error : operand unknown : $operator for $string";
1395 # it's a leaf, do the real SQL query and return the result
1397 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1398 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\///g; # we must not introduce spaces in place of these chars
1399 warn "leaf:$string" if $DEBUG;
1400 # parse the string in in operator/operand/value again
1401 $string =~ /(.*)(>=|<=)(.*)/;
1405 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1406 unless ($operator) {
1407 $string =~ /(.*)(>|<|=)(.*)/;
1411 warn "handling unless (operator)... left:$left operator:$operator right:$right" if $DEBUG;
1414 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1415 $left =~ s/[ ,].*$//;
1416 # automatic replace for short operators
1417 $left='title' if $left =~ '^ti$';
1418 $left='author' if $left =~ '^au$';
1419 $left='publisher' if $left =~ '^pb$';
1420 $left='subject' if $left =~ '^su$';
1421 $left='koha-Auth-Number' if $left =~ '^an$';
1422 $left='keyword' if $left =~ '^kw$';
1423 $left='itemtype' if $left =~ '^mc$'; # we must allow for limit operators since buildQuery will append $limit to $query
1424 if ($operator && $left ne 'keyword' ) {
1425 #do a specific search
1426 my $dbh = C4::Context->dbh;
1427 $operator='LIKE' if $operator eq '=' and $right=~ /%/;
1428 my $sth = $dbh->prepare("SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?");
1429 warn "$left / $operator / $right\n";
1430 # split each word, query the DB and build the biblionumbers result
1431 #sanitizing leftpart
1432 $left=~s/^\s+|\s+$//;
1433 foreach (split / /,$right) {
1437 warn "EXECUTE : $server, $left, $_";
1438 $sth->execute($server, $left, $_) or warn "execute failed: $!";
1439 while (my ($line,$value) = $sth->fetchrow) {
1440 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1441 # otherwise, fill the result
1442 $biblionumbers .= $line unless ($right =~ /^\d+$/ && $value =~ /\D/);
1443 warn "result : $value ". ($right =~ /\d/) . "==".(!$value =~ /\d/) ;#= $line";
1445 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1447 my @leftresult = split /;/, $biblionumbers;
1449 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1450 # remove weight at the end
1451 my $cleaned = $entry;
1452 $cleaned =~ s/-\d*$//;
1453 # if the entry already in the hash, take it & increase weight
1454 warn "===== $cleaned =====" if $DEBUG;
1455 if ($results =~ "$cleaned") {
1456 $temp .= "$entry;$entry;";
1457 warn "INCLUDING $entry" if $DEBUG;
1462 $results = $biblionumbers;
1466 #do a complete search (all indexes), if index='kw' do complete search too.
1467 my $dbh = C4::Context->dbh;
1468 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?");
1469 # split each word, query the DB and build the biblionumbers result
1470 foreach (split / /,$string) {
1471 next if C4::Context->stopwords->{uc($_)}; # skip if stopword
1472 warn "search on all indexes on $_" if $DEBUG;
1475 $sth->execute($server, $_);
1476 while (my $line = $sth->fetchrow) {
1477 $biblionumbers .= $line;
1479 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1481 warn "RES for $_ = $biblionumbers" if $DEBUG;
1482 my @leftresult = split /;/, $biblionumbers;
1484 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1485 # remove weight at the end
1486 my $cleaned = $entry;
1487 $cleaned =~ s/-\d*$//;
1488 # if the entry already in the hash, take it & increase weight
1489 # warn "===== $cleaned =====" if $DEBUG;
1490 if ($results =~ "$cleaned") {
1491 $temp .= "$entry;$entry;";
1492 # warn "INCLUDING $entry" if $DEBUG;
1497 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
1498 $results = $biblionumbers;
1502 warn "return : $results for LEAF : $string" if $DEBUG;
1505 warn "---------" if $DEBUG;
1506 warn "Leave NZanalyse" if $DEBUG;
1507 warn "---------" if $DEBUG;
1512 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
1520 my ($biblionumbers, $ordering,$results_per_page,$offset) = @_;
1521 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
1522 # order title asc by default
1523 # $ordering = '1=36 <i' unless $ordering;
1524 $results_per_page=20 unless $results_per_page;
1525 $offset = 0 unless $offset;
1526 my $dbh = C4::Context->dbh;
1528 # order by POPULARITY
1530 if ($ordering =~ /popularity/) {
1533 # popularity is not in MARC record, it's builded from a specific query
1534 my $sth = $dbh->prepare("select sum(issues) from items where biblionumber=?");
1535 foreach (split /;/,$biblionumbers) {
1536 my ($biblionumber,$title) = split /,/,$_;
1537 $result{$biblionumber}=GetMarcBiblio($biblionumber);
1538 $sth->execute($biblionumber);
1539 my $popularity= $sth->fetchrow ||0;
1540 # hint : the key is popularity.title because we can have
1541 # many results with the same popularity. In this cas, sub-ordering is done by title
1542 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1543 # (un-frequent, I agree, but we won't forget anything that way ;-)
1544 $popularity{sprintf("%10d",$popularity).$title.$biblionumber} = $biblionumber;
1546 # sort the hash and return the same structure as GetRecords (Zebra querying)
1549 if ($ordering eq 'popularity_dsc') { # sort popularity DESC
1550 foreach my $key (sort {$b cmp $a} (keys %popularity)) {
1551 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1553 } else { # sort popularity ASC
1554 foreach my $key (sort (keys %popularity)) {
1555 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1559 $result_hash->{'hits'} = $numbers;
1560 $finalresult->{'biblioserver'} = $result_hash;
1561 return $finalresult;
1565 } elsif ($ordering =~/author/){
1567 foreach (split /;/,$biblionumbers) {
1568 my ($biblionumber,$title) = split /,/,$_;
1569 my $record=GetMarcBiblio($biblionumber);
1571 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1572 $author=$record->subfield('200','f');
1573 $author=$record->subfield('700','a') unless $author;
1575 $author=$record->subfield('100','a');
1577 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1578 # and we don't want to get only 1 result for each of them !!!
