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, $scan) = @_;
691 my @operators = @$operators if $operators;
692 my @indexes = @$indexes if $indexes;
693 my @operands = @$operands if $operands;
694 my @limits = @$limits if $limits;
695 my @sort_by = @$sort_by if $sort_by;
697 my $stemming = C4::Context->preference("QueryStemming") || 0;
698 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
699 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
700 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
701 # no stemming/weight/fuzzy in NoZebra
702 if (C4::Context->preference("NoZebra")) {
707 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
709 my $query = $operands[0];
710 my $simple_query = $operands[0];
719 my $stopwords_removed;
721 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
723 if ( $query =~ /^ccl=/ ) {
724 return ( undef, $', $', $', $', '', '', '', '', 'ccl' );
726 if ( $query =~ /^cql=/ ) {
727 return ( undef, $', $', $', $', '', '', '', '', 'cql' );
729 if ( $query =~ /^pqf=/ ) {
730 return ( undef, $', $', $', $', '', '', '', '', 'pqf' );
733 # pass nested queries directly
734 if ( $query =~ /(\(|\))/ ) {
735 return ( undef, $query, $simple_query, $query_cgi, $query, $limit, $limit_cgi, $limit_desc, $stopwords_removed, 'ccl' );
738 # form-based queries are limited to non-nested at a specific depth, so we can easily
739 # modify the incoming query operands and indexes to do stemming and field weighting
740 # Once we do so, we'll end up with a value in $query, just like if we had an
741 # incoming $query from the user
743 $query = ""; # clear it out so we can populate properly with field-weighted stemmed query
744 my $previous_operand; # a flag used to keep track if there was a previous query
745 # if there was, we can apply the current operator
747 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
749 # COMBINE OPERANDS, INDEXES AND OPERATORS
750 if ( $operands[$i] ) {
752 # a flag to determine whether or not to add the index to the query
755 # if the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
756 if ($operands[$i] =~ /(:|=)/ || $scan) {
759 $remove_stopwords = 0;
761 my $operand = $operands[$i];
762 my $index = $indexes[$i];
764 # add some attributes for certain index types
765 # Date of Publication
766 if ($index eq 'yr') {
767 $index .=",st-numeric";
769 ($stemming,$auto_truncation,$weight_fields, $fuzzy_enabled, $remove_stopwords) = (0,0,0,0,0);
771 # Date of Acquisition
772 elsif ($index eq 'acqdate') {
773 $index.=",st-date-normalized";
775 ($stemming,$auto_truncation,$weight_fields, $fuzzy_enabled, $remove_stopwords) = (0,0,0,0,0);
779 # set default structure attribute (word list)
781 unless (!$index || $index =~ /(st-|phr|ext|wrdl)/) {
782 $struct_attr = ",wrdl";
784 # some helpful index modifs
785 my $index_plus = $index.$struct_attr.":" if $index;
786 my $index_plus_comma=$index.$struct_attr."," if $index;
789 if ($remove_stopwords) {
790 ($operand, $stopwords_removed) = _remove_stopwords($operand,$index);
791 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
792 warn "REMOVED STOPWORDS: @$stopwords_removed" if ($stopwords_removed && $DEBUG);
796 my ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr);
797 my $truncated_operand;
798 ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr) = _detect_truncation($operand,$index);
799 warn "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<" if $DEBUG;
802 if (scalar(@$righttruncated)+scalar(@$lefttruncated)+scalar(@$rightlefttruncated)>0){
803 # don't field weight or add the index to the query, we do it here
805 undef $weight_fields;
806 my $previous_truncation_operand;
807 if (scalar(@$nontruncated)>0) {
808 $truncated_operand.= "$index_plus @$nontruncated ";
809 $previous_truncation_operand = 1;
811 if (scalar(@$righttruncated)>0){
812 $truncated_operand .= "and " if $previous_truncation_operand;
813 $truncated_operand .= "$index_plus_comma"."rtrn:@$righttruncated ";
814 $previous_truncation_operand = 1;
816 if (scalar(@$lefttruncated)>0){
817 $truncated_operand .= "and " if $previous_truncation_operand;
818 $truncated_operand .= "$index_plus_comma"."ltrn:@$lefttruncated ";
819 $previous_truncation_operand = 1;
