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.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'}->{'RECORDS'};
207 return (undef,$result);
213 return ( "No query entered", undef ) unless $query;
215 #@servers = (C4::Context->config("biblioserver")) unless @servers;
217 ("biblioserver") unless @servers
218 ; # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
221 for ( my $i = 0 ; $i < @servers ; $i++ ) {
223 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
226 ->search( new ZOOM::Query::CCL2RPN( $query, $zconns[$i] ) );
228 # getting error message if one occured.
230 $zconns[$i]->errmsg() . " ("
231 . $zconns[$i]->errcode() . ") "
232 . $zconns[$i]->addinfo() . " "
233 . $zconns[$i]->diagset();
235 return ( $error, undef ) if $zconns[$i]->errcode();
238 # caught a ZOOM::Exception
242 . $@->addinfo() . " "
245 return ( $error, undef );
250 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
251 $ev = $zconns[ $i - 1 ]->last_event();
252 if ( $ev == ZOOM::Event::ZEND ) {
253 $hits = $tmpresults[ $i - 1 ]->size();
256 for ( my $j = 0 ; $j < $hits ; $j++ ) {
257 my $record = $tmpresults[ $i - 1 ]->record($j)->raw();
258 push @results, $record;
262 return ( undef, \@results );
266 # performs the search
269 $koha_query, $simple_query, $sort_by_ref,
270 $servers_ref, $results_per_page, $offset,
271 $expanded_facet, $branches, $query_type,
274 # warn "Query : $koha_query";
275 my @servers = @$servers_ref;
276 my @sort_by = @$sort_by_ref;
278 # create the zoom connection and query object
282 my $results_hashref = ();
285 my $facets_counter = ();
286 my $facets_info = ();
287 my $facets = getFacets();
289 #### INITIALIZE SOME VARS USED CREATE THE FACETED RESULTS
290 my @facets_loop; # stores the ref to array of hashes for template
291 for ( my $i = 0 ; $i < @servers ; $i++ ) {
292 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
294 # perform the search, create the results objects
295 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
297 if ( $servers[$i] =~ /biblioserver/ ) {
298 $query_to_use = $koha_query;
301 $query_to_use = $simple_query;
304 #$query_to_use = $simple_query if $scan;
305 #warn $simple_query if ($scan && $DEBUG);
306 # check if we've got a query_type defined
310 if ( $query_type =~ /^ccl/ ) {
312 s/\:/\=/g; # change : to = last minute (FIXME)
314 # warn "CCL : $query_to_use";
317 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
320 elsif ( $query_type =~ /^cql/ ) {
322 # warn "CQL : $query_to_use";
325 new ZOOM::Query::CQL( $query_to_use, $zconns[$i] ) );
327 elsif ( $query_type =~ /^pqf/ ) {
329 # warn "PQF : $query_to_use";
332 new ZOOM::Query::PQF( $query_to_use, $zconns[$i] ) );
337 # warn "preparing to scan:$query_to_use";
340 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
344 # warn "LAST : $query_to_use";
347 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
353 warn "WARNING: query problem with $query_to_use " . $@;
356 # concatenate the sort_by limits and pass them to the results object
358 foreach my $sort (@sort_by) {
359 if ($sort eq "author_az") {
360 $sort_by.="1=1003 <i ";
362 elsif ($sort eq "author_za") {
363 $sort_by.="1=1003 >i ";
365 elsif ($sort eq "popularity_asc") {
366 $sort_by.="1=9003 <i ";
368 elsif ($sort eq "popularity_dsc") {
369 $sort_by.="1=9003 >i ";
371 elsif ($sort eq "call_number_asc") {
372 $sort_by.="1=20 <i ";
374 elsif ($sort eq "call_number_dsc") {
375 $sort_by.="1=20 >i ";
377 elsif ($sort eq "pubdate_asc") {
378 $sort_by.="1=31 <i ";
380 elsif ($sort eq "pubdate_dsc") {
381 $sort_by.="1=31 >i ";
383 elsif ($sort eq "acqdate_asc") {
384 $sort_by.="1=32 <i ";
386 elsif ($sort eq "acqdate_dsc") {
387 $sort_by.="1=32 >i ";
389 elsif ($sort eq "title_az") {
392 elsif ($sort eq "title_za") {
397 if ( $results[$i]->sort( "yaz", $sort_by ) < 0) {
398 warn "WARNING sort $sort_by failed";
402 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
403 my $ev = $zconns[ $i - 1 ]->last_event();
404 if ( $ev == ZOOM::Event::ZEND ) {
405 next unless $results[ $i - 1 ];
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 # srote 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(
451 $tmprecord->append_fields($tmptitle);
452 $results_hash->{'RECORDS'}[$j] =
453 $tmprecord->as_usmarc();
456 $record = $results[ $i - 1 ]->record($j)->raw();
458 #warn "RECORD $j:".$record;
459 $results_hash->{'RECORDS'}[$j] =
460 $record; # making a reference to a hash
461 # Fill the facets while we're looping
462 $facet_record = MARC::Record->new_from_usmarc($record);
464 #warn $servers[$i-1].$facet_record->title();
465 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
466 if ( $facets->[$k] ) {
468 for my $tag ( @{ $facets->[$k]->{'tags'} } ) {
469 push @fields, $facet_record->field($tag);
471 for my $field (@fields) {
472 my @subfields = $field->subfields();
473 for my $subfield (@subfields) {
474 my ( $code, $data ) = @$subfield;
476 $facets->[$k]->{'subfield'} )
478 $facets_counter->{ $facets->[$k]
479 ->{'link_value'} }->{$data}++;
483 $facets_info->{ $facets->[$k]->{'link_value'} }
485 $facets->[$k]->{'label_value'};
486 $facets_info->{ $facets->[$k]->{'link_value'} }
487 ->{'expanded'} = $facets->[$k]->{'expanded'};
492 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
495 #print "connection ", $i-1, ": $size hits";
496 #print $results[$i-1]->record(0)->render() if $size > 0;
499 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
500 keys %$facets_counter
504 my $number_of_facets;
505 my @this_facets_array;
508 $facets_counter->{$link_value}
509 ->{$b} <=> $facets_counter->{$link_value}->{$a}
510 } keys %{ $facets_counter->{$link_value} }
514 if ( ( $number_of_facets < 6 )
515 || ( $expanded_facet eq $link_value )
516 || ( $facets_info->{$link_value}->{'expanded'} ) )
519 # sanitize the link value ), ( will cause errors with CCL
520 my $facet_link_value = $one_facet;
521 $facet_link_value =~ s/(\(|\))/ /g;
523 # fix the length that will display in the label
524 my $facet_label_value = $one_facet;
525 $facet_label_value = substr( $one_facet, 0, 20 ) . "..."