1579 $result{$author.$biblionumber}=$record;
1581 # sort the hash and return the same structure as GetRecords (Zebra querying)
1584 if ($ordering eq 'author_za') { # sort by author desc
1585 foreach my $key (sort { $b cmp $a } (keys %result)) {
1586 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1588 } else { # sort by author ASC
1589 foreach my $key (sort (keys %result)) {
1590 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1594 $result_hash->{'hits'} = $numbers;
1595 $finalresult->{'biblioserver'} = $result_hash;
1596 return $finalresult;
1598 # ORDER BY callnumber
1600 } elsif ($ordering =~/callnumber/){
1602 foreach (split /;/,$biblionumbers) {
1603 my ($biblionumber,$title) = split /,/,$_;
1604 my $record=GetMarcBiblio($biblionumber);
1606 my ($callnumber_tag,$callnumber_subfield)=GetMarcFromKohaField($dbh,'items.itemcallnumber');
1607 ($callnumber_tag,$callnumber_subfield)= GetMarcFromKohaField('biblioitems.callnumber') unless $callnumber_tag;
1608 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1609 $callnumber=$record->subfield('200','f');
1611 $callnumber=$record->subfield('100','a');
1613 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1614 # and we don't want to get only 1 result for each of them !!!
1615 $result{$callnumber.$biblionumber}=$record;
1617 # sort the hash and return the same structure as GetRecords (Zebra querying)
1620 if ($ordering eq 'call_number_dsc') { # sort by title desc
1621 foreach my $key (sort { $b cmp $a } (keys %result)) {
1622 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1624 } else { # sort by title ASC
1625 foreach my $key (sort { $a cmp $b } (keys %result)) {
1626 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1630 $result_hash->{'hits'} = $numbers;
1631 $finalresult->{'biblioserver'} = $result_hash;
1632 return $finalresult;
1633 } elsif ($ordering =~ /pubdate/){ #pub year
1635 foreach (split /;/,$biblionumbers) {
1636 my ($biblionumber,$title) = split /,/,$_;
1637 my $record=GetMarcBiblio($biblionumber);
1638 my ($publicationyear_tag,$publicationyear_subfield)=GetMarcFromKohaField('biblioitems.publicationyear','');
1639 my $publicationyear=$record->subfield($publicationyear_tag,$publicationyear_subfield);
1640 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1641 # and we don't want to get only 1 result for each of them !!!
1642 $result{$publicationyear.$biblionumber}=$record;
1644 # sort the hash and return the same structure as GetRecords (Zebra querying)
1647 if ($ordering eq 'pubdate_dsc') { # sort by pubyear desc
1648 foreach my $key (sort { $b cmp $a } (keys %result)) {
1649 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1651 } else { # sort by pub year ASC
1652 foreach my $key (sort (keys %result)) {
1653 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1657 $result_hash->{'hits'} = $numbers;
1658 $finalresult->{'biblioserver'} = $result_hash;
1659 return $finalresult;
1663 } elsif ($ordering =~ /title/) {
1664 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
1666 foreach (split /;/,$biblionumbers) {
1667 my ($biblionumber,$title) = split /,/,$_;
1668 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1669 # and we don't want to get only 1 result for each of them !!!
1670 # hint & speed improvement : we can order without reading the record
1671 # so order, and read records only for the requested page !