821 if (scalar(@$rightlefttruncated)>0){
822 $truncated_operand .= "and " if $previous_truncation_operand;
823 $truncated_operand .= "$index_plus_comma"."rltrn:@$rightlefttruncated ";
824 $previous_truncation_operand = 1;
827 $operand = $truncated_operand if $truncated_operand;
828 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
832 $stemmed_operand = _build_stemmed_operand($operand) if $stemming;
833 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
835 # Handle Field Weighting
836 my $weighted_operand;
837 $weighted_operand = _build_weighted_query($operand,$stemmed_operand,$index) if $weight_fields;
838 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
839 $operand = $weighted_operand if $weight_fields;
840 $indexes_set = 1 if $weight_fields;
842 # If there's a previous operand, we need to add an operator
843 if ($previous_operand) {
845 # user-specified operator
846 if ( $operators[$i-1] ) {
847 $query .= " $operators[$i-1] ";
848 $query .= " $index_plus " unless $indexes_set;
849 $query .= " $operand";
850 $query_cgi .="&op=$operators[$i-1]";
851 $query_cgi .="&idx=$index" if $index;
852 $query_cgi .="&q=$operands[$i]" if $operands[$i];
853 $query_desc .=" $operators[$i-1] $index_plus $operands[$i]";
856 # the default operator is and
859 $query .= "$index_plus " unless $indexes_set;
860 $query .= "$operand";
861 $query_cgi .="&op=and&idx=$index" if $index;
862 $query_cgi .="&q=$operands[$i]" if $operands[$i];
863 $query_desc .= " and $index_plus $operands[$i]";
867 # there isn't a pervious operand, don't need an operator
869 # field-weighted queries already have indexes set
870 $query .=" $index_plus " unless $indexes_set;
872 $query_desc .= " $index_plus $operands[$i]";
873 $query_cgi.="&idx=$index" if $index;
874 $query_cgi.="&q=$operands[$i]" if $operands[$i];
876 $previous_operand = 1;
881 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
885 my $availability_limit;
886 foreach my $this_limit (@limits) {
887 if ( $this_limit =~ /available/ ) {
888 # available is defined as (items.notloan is NULL) and (items.itemlost > 0 or NULL) (last clause handles NULL values for lost in zebra)
889 $availability_limit .="( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and ((lost,st-numeric ge 0) or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
890 $limit_cgi .= "&limit=available";
894 # these are treated as OR
895 elsif ( $this_limit =~ /mc/ ) {
896 $group_OR_limits .= " or " if $group_OR_limits;
897 $limit_desc .=" or " if $group_OR_limits;
898 $group_OR_limits .= "$this_limit";
899 $limit_cgi .="&limit=$this_limit";
900 $limit_desc .= " $this_limit";
904 $limit .= " and " if $limit || $query;
905 $limit .= "$this_limit";
906 $limit_cgi .="&limit=$this_limit";
907 $limit_desc .=" $this_limit";
910 if ($group_OR_limits) {
911 $limit.=" and " if ($query || $limit );
912 $limit.="($group_OR_limits)";
914 if ($availability_limit) {
915 $limit.=" not " if ($query || $limit );
916 $limit.="$availability_limit";
918 # normalize the strings
921 for ($query, $query_desc, $limit, $limit_desc) {
922 $_ =~ s/ / /g; # remove extra spaces
923 $_ =~ s/^ //g; # remove any beginning spaces
924 $_ =~ s/ $//g; # remove any ending spaces
925 $_ =~ s/==/=/g; # remove double == from query
928 $query_cgi =~ s/^&//;
930 # append the limit to the query
931 $query .= " ".$limit;
933 warn "QUERY:".$query if $DEBUG;
934 warn "QUERY CGI:".$query_cgi if $DEBUG;
935 warn "QUERY DESC:".$query_desc if $DEBUG;
936 warn "LIMIT:".$limit if $DEBUG;
937 warn "LIMIT CGI:".$limit_cgi if $DEBUG;
938 warn "LIMIT DESC:".$limit_desc if $DEBUG;
940 return ( undef, $query,$simple_query,$query_cgi,$query_desc,$limit,$limit_cgi,$limit_desc,$stopwords_removed,$query_type );
943 # IMO this subroutine is pretty messy still -- it's responsible for
944 # building the HTML output for the template
946 my ( $searchdesc, $hits, $results_per_page, $offset, @marcresults ) = @_;
948 my $dbh = C4::Context->dbh;
952 my $span_terms_hashref;
953 for my $span_term ( split( / /, $searchdesc ) ) {
954 $span_term =~ s/(.*=|\)|\(|\+|\.)//g;
955 $span_terms_hashref->{$span_term}++;
958 #Build brancnames hash
960 #get branch information.....