526 unless length($facet_label_value) <= 20;
528 # well, if it's a branch, label by the name, not the code
529 if ( $link_value =~ /branch/ ) {
531 $branches->{$one_facet}->{'branchname'};
534 # but we're down with the whole label being in the link's title
535 my $facet_title_value = $one_facet;
537 push @this_facets_array,
541 $facets_counter->{$link_value}->{$one_facet},
542 facet_label_value => $facet_label_value,
543 facet_title_value => $facet_title_value,
544 facet_link_value => $facet_link_value,
545 type_link_value => $link_value,
550 unless ( $facets_info->{$link_value}->{'expanded'} ) {
552 if ( ( $number_of_facets > 6 )
553 && ( $expanded_facet ne $link_value ) );
558 type_link_value => $link_value,
559 type_id => $link_value . "_id",
561 $facets_info->{$link_value}->{'label_value'},
562 facets => \@this_facets_array,
563 expandable => $expandable,
564 expand => $link_value,
570 return ( undef, $results_hashref, \@facets_loop );
574 sub _remove_stopwords {
575 my ($operand,$index) = @_;
576 my @stopwords_removed;
577 # phrase and exact-qualified indexes shouldn't have stopwords removed
578 if ($index!~m/phr|ext/){
579 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
580 # we use IsAlpha unicode definition, to deal correctly with diacritics.
581 # otherwise, a French word like "leçon" woudl be split into "le" "çon", le
582 # is an empty word, we'd get "çon" and wouldn't find anything...
583 foreach (keys %{C4::Context->stopwords}) {
584 next if ($_ =~/(and|or|not)/); # don't remove operators
585 if ($operand =~ /(\P{IsAlpha}$_\P{IsAlpha}|^$_\P{IsAlpha}|\P{IsAlpha}$_$)/) {
586 $operand=~ s/\P{IsAlpha}$_\P{IsAlpha}/ /gi;
587 $operand=~ s/^$_\P{IsAlpha}/ /gi;
588 $operand=~ s/\P{IsAlpha}$_$/ /gi;
589 push @stopwords_removed, $_;
593 return ($operand, \@stopwords_removed);
597 sub _detect_truncation {
598 my ($operand,$index) = @_;
599 my (@nontruncated,@righttruncated,@lefttruncated,@rightlefttruncated,@regexpr);
601 my @wordlist= split (/\s/,$operand);
602 foreach my $word (@wordlist){
603 if ($word=~s/^\*([^\*]+)\*$/$1/){
604 push @rightlefttruncated,$word;
606 elsif($word=~s/^\*([^\*]+)$/$1/){
607 push @lefttruncated,$word;
609 elsif ($word=~s/^([^\*]+)\*$/$1/){
610 push @righttruncated,$word;
612 elsif (index($word,"*")<0){
613 push @nontruncated,$word;
619 return (\@nontruncated,\@righttruncated,\@lefttruncated,\@rightlefttruncated,\@regexpr);
622 sub _build_stemmed_operand {
625 # FIXME: the locale should be set based on the user's language and/or search choice
626 my $stemmer = Lingua::Stem->new( -locale => 'EN-US' );
627 # FIXME: these should be stored in the db so the librarian can modify the behavior
628 $stemmer->add_exceptions(
636 my @words = split( / /, $operand );
637 my $stems = $stemmer->stem(@words);
638 for my $stem (@$stems) {
639 $stemmed_operand .= "$stem";
640 $stemmed_operand .= "?" unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
641 $stemmed_operand .= " ";
643 #warn "STEMMED OPERAND: $stemmed_operand";
644 return $stemmed_operand;
647 sub _build_weighted_query {
648 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
649 # pretty well but will work much better when we have an actual query parser
650 my ($operand,$stemmed_operand,$index) = @_;
651 my $stemming = C4::Context->preference("QueryStemming") || 0;
652 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
653 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
655 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
657 # Keyword, or, no index specified
658 if ( ( $index eq 'kw' ) || ( !$index ) ) {
659 $weighted_query .= "Title-cover,ext,r1=\"$operand\""; # exact title-cover
660 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
661 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
662 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
663 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
664 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\"" if $fuzzy_enabled; # add fuzzy, word list
665 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\"" if ($stemming and $stemmed_operand); # add stemming, right truncation
666 # embedded sorting: 0 a-z; 1 z-a
667 # $weighted_query .= ") or (sort1,aut=1";
669 elsif ( $index eq 'bc' ) {
670 $weighted_query .= "bc=\"$operand\"";
672 # if the index already has more than one qualifier, just wrap the operand
673 # in quotes and pass it back
674 elsif ($index =~ ',') {
675 $weighted_query .=" $index=\"$operand\"";
677 #TODO: build better cases based on specific search indexes
679 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
680 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
681 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
682 $weighted_query .= " or $index,rt,wrd,r3=\"$operand\""; # word list index
684 $weighted_query .= "))"; # close rank specification
685 return $weighted_query;
688 # build the query itself
690 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan) = @_;
692 my @operators = @$operators if $operators;
693 my @indexes = @$indexes if $indexes;
694 my @operands = @$operands if $operands;
695 my @limits = @$limits if $limits;
696 my @sort_by = @$sort_by if $sort_by;
698 my $stemming = C4::Context->preference("QueryStemming") || 0;
699 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