1672 $result{$title.$biblionumber}=$biblionumber;
1674 # sort the hash and return the same structure as GetRecords (Zebra querying)
1677 if ($ordering eq 'title_az') { # sort by title desc
1678 foreach my $key (sort (keys %result)) {
1679 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1681 } else { # sort by title ASC
1682 foreach my $key (sort { $b cmp $a } (keys %result)) {
1683 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1686 # limit the $results_per_page to result size if it's more
1687 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1688 # for the requested page, replace biblionumber by the complete record
1689 # speed improvement : avoid reading too much things
1690 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1691 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1694 $result_hash->{'hits'} = $numbers;
1695 $finalresult->{'biblioserver'} = $result_hash;
1696 return $finalresult;
1701 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
1704 foreach (split /;/,$biblionumbers) {
1705 my ($biblionumber,$title) = split /,/,$_;
1706 $title =~ /(.*)-(\d)/;
1709 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
1710 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
1711 # biblio N has ranking = 6
1712 $count_ranking{$biblionumber} += $ranking;
1714 # build the result by "inverting" the count_ranking hash
1715 # 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
1717 foreach (keys %count_ranking) {
1718 $result{sprintf("%10d",$count_ranking{$_}).'-'.$_} = $_;
1720 # sort the hash and return the same structure as GetRecords (Zebra querying)
1723 foreach my $key (sort {$b cmp $a} (keys %result)) {
1724 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1726 # limit the $results_per_page to result size if it's more
1727 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1728 # for the requested page, replace biblionumber by the complete record
1729 # speed improvement : avoid reading too much things
1730 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1731 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc if $result_hash->{'RECORDS'}[$counter];
1734 $result_hash->{'hits'} = $numbers;
1735 $finalresult->{'biblioserver'} = $result_hash;
1736 return $finalresult;
1741 ($countchanged,$listunchanged) = ModBiblios($listbiblios, $tagsubfield,$initvalue,$targetvalue,$test);
1743 this function changes all the values $initvalue in subfield $tag$subfield in any record in $listbiblios
1744 test parameter if set donot perform change to records in database.
1750 * $listbiblios is an array ref to marcrecords to be changed
1751 * $tagsubfield is the reference of the subfield to change.
1752 * $initvalue is the value to search the record for
1753 * $targetvalue is the value to set the subfield to
1754 * $test is to be set only not to perform changes in database.
1756 =item C<Output arg:>
1757 * $countchanged counts all the changes performed.
1758 * $listunchanged contains the list of all the biblionumbers of records unchanged.
1760 =item C<usage in the script:>
1764 my ($countchanged, $listunchanged) = EditBiblios($results->{RECORD}, $tagsubfield,$initvalue,$targetvalue);;
1765 #If one wants to display unchanged records, you should get biblios foreach @$listunchanged
1766 $template->param(countchanged => $countchanged, loopunchanged=>$listunchanged);
1771 my ($listbiblios,$tagsubfield,$initvalue,$targetvalue,$test)=@_;
1774 my ($tag,$subfield)=($1,$2) if ($tagsubfield=~/^(\d{1,3})([a-z0-9A-Z@])?$/);
1775 if ((length($tag)<3)&& $subfield=~/0-9/){
1776 $tag=$tag.$subfield;
1779 my ($bntag,$bnsubf) = GetMarcFromKohaField('biblio.biblionumber');
1780 my ($itemtag,$itemsubf) = GetMarcFromKohaField('items.itemnumber');
1781 foreach my $usmarc (@$listbiblios){
1783 $record=eval{MARC::Record->new_from_usmarc($usmarc)};
1786 # usmarc is not a valid usmarc May be a biblionumber
1787 if ($tag eq $itemtag){
1788 my $bib=GetBiblioFromItemNumber($usmarc);
1789 $record=GetMarcItem($bib->{'biblionumber'},$usmarc) ;
1790 $biblionumber=$bib->{'biblionumber'};
1792 $record=GetMarcBiblio($usmarc);
1793 $biblionumber=$usmarc;
1797 $biblionumber = $record->subfield($bntag,$bnsubf);
1799 $biblionumber=$record->field($bntag)->data;
1802 #GetBiblionumber is to be written.
1803 #Could be replaced by TransformMarcToKoha (But Would be longer)
1804 if ($record->field($tag)){
1806 foreach my $field ($record->field($tag)){
1808 if ($field->delete_subfield('code' =>$subfield,'match'=>qr($initvalue))){
1811 $field->update($subfield,$targetvalue) if ($targetvalue);
1815 if ($field->delete_field($field)){
1820 $field->data=$targetvalue if ($field->data=~qr($initvalue));
1824 # warn $record->as_formatted;
1826 ModBiblio($record,$biblionumber,GetFrameworkCode($biblionumber)) unless ($test);
1828 push @unmatched, $biblionumber;
1831 push @unmatched, $biblionumber;
1834 return ($countmatched,\@unmatched);
1837 END { } # module clean-up code here (global destructor)
1844 Koha Developement team <info@koha.org>