963 $dbh->prepare("SELECT branchcode,branchname FROM branches")
964 ; # FIXME : use C4::Koha::GetBranches
966 while ( my $bdata = $bsth->fetchrow_hashref ) {
967 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
971 #find itemtype & itemtype image
974 $dbh->prepare("SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes");
976 while ( my $bdata = $bsth->fetchrow_hashref ) {
977 $itemtypes{ $bdata->{'itemtype'} }->{description} =
978 $bdata->{'description'};
979 $itemtypes{ $bdata->{'itemtype'} }->{imageurl} = $bdata->{'imageurl'};
980 $itemtypes{ $bdata->{'itemtype'} }->{summary} = $bdata->{'summary'};
981 $itemtypes{ $bdata->{'itemtype'} }->{notforloan} = $bdata->{'notforloan'};
984 #search item field code
987 "select tagfield from marc_subfield_structure where kohafield like 'items.itemnumber'"
990 my ($itemtag) = $sth->fetchrow;
992 ## find column names of items related to MARC
993 my $sth2 = $dbh->prepare("SHOW COLUMNS from items");
995 my %subfieldstosearch;
996 while ( ( my $column ) = $sth2->fetchrow ) {
997 my ( $tagfield, $tagsubfield ) =
998 &GetMarcFromKohaField( "items." . $column, "" );
999 $subfieldstosearch{$column} = $tagsubfield;
1003 if ( $hits && $offset + $results_per_page <= $hits ) {
1004 $times = $offset + $results_per_page;
1010 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1012 $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1013 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, '' );
1014 $oldbiblio->{result_number} = $i+1;
1015 # add image url if there is one
1016 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} =~ /^http:/ ) {
1017 $oldbiblio->{imageurl} =
1018 $itemtypes{ $oldbiblio->{itemtype} }->{imageurl};
1019 $oldbiblio->{description} =
1020 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1023 $oldbiblio->{imageurl} =
1024 getitemtypeimagesrc() . "/"
1025 . $itemtypes{ $oldbiblio->{itemtype} }->{imageurl}
1026 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1027 $oldbiblio->{description} =
1028 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1031 # build summary if there is one (the summary is defined in itemtypes table
1033 if ($itemtypes{ $oldbiblio->{itemtype} }->{summary}) {
1034 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1035 my @fields = $marcrecord->fields();
1036 foreach my $field (@fields) {
1037 my $tag = $field->tag();
1038 my $tagvalue = $field->as_string();
1039 $summary =~ s/\[(.?.?.?.?)$tag\*(.*?)]/$1$tagvalue$2\[$1$tag$2]/g;
1041 my @subf = $field->subfields;
1042 for my $i (0..$#subf) {
1043 my $subfieldcode = $subf[$i][0];
1044 my $subfieldvalue = $subf[$i][1];
1045 my $tagsubf = $tag.$subfieldcode;
1046 $summary =~ s/\[(.?.?.?.?)$tagsubf(.*?)]/$1$subfieldvalue$2\[$1$tagsubf$2]/g;
1050 $summary =~ s/\[(.*?)]//g;
1051 $summary =~ s/\n/<br>/g;
1052 $oldbiblio->{summary} = $summary;
1054 # add spans to search term in results for search term highlighting
1055 # save a native author, for the <a href=search.lq=<!--tmpl_var name="author"-->> link
1056 my $searchhighlightblob;
1057 for my $highlight_field ($marcrecord->fields) {
1058 next if $highlight_field->tag() =~ /(^00)/; # skip fixed fields
1060 my $field = $highlight_field->as_string();
1061 for my $term ( keys %$span_terms_hashref ) {
1062 if (($field =~ /$term/i) && (length($term) > 3)) {
1063 $field =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1067 $searchhighlightblob .= $field." ... " if $match;
1069 $oldbiblio->{'searchhighlightblob'} = $searchhighlightblob;
1071 $oldbiblio->{'author_nospan'} = $oldbiblio->{'author'};
1072 for my $term ( keys %$span_terms_hashref ) {
1073 my $old_term = $term;
1074 if ( length($term) > 3 ) {
1075 $term =~ s/(.*=|\)|\(|\+|\.|\?|\[|\]|\\|\*)//g;
1076 $oldbiblio->{'title'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1077 $oldbiblio->{'subtitle'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1078 $oldbiblio->{'author'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1079 $oldbiblio->{'publishercode'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1080 $oldbiblio->{'place'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1081 $oldbiblio->{'pages'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1082 $oldbiblio->{'notes'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1083 $oldbiblio->{'size'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1088 $toggle = "#ffffcc";