700 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
701 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
702 # no stemming/weight/fuzzy in NoZebra
703 if (C4::Context->preference("NoZebra")) {
708 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
710 my $query = $operands[0];
711 my $simple_query = $operands[0];
720 my $stopwords_removed;
722 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
724 if ( $query =~ /^ccl=/ ) {
725 return ( undef, $', $', $', $', '', '', '', '', 'ccl' );
727 if ( $query =~ /^cql=/ ) {
728 return ( undef, $', $', $', $', '', '', '', '', 'cql' );
730 if ( $query =~ /^pqf=/ ) {
731 return ( undef, $', $', $', $', '', '', '', '', 'pqf' );
734 # pass nested queries directly
735 if ( $query =~ /(\(|\))/ ) {
736 return ( undef, $query, $simple_query, $query_cgi, $query, $limit, $limit_cgi, $limit_desc, $stopwords_removed, 'ccl' );
739 # form-based queries are limited to non-nested at a specific depth, so we can easily
740 # modify the incoming query operands and indexes to do stemming and field weighting
741 # Once we do so, we'll end up with a value in $query, just like if we had an
742 # incoming $query from the user
744 $query = ""; # clear it out so we can populate properly with field-weighted stemmed query
745 my $previous_operand; # a flag used to keep track if there was a previous query
746 # if there was, we can apply the current operator
748 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
750 # COMBINE OPERANDS, INDEXES AND OPERATORS
751 if ( $operands[$i] ) {
753 # 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 some defaults
756 if ($operands[$i] =~ /(:|=)/ || $scan) {
759 $remove_stopwords = 0;
761 my $operand = $operands[$i];
762 my $index = $indexes[$i];
764 # some helpful index modifs
765 my $index_plus = "$index:" if $index;
766 my $index_plus_comma="$index," if $index;
769 if ($remove_stopwords) {
770 ($operand, $stopwords_removed) = _remove_stopwords($operand,$index);
771 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
772 warn "REMOVED STOPWORDS: @$stopwords_removed" if ($stopwords_removed && $DEBUG);
776 my ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr);
777 my $truncated_operand;
778 ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr) = _detect_truncation($operand,$index);
779 warn "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<" if $DEBUG;
782 if (scalar(@$righttruncated)+scalar(@$lefttruncated)+scalar(@$rightlefttruncated)>0){
783 # don't field weight or add the index to the query, we do it here
785 undef $weight_fields;
786 my $previous_truncation_operand;
787 if (scalar(@$nontruncated)>0) {
788 $truncated_operand.= "$index_plus @$nontruncated ";
789 $previous_truncation_operand = 1;
791 if (scalar(@$righttruncated)>0){
792 $truncated_operand .= "and " if $previous_truncation_operand;
793 $truncated_operand .= "$index_plus_comma"."rtrn:@$righttruncated ";
794 $previous_truncation_operand = 1;
796 if (scalar(@$lefttruncated)>0){
797 $truncated_operand .= "and " if $previous_truncation_operand;
798 $truncated_operand .= "$index_plus_comma"."ltrn:@$lefttruncated ";
799 $previous_truncation_operand = 1;
801 if (scalar(@$rightlefttruncated)>0){
802 $truncated_operand .= "and " if $previous_truncation_operand;
803 $truncated_operand .= "$index_plus_comma"."rltrn:@$rightlefttruncated ";
804 $previous_truncation_operand = 1;
807 $operand = $truncated_operand if $truncated_operand;
808 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
812 $stemmed_operand = _build_stemmed_operand($operand) if $stemming;
813 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
815 # Handle Field Weighting
816 my $weighted_operand;
817 $weighted_operand = _build_weighted_query($operand,$stemmed_operand,$index) if $weight_fields;
818 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
819 $operand = $weighted_operand if $weight_fields;
820 $indexes_set = 1 if $weight_fields;
822 # If there's a previous operand, we need to add an operator
823 if ($previous_operand) {
825 # user-specified operator
826 if ( $operators[$i-1] ) {
827 $query .= " $operators[$i-1] ";
828 $query .= " $index_plus " unless $indexes_set;
829 $query .= " $operand";
830 $query_cgi .="&op=$operators[$i-1]";
831 $query_cgi .="&idx=$index" if $index;
832 $query_cgi .="&q=$operands[$i]" if $operands[$i];
833 $query_desc .=" $operators[$i-1] $index_plus $operands[$i]";
836 # the default operator is and
839 $query .= "$index_plus " unless $indexes_set;
840 $query .= "$operand";
841 $query_cgi .="&op=and&idx=$index" if $index;
842 $query_cgi .="&q=$operands[$i]" if $operands[$i];
843 $query_desc .= " and $index_plus $operands[$i]";
847 # there isn't a pervious operand, don't need an operator
849 # field-weighted queries already have indexes set
850 $query .=" $index_plus " unless $indexes_set;
852 $query_desc .= " $index_plus $operands[$i]";
853 $query_cgi.="&idx=$index" if $index;
854 $query_cgi.="&q=$operands[$i]" if $operands[$i];
856 $previous_operand = 1;
861 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
865 my $availability_limit;
866 foreach my $this_limit (@limits) {
867 if ( $this_limit =~ /available/ ) {
868 # available is defined as (items.notloan is NULL) and (items.itemlost > 0 or NULL) (last clause handles NULL values for lost in zebra)
869 $availability_limit .="( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and ((lost,st-numeric gt 0) or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
870 $limit_cgi .= "&limit=available";
874 # these are treated as OR
875 elsif ( $this_limit =~ /mc/ ) {
876 $group_OR_limits .= " or " if $group_OR_limits;
877 $limit_desc .=" or " if $group_OR_limits;
878 $group_OR_limits .= "$this_limit";
879 $limit_cgi .="&limit=$this_limit";
880 $limit_desc .= "$this_limit";
885 $limit .= " and " if $limit || $query;
886 $limit .= "$this_limit";
887 $limit_cgi .="&limit=$this_limit";
888 $limit_desc .=" and $this_limit";
891 if ($group_OR_limits) {
892 $limit.=" and " if ($query || $limit );
893 $limit.="($group_OR_limits)";
895 if ($availability_limit) {
896 $limit.=" and " if ($query || $limit );
897 $limit.="($availability_limit)";
899 # normalize the strings
902 for ($query, $query_desc, $limit, $limit_desc) {
903 $_ =~ s/ / /g; # remove extra spaces
904 $_ =~ s/^ //g; # remove any beginning spaces
905 $_ =~ s/ $//g; # remove any ending spaces
906 $_ =~ s/==/=/g; # remove double == from query
909 $query_cgi =~ s/^&//;
911 # append the limit to the query
914 warn "query=$query and limit=$limit" if $DEBUG;
916 warn "QUERY:".$query if $DEBUG;
917 warn "QUERY CGI:".$query_cgi if $DEBUG;
918 warn "QUERY DESC:".$query_desc if $DEBUG;
919 warn "LIMIT:".$limit if $DEBUG;
920 warn "LIMIT CGI:".$limit_cgi if $DEBUG;
921 warn "LIMIT DESC:".$limit_desc if $DEBUG;
923 return ( undef, $query,$simple_query,$query_cgi,$query_desc,$limit,$limit_cgi,$limit_desc,$stopwords_removed,$query_type );
926 # IMO this subroutine is pretty messy still -- it's responsible for
927 # building the HTML output for the template
929 my ( $searchdesc, $hits, $results_per_page, $offset, @marcresults ) = @_;
930 my $dbh = C4::Context->dbh;
934 my $span_terms_hashref;
935 for my $span_term ( split( / /, $searchdesc ) ) {
936 $span_term =~ s/(.*=|\)|\(|\+|\.)//g;
937 $span_terms_hashref->{$span_term}++;
940 #Build branchnames hash
942 #get branch information.....
945 $dbh->prepare("SELECT branchcode,branchname FROM branches")
946 ; # FIXME : use C4::Koha::GetBranches
948 while ( my $bdata = $bsth->fetchrow_hashref ) {
949 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
953 #find itemtype & itemtype image
956 $dbh->prepare("SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes");
958 while ( my $bdata = $bsth->fetchrow_hashref ) {
959 $itemtypes{ $bdata->{'itemtype'} }->{description} =
960 $bdata->{'description'};
961 $itemtypes{ $bdata->{'itemtype'} }->{imageurl} = $bdata->{'imageurl'};
962 $itemtypes{ $bdata->{'itemtype'} }->{summary} = $bdata->{'summary'};
963 $itemtypes{ $bdata->{'itemtype'} }->{notforloan} = $bdata->{'notforloan'};
966 #search item field code
967 my $sth = $dbh->prepare("SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'");
969 my ($itemtag) = $sth->fetchrow;
971 ## find column names of items related to MARC
972 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
974 my %subfieldstosearch;
975 while ( ( my $column ) = $sth2->fetchrow ) {
976 my ( $tagfield, $tagsubfield ) =
977 &GetMarcFromKohaField( "items." . $column, "" );
978 $subfieldstosearch{$column} = $tagsubfield;
982 if ( $hits && $offset + $results_per_page <= $hits ) {
983 $times = $offset + $results_per_page;
989 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
991 $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
992 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, '' );
993 $oldbiblio->{result_number} = $i+1;
994 # add image url if there is one
995 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} =~ /^http:/ ) {
996 $oldbiblio->{imageurl} =
997 $itemtypes{ $oldbiblio->{itemtype} }->{imageurl};
998 $oldbiblio->{description} =
999 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1002 $oldbiblio->{imageurl} =
1003 getitemtypeimagesrc() . "/"
1004 . $itemtypes{ $oldbiblio->{itemtype} }->{imageurl}
1005 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1006 $oldbiblio->{description} =
1007 $itemtypes{ $oldbiblio->{itemtype} }->{description};
1010 # build summary if there is one (the summary is defined in itemtypes table
1012 if ($itemtypes{ $oldbiblio->{itemtype} }->{summary}) {
1013 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1014 my @fields = $marcrecord->fields();
1015 foreach my $field (@fields) {
1016 my $tag = $field->tag();
1017 my $tagvalue = $field->as_string();
1018 $summary =~ s/\[(.?.?.?.?)$tag\*(.*?)]/$1$tagvalue$2\[$1$tag$2]/g;
1020 my @subf = $field->subfields;
1021 for my $i (0..$#subf) {
1022 my $subfieldcode = $subf[$i][0];
1023 my $subfieldvalue = $subf[$i][1];
1024 my $tagsubf = $tag.$subfieldcode;
1025 $summary =~ s/\[(.?.?.?.?)$tagsubf(.*?)]/$1$subfieldvalue$2\[$1$tagsubf$2]/g;
1029 $summary =~ s/\[(.*?)]//g;
1030 $summary =~ s/\n/<br>/g;
1031 $oldbiblio->{summary} = $summary;
1033 # add spans to search term in results for search term highlighting
1034 # save a native author, for the <a href=search.lq=<!--tmpl_var name="author"-->> link
1035 $oldbiblio->{'author_nospan'} = $oldbiblio->{'author'};
1036 foreach my $term ( keys %$span_terms_hashref ) {
1037 my $old_term = $term;
1038 if ( length($term) > 3 ) {
1039 $term =~ s/(.*=|\)|\(|\+|\.|\?|\[|\])//g;
1043 #FIXME: is there a better way to do this?