1093 $oldbiblio->{'toggle'} = $toggle;
1094 my @fields = $marcrecord->field($itemtag);
1097 my $ordered_count = 0;
1098 my $onloan_count = 0;
1099 my $wthdrawn_count = 0;
1100 my $itemlost_count = 0;
1104 # check the loan status of the item :
1105 # it is not stored in the MARC record, for pref (zebra reindexing)
1106 # reason. Thus, we have to get the status from a specific SQL query
1108 my $sth_issue = $dbh->prepare("
1109 SELECT date_due,returndate
1111 WHERE itemnumber=? AND returndate IS NULL");
1112 my $items_count=scalar(@fields);
1113 foreach my $field (@fields) {
1115 foreach my $code ( keys %subfieldstosearch ) {
1116 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1118 $sth_issue->execute($item->{itemnumber});
1119 $item->{due_date} = format_date($sth_issue->fetchrow) if $sth_issue->fetchrow;
1120 $item->{onloan} = 1 if $item->{due_date};
1121 # at least one item can be reserved : suppose no
1123 if ( $item->{wthdrawn} ) {
1125 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1126 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{wthdrawn}=1;
1128 elsif ( $item->{itemlost} ) {
1130 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1131 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemlost}=1;
1133 unless ( $item->{notforloan}) {
1134 # OK, this one can be issued, so at least one can be reserved
1137 if ( ( $item->{onloan} ) && ( $item->{onloan} != '0000-00-00' ) )
1139 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1140 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{onloancount} = 1;
1141 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{due_date} = $item->{due_date};
1144 if ( $item->{'homebranch'} ) {
1145 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{count}++;
1149 elsif ( $item->{'holdingbranch'} ) {
1150 $items->{ $item->{'holdingbranch'} }->{count}++;
1152 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemcallnumber} = $item->{itemcallnumber};
1153 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{location} = $item->{location};
1154 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{branchcode} = $item->{homebranch};
1155 } # notforloan, item level and biblioitem level
1157 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1158 $norequests = 1 if $itemtypes{$oldbiblio->{itemtype}}->{notforloan};
1160 for my $key ( sort keys %$items ) {
1163 branchname => $branches{$items->{$key}->{branchcode}},
1164 branchcode => $items->{$key}->{branchcode},
1165 count => $items->{$key}->{count},
1166 itemcallnumber => $items->{$key}->{itemcallnumber},
1167 location => $items->{$key}->{location},
1168 onloancount => $items->{$key}->{onloancount},
1169 due_date => $items->{$key}->{due_date},
1170 wthdrawn => $items->{$key}->{wthdrawn},
1171 lost => $items->{$key}->{itemlost},
1173 # only show the number specified by the user
1174 my $maxitems = (C4::Context->preference('maxItemsinSearchResults')) ? C4::Context->preference('maxItemsinSearchResults')- 1 : 1;
1175 push @items_loop, $this_item unless $itemscount > $maxitems;;
1177 $oldbiblio->{norequests} = $norequests;
1178 $oldbiblio->{items_count} = $items_count;
1179 $oldbiblio->{items_loop} = \@items_loop;
1180 $oldbiblio->{onloancount} = $onloan_count;
1181 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1182 $oldbiblio->{itemlostcount} = $itemlost_count;
1183 $oldbiblio->{orderedcount} = $ordered_count;
1184 $oldbiblio->{isbn} =~ s/-//g; # deleting - in isbn to enable amazon content
1185 push( @newresults, $oldbiblio );
1192 #----------------------------------------------------------------------
1194 # Non-Zebra GetRecords#
1195 #----------------------------------------------------------------------
1199 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1203 my ($query,$simple_query,$sort_by_ref,$servers_ref,$results_per_page,$offset,$expanded_facet,$branches,$query_type,$scan) = @_;
1204 my $result = NZanalyse($query);
1205 return (undef,NZorder($result,@$sort_by_ref[0],$results_per_page,$offset),undef);
1210 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1211 the list is built from an inverted index in the nozebra SQL table
1212 note that title is here only for convenience : the sorting will be very fast when requested on title
1213 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1218 my ($string,$server) = @_;
1219 # $server contains biblioserver or authorities, depending on what we search on.