1044 $oldbiblio->{'title'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1045 $oldbiblio->{'subtitle'} =~
1046 s/$term/<span class=\"term\">$&<\/span>/gi;
1048 $oldbiblio->{'author'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1049 $oldbiblio->{'publishercode'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1050 $oldbiblio->{'place'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1051 $oldbiblio->{'pages'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1052 $oldbiblio->{'notes'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1053 $oldbiblio->{'size'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1058 $toggle = "#ffffcc";
1063 $oldbiblio->{'toggle'} = $toggle;
1064 my @fields = $marcrecord->field($itemtag);
1067 my $ordered_count = 0;
1068 my $onloan_count = 0;
1069 my $wthdrawn_count = 0;
1070 my $itemlost_count = 0;
1074 # check the loan status of the item :
1075 # it is not stored in the MARC record, for pref (zebra reindexing)
1076 # reason. Thus, we have to get the status from a specific SQL query
1078 my $sth_issue = $dbh->prepare("
1079 SELECT date_due,returndate
1081 WHERE itemnumber=? AND returndate IS NULL");
1082 my $items_count=scalar(@fields);
1083 foreach my $field (@fields) {
1085 foreach my $code ( keys %subfieldstosearch ) {
1086 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1088 $sth_issue->execute($item->{itemnumber});
1089 $item->{due_date} = format_date($sth_issue->fetchrow) if $sth_issue->fetchrow;
1090 $item->{onloan} = 1 if $item->{due_date};
1091 # at least one item can be reserved : suppose no
1093 if ( $item->{wthdrawn} ) {
1095 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1096 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{wthdrawn}=1;
1098 elsif ( $item->{itemlost} ) {
1100 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1101 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemlost}=1;
1103 unless ( $item->{notforloan}) {
1104 # OK, this one can be issued, so at least one can be reserved
1107 if ( ( $item->{onloan} ) && ( $item->{onloan} != '0000-00-00' ) )
1109 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1110 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{onloancount} = 1;
1111 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{due_date} = $item->{due_date};
1114 if ( $item->{'homebranch'} ) {
1115 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{count}++;
1119 elsif ( $item->{'holdingbranch'} ) {
1120 $items->{ $item->{'holdingbranch'} }->{count}++;
1122 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemcallnumber} = $item->{itemcallnumber};
1123 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{location} = $item->{location};
1124 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{branchcode} = $item->{homebranch};
1125 } # notforloan, item level and biblioitem level
1127 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1128 $norequests = 1 if $itemtypes{$oldbiblio->{itemtype}}->{notforloan};
1130 for my $key ( sort keys %$items ) {
1133 branchname => $branches{$items->{$key}->{branchcode}},
1134 branchcode => $items->{$key}->{branchcode},
1135 count => $items->{$key}->{count},
1136 itemcallnumber => $items->{$key}->{itemcallnumber},
1137 location => $items->{$key}->{location},
1138 onloancount => $items->{$key}->{onloancount},
1139 due_date => $items->{$key}->{due_date},
1140 wthdrawn => $items->{$key}->{wthdrawn},
1141 lost => $items->{$key}->{itemlost},
1143 # only show the number specified by the user
1144 my $maxitems = (C4::Context->preference('maxItemsinSearchResults')) ? C4::Context->preference('maxItemsinSearchResults')- 1 : 1;
1145 push @items_loop, $this_item unless $itemscount > $maxitems;;
1147 $oldbiblio->{norequests} = $norequests;
1148 $oldbiblio->{items_count} = $items_count;
1149 $oldbiblio->{items_loop} = \@items_loop;
1150 $oldbiblio->{onloancount} = $onloan_count;
1151 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1152 $oldbiblio->{itemlostcount} = $itemlost_count;
1153 $oldbiblio->{orderedcount} = $ordered_count;
1154 $oldbiblio->{isbn} =~ s/-//g; # deleting - in isbn to enable amazon content
1155 push( @newresults, $oldbiblio );
1162 #----------------------------------------------------------------------
1164 # Non-Zebra GetRecords#
1165 #----------------------------------------------------------------------
1169 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1173 my ($query,$simple_query,$sort_by_ref,$servers_ref,$results_per_page,$offset,$expanded_facet,$branches,$query_type,$scan) = @_;
1174 my $result = NZanalyse($query);
1175 return (undef,NZorder($result,@$sort_by_ref[0],$results_per_page,$offset),undef);
1180 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1181 the list is built from an inverted index in the nozebra SQL table
1182 note that title is here only for convenience : the sorting will be very fast when requested on title
1183 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1188 my ($string,$server) = @_;
1189 # $server contains biblioserver or authorities, depending on what we search on.