1220 #warn "querying : $string on $server";
1221 $server='biblioserver' unless $server;
1223 # if we have a ", replace the content to discard temporarily any and/or/not inside
1225 if ($string =~/"/) {
1226 $string =~ s/"(.*?)"/__X__/;
1228 warn "commacontent : $commacontent" if $DEBUG;
1230 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1231 # then, call again NZanalyse with $left and $right
1232 # (recursive until we find a leaf (=> something without and/or/not)
1233 # delete repeated operator... Would then go in infinite loop
1234 while ($string =~s/( and| or| not| AND| OR| NOT)\1/$1/g){
1236 #process parenthesis before.
1237 if ($string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/){
1239 # warn "left :".$left;
1241 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1242 my $leftresult = NZanalyse($left,$server);
1244 my $rightresult = NZanalyse($right,$server);
1245 # OK, we have the results for right and left part of the query
1246 # depending of operand, intersect, union or exclude both lists
1247 # to get a result list
1248 if ($operator eq ' and ') {
1249 my @leftresult = split /;/, $leftresult;
1250 # my @rightresult = split /;/,$leftresult;
1252 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1253 # the result is stored twice, to have the same weight for AND than OR.
1254 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1255 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1256 foreach (@leftresult) {
1257 if ($rightresult =~ "$_;") {
1258 $finalresult .= "$_;$_;";
1261 return $finalresult;
1262 } elsif ($operator eq ' or ') {
1263 # just merge the 2 strings
1264 return $leftresult.$rightresult;
1265 } elsif ($operator eq ' not ') {
1266 my @leftresult = split /;/, $leftresult;
1267 # my @rightresult = split /;/,$leftresult;
1269 foreach (@leftresult) {
1270 unless ($rightresult =~ "$_;") {
1271 $finalresult .= "$_;";
1274 return $finalresult;
1276 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1282 warn "string :".$string if $DEBUG;
1283 $string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/;
1286 my $operand = lc($2); # FIXME: and/or/not are operators, not operands
1287 # it's not a leaf, we have a and/or/not
1289 # reintroduce comma content if needed
1290 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1291 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1292 warn "node : $left / $operand / $right\n" if $DEBUG;
1293 my $leftresult = NZanalyse($left,$server);
1294 my $rightresult = NZanalyse($right,$server);
1295 # OK, we have the results for right and left part of the query
1296 # depending of operand, intersect, union or exclude both lists
1297 # to get a result list
1298 if ($operand eq ' and ') {
1299 my @leftresult = split /;/, $leftresult;
1300 # my @rightresult = split /;/,$leftresult;
1302 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1303 # the result is stored twice, to have the same weight for AND than OR.
1304 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1305 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1306 foreach (@leftresult) {
1307 if ($rightresult =~ "$_;") {
1308 $finalresult .= "$_;$_;";
1311 return $finalresult;
1312 } elsif ($operand eq ' or ') {
1313 # just merge the 2 strings
1314 return $leftresult.$rightresult;
1315 } elsif ($operand eq ' not ') {
1316 my @leftresult = split /;/, $leftresult;
1317 # my @rightresult = split /;/,$leftresult;
1319 foreach (@leftresult) {
1320 unless ($rightresult =~ "$_;") {
1321 $finalresult .= "$_;";
1324 return $finalresult;
1326 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1327 die "error : operand unknown : $operand for $string";
1329 # it's a leaf, do the real SQL query and return the result
1331 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1332 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1333 warn "leaf : $string\n" if $DEBUG;
1334 # parse the string in in operator/operand/value again
1335 $string =~ /(.*)(>=|<=)(.*)/;
1339 unless ($operator) {
1340 $string =~ /(.*)(>|<|=)(.*)/;
1346 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1347 $left =~ s/[ ,].*$//;
1348 # automatic replace for short operators
1349 $left='title' if $left =~ '^ti$';
1350 $left='author' if $left =~ '^au$';
1351 $left='publisher' if $left =~ '^pb$';
1352 $left='subject' if $left =~ '^su$';
1353 $left='koha-Auth-Number' if $left =~ '^an$';
1354 $left='keyword' if $left =~ '^kw$';
1355 if ($operator && $left ne 'keyword' ) {
1356 #do a specific search
1357 my $dbh = C4::Context->dbh;
1358 $operator='LIKE' if $operator eq '=' and $right=~ /%/;