1190 #warn "querying : $string on $server";
1191 $server='biblioserver' unless $server;
1193 # if we have a ", replace the content to discard temporarily any and/or/not inside
1195 if ($string =~/"/) {
1196 $string =~ s/"(.*?)"/__X__/;
1198 warn "commacontent : $commacontent" if $DEBUG;
1200 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1201 # then, call again NZanalyse with $left and $right
1202 # (recursive until we find a leaf (=> something without and/or/not)
1203 # delete repeated operator... Would then go in infinite loop
1204 while ($string =~s/( and| or| not| AND| OR| NOT)\1/$1/g){
1206 #process parenthesis before.
1207 if ($string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/){
1209 # warn "left :".$left;
1211 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1212 my $leftresult = NZanalyse($left,$server);
1214 my $rightresult = NZanalyse($right,$server);
1215 # OK, we have the results for right and left part of the query
1216 # depending of operand, intersect, union or exclude both lists
1217 # to get a result list
1218 if ($operator eq ' and ') {
1219 my @leftresult = split /;/, $leftresult;
1220 # my @rightresult = split /;/,$leftresult;
1222 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1223 # the result is stored twice, to have the same weight for AND than OR.
1224 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1225 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1226 foreach (@leftresult) {
1227 if ($rightresult =~ "$_;") {
1228 $finalresult .= "$_;$_;";
1231 return $finalresult;
1232 } elsif ($operator eq ' or ') {
1233 # just merge the 2 strings
1234 return $leftresult.$rightresult;
1235 } elsif ($operator eq ' not ') {
1236 my @leftresult = split /;/, $leftresult;
1237 # my @rightresult = split /;/,$leftresult;
1239 foreach (@leftresult) {
1240 unless ($rightresult =~ "$_;") {
1241 $finalresult .= "$_;";
1244 return $finalresult;
1246 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1252 warn "string :".$string if $DEBUG;
1253 $string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/;
1256 my $operand = lc($2); # FIXME: and/or/not are operators, not operands
1257 # it's not a leaf, we have a and/or/not
1259 # reintroduce comma content if needed
1260 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1261 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1262 warn "node : $left / $operand / $right\n" if $DEBUG;
1263 my $leftresult = NZanalyse($left,$server);
1264 my $rightresult = NZanalyse($right,$server);
1265 # OK, we have the results for right and left part of the query
1266 # depending of operand, intersect, union or exclude both lists
1267 # to get a result list
1268 if ($operand eq ' and ') {
1269 my @leftresult = split /;/, $leftresult;
1270 # my @rightresult = split /;/,$leftresult;
1272 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1273 # the result is stored twice, to have the same weight for AND than OR.
1274 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1275 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1276 foreach (@leftresult) {
1277 if ($rightresult =~ "$_;") {
1278 $finalresult .= "$_;$_;";
1281 return $finalresult;
1282 } elsif ($operand eq ' or ') {
1283 # just merge the 2 strings
1284 return $leftresult.$rightresult;
1285 } elsif ($operand eq ' not ') {
1286 my @leftresult = split /;/, $leftresult;
1287 # my @rightresult = split /;/,$leftresult;
1289 foreach (@leftresult) {
1290 unless ($rightresult =~ "$_;") {
1291 $finalresult .= "$_;";
1294 return $finalresult;
1296 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1297 die "error : operand unknown : $operand for $string";
1299 # it's a leaf, do the real SQL query and return the result
1301 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1302 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1303 warn "leaf:$string" if $DEBUG;
1304 # parse the string in in operator/operand/value again
1305 $string =~ /(.*)(>=|<=)(.*)/;
1309 unless ($operator) {
1310 $string =~ /(.*)(>|<|=)(.*)/;
1316 # automatic replace for short operators
1317 $left='title' if $left =~ '^ti$';
1318 $left='author' if $left =~ '^au$';
1319 $left='publisher' if $left =~ '^pb$';
1320 $left='subject' if $left =~ '^su$';
1321 $left='koha-Auth-Number' if $left =~ '^an$';
1322 $left='keyword' if $left =~ '^kw$';
1323 if ($operator && $left ne 'keyword' ) {
1324 # do a specific search
1325 my $dbh = C4::Context->dbh;
1326 $operator='LIKE' if $operator eq '=' and $right=~ /%/;
1327 my $sth = $dbh->prepare("SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?");
1328 warn "$left / $operator / $right\n";
1329 # split each word, query the DB and build the biblionumbers result
1330 #sanitizing leftpart
1331 $left=~s/^\s+|\s+$//;
1332 my ($biblionumbers,$value);
1333 foreach (split / /,$right) {
1335 warn "EXECUTE : $server, $left, $_";
1336 $sth->execute($server, $left, $_) or warn "execute failed: $!";
1337 while (my ($line,$value) = $sth->fetchrow) {
1338 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1339 # otherwise, fill the result
1340 $biblionumbers .= $line unless ($right =~ /\d/ && $value =~ /\D/);
1341 # warn "result : $value ". ($right =~ /\d/) . "==".(!$value =~ /\d/) ;#= $line";
1343 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1345 my @leftresult = split /;/, $biblionumbers;
1347 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1348 # remove weight at the end
1349 my $cleaned = $entry;
1350 $cleaned =~ s/-\d*$//;
1351 # if the entry already in the hash, take it & increase weight
1352 warn "===== $cleaned =====" if $DEBUG;
1353 if ($results =~ "$cleaned") {
1354 $temp .= "$entry;$entry;";
1355 warn "INCLUDING $entry" if $DEBUG;
1360 $results = $biblionumbers;
1364 #do a complete search (all indexes)
1365 my $dbh = C4::Context->dbh;
1366 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?");
1367 # split each word, query the DB and build the biblionumbers result
1368 foreach (split / /,$string) {
1369 next if C4::Context->stopwords->{uc($_)}; # skip if stopword
1370 warn "search on all indexes on $_" if $DEBUG;
1373 $sth->execute($server, $_);
1374 while (my $line = $sth->fetchrow) {
1375 $biblionumbers .= $line;
1377 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1379 warn "RES for $_ = $biblionumbers" if $DEBUG;
1380 my @leftresult = split /;/, $biblionumbers;
1382 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1383 # remove weight at the end
1384 my $cleaned = $entry;
1385 $cleaned =~ s/-\d*$//;
1386 # if the entry already in the hash, take it & increase weight
1387 warn "===== $cleaned =====" if $DEBUG;
1388 if ($results =~ "$cleaned") {
1389 $temp .= "$entry;$entry;";
1390 warn "INCLUDING $entry" if $DEBUG;
1395 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
1396 $results = $biblionumbers;
1400 # warn "return : $results for LEAF : $string" if $DEBUG;
1407 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
1415 my ($biblionumbers, $ordering,$results_per_page,$offset) = @_;
1416 # order title asc by default
1417 # $ordering = '1=36 <i' unless $ordering;
1418 $results_per_page=20 unless $results_per_page;
1419 $offset = 0 unless $offset;
1420 my $dbh = C4::Context->dbh;
1422 # order by POPULARITY
1424 if ($ordering =~ /popularity/) {
1427 # popularity is not in MARC record, it's builded from a specific query
1428 my $sth = $dbh->prepare("select sum(issues) from items where biblionumber=?");
1429 foreach (split /;/,$biblionumbers) {
1430 my ($biblionumber,$title) = split /,/,$_;
1431 $result{$biblionumber}=GetMarcBiblio($biblionumber);
1432 $sth->execute($biblionumber);
1433 my $popularity= $sth->fetchrow ||0;
1434 # hint : the key is popularity.title because we can have
1435 # many results with the same popularity. In this cas, sub-ordering is done by title
1436 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1437 # (un-frequent, I agree, but we won't forget anything that way ;-)
1438 $popularity{sprintf("%10d",$popularity).$title.$biblionumber} = $biblionumber;
1440 # sort the hash and return the same structure as GetRecords (Zebra querying)
1443 if ($ordering eq 'popularity_dsc') { # sort popularity DESC
1444 foreach my $key (sort {$b cmp $a} (keys %popularity)) {
1445 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1447 } else { # sort popularity ASC
1448 foreach my $key (sort (keys %popularity)) {
1449 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1453 $result_hash->{'hits'} = $numbers;
1454 $finalresult->{'biblioserver'} = $result_hash;
1455 return $finalresult;
1459 } elsif ($ordering =~/author/){
1461 foreach (split /;/,$biblionumbers) {
1462 my ($biblionumber,$title) = split /,/,$_;
1463 my $record=GetMarcBiblio($biblionumber);
1465 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1466 $author=$record->subfield('200','f');
1467 $author=$record->subfield('700','a') unless $author;
1469 $author=$record->subfield('100','a');
1471 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1472 # and we don't want to get only 1 result for each of them !!!
1473 $result{$author.$biblionumber}=$record;
1475 # sort the hash and return the same structure as GetRecords (Zebra querying)
1478 if ($ordering eq 'author_za') { # sort by author desc
1479 foreach my $key (sort { $b cmp $a } (keys %result)) {
1480 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1482 } else { # sort by author ASC
1483 foreach my $key (sort (keys %result)) {
1484 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1488 $result_hash->{'hits'} = $numbers;
1489 $finalresult->{'biblioserver'} = $result_hash;
1490 return $finalresult;
1492 # ORDER BY callnumber
1494 } elsif ($ordering =~/callnumber/){
1496 foreach (split /;/,$biblionumbers) {
1497 my ($biblionumber,$title) = split /,/,$_;
1498 my $record=GetMarcBiblio($biblionumber);
1500 my ($callnumber_tag,$callnumber_subfield)=GetMarcFromKohaField($dbh,'items.itemcallnumber');
1501 ($callnumber_tag,$callnumber_subfield)= GetMarcFromKohaField('biblioitems.callnumber') unless $callnumber_tag;
1502 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1503 $callnumber=$record->subfield('200','f');
1505 $callnumber=$record->subfield('100','a');
1507 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1508 # and we don't want to get only 1 result for each of them !!!
1509 $result{$callnumber.$biblionumber}=$record;
1511 # sort the hash and return the same structure as GetRecords (Zebra querying)
1514 if ($ordering eq 'call_number_dsc') { # sort by title desc
1515 foreach my $key (sort { $b cmp $a } (keys %result)) {
1516 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1518 } else { # sort by title ASC
1519 foreach my $key (sort { $a cmp $b } (keys %result)) {
1520 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1524 $result_hash->{'hits'} = $numbers;
1525 $finalresult->{'biblioserver'} = $result_hash;
1526 return $finalresult;
1527 } elsif ($ordering =~ /pubdate/){ #pub year
1529 foreach (split /;/,$biblionumbers) {
1530 my ($biblionumber,$title) = split /,/,$_;
1531 my $record=GetMarcBiblio($biblionumber);
1532 my ($publicationyear_tag,$publicationyear_subfield)=GetMarcFromKohaField('biblioitems.publicationyear','');
1533 my $publicationyear=$record->subfield($publicationyear_tag,$publicationyear_subfield);
1534 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1535 # and we don't want to get only 1 result for each of them !!!