1359 my $sth = $dbh->prepare("SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?");
1360 warn "$left / $operator / $right\n";
1361 # split each word, query the DB and build the biblionumbers result
1362 #sanitizing leftpart
1363 $left=~s/^\s+|\s+$//;
1364 my ($biblionumbers,$value);
1365 foreach (split / /,$right) {
1367 warn "EXECUTE : $server, $left, $_";
1368 $sth->execute($server, $left, $_) or warn "execute failed: $!";
1369 while (my ($line,$value) = $sth->fetchrow) {
1370 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1371 # otherwise, fill the result
1372 $biblionumbers .= $line unless ($right =~ /\d/ && $value =~ /\D/);
1373 # warn "result : $value ". ($right =~ /\d/) . "==".(!$value =~ /\d/) ;#= $line";
1375 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1377 my @leftresult = split /;/, $biblionumbers;
1379 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1380 # remove weight at the end
1381 my $cleaned = $entry;
1382 $cleaned =~ s/-\d*$//;
1383 # if the entry already in the hash, take it & increase weight
1384 warn "===== $cleaned =====" if $DEBUG;
1385 if ($results =~ "$cleaned") {
1386 $temp .= "$entry;$entry;";
1387 warn "INCLUDING $entry" if $DEBUG;
1392 $results = $biblionumbers;
1396 #do a complete search (all indexes), if index='kw' do complete search too.
1397 my $dbh = C4::Context->dbh;
1398 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?");
1399 # split each word, query the DB and build the biblionumbers result
1400 foreach (split / /,$string) {
1401 next if C4::Context->stopwords->{uc($_)}; # skip if stopword
1402 warn "search on all indexes on $_" if $DEBUG;
1405 $sth->execute($server, $_);
1406 while (my $line = $sth->fetchrow) {
1407 $biblionumbers .= $line;
1409 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1411 warn "RES for $_ = $biblionumbers" if $DEBUG;
1412 my @leftresult = split /;/, $biblionumbers;
1414 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1415 # remove weight at the end
1416 my $cleaned = $entry;
1417 $cleaned =~ s/-\d*$//;
1418 # if the entry already in the hash, take it & increase weight
1419 warn "===== $cleaned =====" if $DEBUG;
1420 if ($results =~ "$cleaned") {
1421 $temp .= "$entry;$entry;";
1422 warn "INCLUDING $entry" if $DEBUG;
1427 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
1428 $results = $biblionumbers;
1432 warn "return : $results for LEAF : $string" if $DEBUG;
1439 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
1447 my ($biblionumbers, $ordering,$results_per_page,$offset) = @_;
1448 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
1449 # order title asc by default
1450 # $ordering = '1=36 <i' unless $ordering;
1451 $results_per_page=20 unless $results_per_page;
1452 $offset = 0 unless $offset;
1453 my $dbh = C4::Context->dbh;
1455 # order by POPULARITY
1457 if ($ordering =~ /popularity/) {
1460 # popularity is not in MARC record, it's builded from a specific query
1461 my $sth = $dbh->prepare("select sum(issues) from items where biblionumber=?");
1462 foreach (split /;/,$biblionumbers) {
1463 my ($biblionumber,$title) = split /,/,$_;
1464 $result{$biblionumber}=GetMarcBiblio($biblionumber);
1465 $sth->execute($biblionumber);
1466 my $popularity= $sth->fetchrow ||0;
1467 # hint : the key is popularity.title because we can have
1468 # many results with the same popularity. In this cas, sub-ordering is done by title
1469 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1470 # (un-frequent, I agree, but we won't forget anything that way ;-)
1471 $popularity{sprintf("%10d",$popularity).$title.$biblionumber} = $biblionumber;
1473 # sort the hash and return the same structure as GetRecords (Zebra querying)
1476 if ($ordering eq 'popularity_dsc') { # sort popularity DESC
1477 foreach my $key (sort {$b cmp $a} (keys %popularity)) {
1478 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1480 } else { # sort popularity ASC
1481 foreach my $key (sort (keys %popularity)) {
1482 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1486 $result_hash->{'hits'} = $numbers;
1487 $finalresult->{'biblioserver'} = $result_hash;
1488 return $finalresult;
1492 } elsif ($ordering =~/author/){
1494 foreach (split /;/,$biblionumbers) {
1495 my ($biblionumber,$title) = split /,/,$_;
1496 my $record=GetMarcBiblio($biblionumber);
1498 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1499 $author=$record->subfield('200','f');
1500 $author=$record->subfield('700','a') unless $author;
1502 $author=$record->subfield('100','a');
1504 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1505 # and we don't want to get only 1 result for each of them !!!