1536 $result{$publicationyear.$biblionumber}=$record;
1538 # sort the hash and return the same structure as GetRecords (Zebra querying)
1541 if ($ordering eq 'pubdate_dsc') { # sort by pubyear desc
1542 foreach my $key (sort { $b cmp $a } (keys %result)) {
1543 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1545 } else { # sort by pub year ASC
1546 foreach my $key (sort (keys %result)) {
1547 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1551 $result_hash->{'hits'} = $numbers;
1552 $finalresult->{'biblioserver'} = $result_hash;
1553 return $finalresult;
1557 } elsif ($ordering =~ /title/) {
1558 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
1560 foreach (split /;/,$biblionumbers) {
1561 my ($biblionumber,$title) = split /,/,$_;
1562 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1563 # and we don't want to get only 1 result for each of them !!!
1564 # hint & speed improvement : we can order without reading the record
1565 # so order, and read records only for the requested page !
1566 $result{$title.$biblionumber}=$biblionumber;
1568 # sort the hash and return the same structure as GetRecords (Zebra querying)
1571 if ($ordering eq 'title_az') { # sort by title desc
1572 foreach my $key (sort (keys %result)) {
1573 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1575 } else { # sort by title ASC
1576 foreach my $key (sort { $b cmp $a } (keys %result)) {
1577 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1580 # limit the $results_per_page to result size if it's more
1581 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1582 # for the requested page, replace biblionumber by the complete record
1583 # speed improvement : avoid reading too much things
1584 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1585 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1588 $result_hash->{'hits'} = $numbers;
1589 $finalresult->{'biblioserver'} = $result_hash;
1590 return $finalresult;
1595 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
1598 foreach (split /;/,$biblionumbers) {
1599 my ($biblionumber,$title) = split /,/,$_;
1600 $title =~ /(.*)-(\d)/;
1603 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
1604 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
1605 # biblio N has ranking = 6
1606 $count_ranking{$biblionumber} += $ranking;
1608 # build the result by "inverting" the count_ranking hash
1609 # 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
1611 foreach (keys %count_ranking) {
1612 $result{sprintf("%10d",$count_ranking{$_}).'-'.$_} = $_;
1614 # sort the hash and return the same structure as GetRecords (Zebra querying)
1617 foreach my $key (sort {$b cmp $a} (keys %result)) {
1618 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1620 # limit the $results_per_page to result size if it's more
1621 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1622 # for the requested page, replace biblionumber by the complete record
1623 # speed improvement : avoid reading too much things
1624 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1625 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc if $result_hash->{'RECORDS'}[$counter];
1628 $result_hash->{'hits'} = $numbers;
1629 $finalresult->{'biblioserver'} = $result_hash;
1630 return $finalresult;
1635 ($countchanged,$listunchanged) = ModBiblios($listbiblios, $tagsubfield,$initvalue,$targetvalue,$test);
1637 this function changes all the values $initvalue in subfield $tag$subfield in any record in $listbiblios
1638 test parameter if set donot perform change to records in database.
1644 * $listbiblios is an array ref to marcrecords to be changed
1645 * $tagsubfield is the reference of the subfield to change.
1646 * $initvalue is the value to search the record for
1647 * $targetvalue is the value to set the subfield to
1648 * $test is to be set only not to perform changes in database.
1650 =item C<Output arg:>
1651 * $countchanged counts all the changes performed.
1652 * $listunchanged contains the list of all the biblionumbers of records unchanged.
1654 =item C<usage in the script:>
1658 my ($countchanged, $listunchanged) = EditBiblios($results->{RECORD}, $tagsubfield,$initvalue,$targetvalue);;
1659 #If one wants to display unchanged records, you should get biblios foreach @$listunchanged
1660 $template->param(countchanged => $countchanged, loopunchanged=>$listunchanged);
1665 my ($listbiblios,$tagsubfield,$initvalue,$targetvalue,$test)=@_;
1668 my ($tag,$subfield)=($1,$2) if ($tagsubfield=~/^(\d{1,3})([a-z0-9A-Z@])?$/);
1669 if ((length($tag)<3)&& $subfield=~/0-9/){
1670 $tag=$tag.$subfield;
1673 my ($bntag,$bnsubf) = GetMarcFromKohaField('biblio.biblionumber');
1674 my ($itemtag,$itemsubf) = GetMarcFromKohaField('items.itemnumber');
1675 foreach my $usmarc (@$listbiblios){
1677 $record=eval{MARC::Record->new_from_usmarc($usmarc)};
1680 # usmarc is not a valid usmarc May be a biblionumber
1681 if ($tag eq $itemtag){
1682 my $bib=GetBiblioFromItemNumber($usmarc);
1683 $record=GetMarcItem($bib->{'biblionumber'},$usmarc) ;
1684 $biblionumber=$bib->{'biblionumber'};
1686 $record=GetMarcBiblio($usmarc);
1687 $biblionumber=$usmarc;
1691 $biblionumber = $record->subfield($bntag,$bnsubf);
1693 $biblionumber=$record->field($bntag)->data;
1696 #GetBiblionumber is to be written.
1697 #Could be replaced by TransformMarcToKoha (But Would be longer)
1698 if ($record->field($tag)){
1700 foreach my $field ($record->field($tag)){
1702 if ($field->delete_subfield('code' =>$subfield,'match'=>qr($initvalue))){
1705 $field->update($subfield,$targetvalue) if ($targetvalue);
1709 if ($field->delete_field($field)){
1714 $field->data=$targetvalue if ($field->data=~qr($initvalue));
1718 # warn $record->as_formatted;
1720 ModBiblio($record,$biblionumber,GetFrameworkCode($biblionumber)) unless ($test);
1722 push @unmatched, $biblionumber;
1725 push @unmatched, $biblionumber;
1728 return ($countmatched,\@unmatched);
1731 END { } # module clean-up code here (global destructor)
1738 Koha Developement team <info@koha.org>