1506 $result{$author.$biblionumber}=$record;
1508 # sort the hash and return the same structure as GetRecords (Zebra querying)
1511 if ($ordering eq 'author_za') { # sort by author desc
1512 foreach my $key (sort { $b cmp $a } (keys %result)) {
1513 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1515 } else { # sort by author ASC
1516 foreach my $key (sort (keys %result)) {
1517 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1521 $result_hash->{'hits'} = $numbers;
1522 $finalresult->{'biblioserver'} = $result_hash;
1523 return $finalresult;
1525 # ORDER BY callnumber
1527 } elsif ($ordering =~/callnumber/){
1529 foreach (split /;/,$biblionumbers) {
1530 my ($biblionumber,$title) = split /,/,$_;
1531 my $record=GetMarcBiblio($biblionumber);
1533 my ($callnumber_tag,$callnumber_subfield)=GetMarcFromKohaField($dbh,'items.itemcallnumber');
1534 ($callnumber_tag,$callnumber_subfield)= GetMarcFromKohaField('biblioitems.callnumber') unless $callnumber_tag;
1535 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1536 $callnumber=$record->subfield('200','f');
1538 $callnumber=$record->subfield('100','a');
1540 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1541 # and we don't want to get only 1 result for each of them !!!
1542 $result{$callnumber.$biblionumber}=$record;
1544 # sort the hash and return the same structure as GetRecords (Zebra querying)
1547 if ($ordering eq 'call_number_dsc') { # sort by title desc
1548 foreach my $key (sort { $b cmp $a } (keys %result)) {
1549 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1551 } else { # sort by title ASC
1552 foreach my $key (sort { $a cmp $b } (keys %result)) {
1553 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1557 $result_hash->{'hits'} = $numbers;
1558 $finalresult->{'biblioserver'} = $result_hash;
1559 return $finalresult;
1560 } elsif ($ordering =~ /pubdate/){ #pub year
1562 foreach (split /;/,$biblionumbers) {
1563 my ($biblionumber,$title) = split /,/,$_;
1564 my $record=GetMarcBiblio($biblionumber);
1565 my ($publicationyear_tag,$publicationyear_subfield)=GetMarcFromKohaField('biblioitems.publicationyear','');
1566 my $publicationyear=$record->subfield($publicationyear_tag,$publicationyear_subfield);
1567 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1568 # and we don't want to get only 1 result for each of them !!!
1569 $result{$publicationyear.$biblionumber}=$record;
1571 # sort the hash and return the same structure as GetRecords (Zebra querying)
1574 if ($ordering eq 'pubdate_dsc') { # sort by pubyear desc
1575 foreach my $key (sort { $b cmp $a } (keys %result)) {
1576 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1578 } else { # sort by pub year ASC
1579 foreach my $key (sort (keys %result)) {
1580 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1584 $result_hash->{'hits'} = $numbers;
1585 $finalresult->{'biblioserver'} = $result_hash;
1586 return $finalresult;
1590 } elsif ($ordering =~ /title/) {
1591 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
1593 foreach (split /;/,$biblionumbers) {
1594 my ($biblionumber,$title) = split /,/,$_;
1595 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1596 # and we don't want to get only 1 result for each of them !!!
1597 # hint & speed improvement : we can order without reading the record
1598 # so order, and read records only for the requested page !
1599 $result{$title.$biblionumber}=$biblionumber;
1601 # sort the hash and return the same structure as GetRecords (Zebra querying)
1604 if ($ordering eq 'title_az') { # sort by title desc
1605 foreach my $key (sort (keys %result)) {
1606 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1608 } else { # sort by title ASC
1609 foreach my $key (sort { $b cmp $a } (keys %result)) {
1610 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1613 # limit the $results_per_page to result size if it's more
1614 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1615 # for the requested page, replace biblionumber by the complete record
1616 # speed improvement : avoid reading too much things
1617 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1618 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1621 $result_hash->{'hits'} = $numbers;
1622 $finalresult->{'biblioserver'} = $result_hash;
1623 return $finalresult;
1628 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
1631 foreach (split /;/,$biblionumbers) {
1632 my ($biblionumber,$title) = split /,/,$_;
1633 $title =~ /(.*)-(\d)/;
1636 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
1637 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
1638 # biblio N has ranking = 6
1639 $count_ranking{$biblionumber} += $ranking;
1641 # build the result by "inverting" the count_ranking hash
1642 # 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
1644 foreach (keys %count_ranking) {
1645 $result{sprintf("%10d",$count_ranking{$_}).'-'.$_} = $_;
1647 # sort the hash and return the same structure as GetRecords (Zebra querying)
1650 foreach my $key (sort {$b cmp $a} (keys %result)) {
1651 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1653 # limit the $results_per_page to result size if it's more
1654 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1655 # for the requested page, replace biblionumber by the complete record
1656 # speed improvement : avoid reading too much things
1657 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1658 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc if $result_hash->{'RECORDS'}[$counter];
1661 $result_hash->{'hits'} = $numbers;
1662 $finalresult->{'biblioserver'} = $result_hash;
1663 return $finalresult;
1668 ($countchanged,$listunchanged) = ModBiblios($listbiblios, $tagsubfield,$initvalue,$targetvalue,$test);
1670 this function changes all the values $initvalue in subfield $tag$subfield in any record in $listbiblios
1671 test parameter if set donot perform change to records in database.
1677 * $listbiblios is an array ref to marcrecords to be changed
1678 * $tagsubfield is the reference of the subfield to change.
1679 * $initvalue is the value to search the record for
1680 * $targetvalue is the value to set the subfield to
1681 * $test is to be set only not to perform changes in database.
1683 =item C<Output arg:>
1684 * $countchanged counts all the changes performed.
1685 * $listunchanged contains the list of all the biblionumbers of records unchanged.
1687 =item C<usage in the script:>
1691 my ($countchanged, $listunchanged) = EditBiblios($results->{RECORD}, $tagsubfield,$initvalue,$targetvalue);;
1692 #If one wants to display unchanged records, you should get biblios foreach @$listunchanged
1693 $template->param(countchanged => $countchanged, loopunchanged=>$listunchanged);
1698 my ($listbiblios,$tagsubfield,$initvalue,$targetvalue,$test)=@_;
1701 my ($tag,$subfield)=($1,$2) if ($tagsubfield=~/^(\d{1,3})([a-z0-9A-Z@])?$/);
1702 if ((length($tag)<3)&& $subfield=~/0-9/){
1703 $tag=$tag.$subfield;
1706 my ($bntag,$bnsubf) = GetMarcFromKohaField('biblio.biblionumber');
1707 my ($itemtag,$itemsubf) = GetMarcFromKohaField('items.itemnumber');
1708 foreach my $usmarc (@$listbiblios){
1710 $record=eval{MARC::Record->new_from_usmarc($usmarc)};
1713 # usmarc is not a valid usmarc May be a biblionumber
1714 if ($tag eq $itemtag){
1715 my $bib=GetBiblioFromItemNumber($usmarc);
1716 $record=GetMarcItem($bib->{'biblionumber'},$usmarc) ;
1717 $biblionumber=$bib->{'biblionumber'};
1719 $record=GetMarcBiblio($usmarc);
1720 $biblionumber=$usmarc;
1724 $biblionumber = $record->subfield($bntag,$bnsubf);
1726 $biblionumber=$record->field($bntag)->data;
1729 #GetBiblionumber is to be written.
1730 #Could be replaced by TransformMarcToKoha (But Would be longer)
1731 if ($record->field($tag)){
1733 foreach my $field ($record->field($tag)){
1735 if ($field->delete_subfield('code' =>$subfield,'match'=>qr($initvalue))){
1738 $field->update($subfield,$targetvalue) if ($targetvalue);
1742 if ($field->delete_field($field)){
1747 $field->data=$targetvalue if ($field->data=~qr($initvalue));
1751 # warn $record->as_formatted;
1753 ModBiblio($record,$biblionumber,GetFrameworkCode($biblionumber)) unless ($test);
1755 push @unmatched, $biblionumber;
1758 push @unmatched, $biblionumber;
1761 return ($countmatched,\@unmatched);
1764 END { } # module clean-up code here (global destructor)
1771 Koha Developement team <info@koha.org>