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
26 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);
28 # set the version for version checking
34 C4::Search - Functions for searching the Koha catalog.
38 see opac/opac-search.pl or catalogue/search.pl for example of usage
42 This module provides the searching facilities for the Koha into a zebra catalog.
60 # make all your functions, whether exported or not;
62 =head2 findseealso($dbh,$fields);
64 C<$dbh> is a link to the DB handler.
67 my $dbh =C4::Context->dbh;
69 C<$fields> is a reference to the fields array
71 This function modify the @$fields array and add related fields to search on.
76 my ( $dbh, $fields ) = @_;
77 my $tagslib = GetMarcStructure( 1 );
78 for ( my $i = 0 ; $i <= $#{$fields} ; $i++ ) {
79 my ($tag) = substr( @$fields[$i], 1, 3 );
80 my ($subfield) = substr( @$fields[$i], 4, 1 );
81 @$fields[$i] .= ',' . $tagslib->{$tag}->{$subfield}->{seealso}
82 if ( $tagslib->{$tag}->{$subfield}->{seealso} );
88 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
94 my $dbh = C4::Context->dbh;
95 my $result = TransformMarcToKoha( $dbh, $record, '' );
100 my ( $biblionumber, $title );
102 # search duplicate on ISBN, easy and fast..
103 # ... normalize first
104 if ( $result->{isbn} ) {
105 $result->{isbn} =~ s/\(.*$//;
106 $result->{isbn} =~ s/\s+$//;
108 #$search->{'avoidquerylog'}=1;
109 if ( $result->{isbn} ) {
110 $query = "isbn=$result->{isbn}";
113 $result->{title} =~ s /\\//g;
114 $result->{title} =~ s /\"//g;
115 $result->{title} =~ s /\(//g;
116 $result->{title} =~ s /\)//g;
117 # remove valid operators
118 $result->{title} =~ s/(and|or|not)//g;
119 $query = "ti,ext=$result->{title}";
120 $query .= " and itemtype=$result->{itemtype}" if ($result->{itemtype});
121 if ($result->{author}){
122 $result->{author} =~ s /\\//g;
123 $result->{author} =~ s /\"//g;
124 $result->{author} =~ s /\(//g;
125 $result->{author} =~ s /\)//g;
126 # remove valid operators
127 $result->{author} =~ s/(and|or|not)//g;
128 $query .= " and au,ext=$result->{author}";
131 my ($error,$searchresults) =
132 SimpleSearch($query); # FIXME :: hardcoded !
134 foreach my $possible_duplicate_record (@$searchresults) {
136 MARC::Record->new_from_usmarc($possible_duplicate_record);
137 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
139 # FIXME :: why 2 $biblionumber ?
141 push @results, $result->{'biblionumber'};
142 push @results, $result->{'title'};
150 ($error,$results) = SimpleSearch($query,@servers);
152 this function performs a simple search on the catalog using zoom.
158 * $query could be a simple keyword or a complete CCL query wich is depending on your ccl file.
159 * @servers is optionnal. default one is read on koha.xml
162 * $error is a string which containt the description error if there is one. Else it's empty.
163 * \@results is an array of marc record.
165 =item C<usage in the script:>
169 my ($error, $marcresults) = SimpleSearch($query);
171 if (defined $error) {
172 $template->param(query_error => $error);
173 warn "error: ".$error;
174 output_html_with_http_headers $input, $cookie, $template->output;
178 my $hits = scalar @$marcresults;
181 for(my $i=0;$i<$hits;$i++) {
183 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
184 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
186 #build the hash for the template.
187 $resultsloop{highlight} = ($i % 2)?(1):(0);
188 $resultsloop{title} = $biblio->{'title'};
189 $resultsloop{subtitle} = $biblio->{'subtitle'};
190 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
191 $resultsloop{author} = $biblio->{'author'};
192 $resultsloop{publishercode} = $biblio->{'publishercode'};
193 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
195 push @results, \%resultsloop;
197 $template->param(result=>\@results);
203 if (C4::Context->preference('NoZebra')) {
204 my $result = NZorder(NZanalyse($query))->{'biblioserver'}->{'RECORDS'};
205 return (undef,$result);
211 return ( "No query entered", undef ) unless $query;
213 #@servers = (C4::Context->config("biblioserver")) unless @servers;
215 ("biblioserver") unless @servers
216 ; # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
219 for ( my $i = 0 ; $i < @servers ; $i++ ) {
220 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
223 ->search( new ZOOM::Query::CCL2RPN( $query, $zconns[$i] ) );
225 # getting error message if one occured.
227 $zconns[$i]->errmsg() . " ("
228 . $zconns[$i]->errcode() . ") "
229 . $zconns[$i]->addinfo() . " "
230 . $zconns[$i]->diagset();
232 return ( $error, undef ) if $zconns[$i]->errcode();
236 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
237 $ev = $zconns[ $i - 1 ]->last_event();
238 if ( $ev == ZOOM::Event::ZEND ) {
239 $hits = $tmpresults[ $i - 1 ]->size();
242 for ( my $j = 0 ; $j < $hits ; $j++ ) {
243 my $record = $tmpresults[ $i - 1 ]->record($j)->raw();
244 push @results, $record;
248 return ( undef, \@results );
252 # performs the search
255 $koha_query, $simple_query, $sort_by_ref,
256 $servers_ref, $results_per_page, $offset,
257 $expanded_facet, $branches, $query_type,
260 # warn "Query : $koha_query";
261 my @servers = @$servers_ref;
262 my @sort_by = @$sort_by_ref;
264 # create the zoom connection and query object
268 my $results_hashref = ();
271 my $facets_counter = ();
272 my $facets_info = ();
273 my $facets = getFacets();
275 #### INITIALIZE SOME VARS USED CREATE THE FACETED RESULTS
276 my @facets_loop; # stores the ref to array of hashes for template
277 for ( my $i = 0 ; $i < @servers ; $i++ ) {
278 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
280 # perform the search, create the results objects
281 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
283 if ( $servers[$i] =~ /biblioserver/ ) {
284 $query_to_use = $koha_query;
287 $query_to_use = $simple_query;
290 #$query_to_use = $simple_query if $scan;
291 #warn $simple_query if ($scan && $DEBUG);
292 # check if we've got a query_type defined
296 if ( $query_type =~ /^ccl/ ) {
298 s/\:/\=/g; # change : to = last minute (FIXME)
300 # warn "CCL : $query_to_use";
303 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
306 elsif ( $query_type =~ /^cql/ ) {
308 # warn "CQL : $query_to_use";
311 new ZOOM::Query::CQL( $query_to_use, $zconns[$i] ) );
313 elsif ( $query_type =~ /^pqf/ ) {
315 # warn "PQF : $query_to_use";
318 new ZOOM::Query::PQF( $query_to_use, $zconns[$i] ) );
323 # warn "preparing to scan:$query_to_use";
326 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
330 # warn "LAST : $query_to_use";
333 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
339 warn "WARNING: query problem with $query_to_use " . $@;
342 # concatenate the sort_by limits and pass them to the results object
344 foreach my $sort (@sort_by) {
345 if ($sort eq "author_az") {
346 $sort_by.="1=1003 <i ";
348 elsif ($sort eq "author_za") {
349 $sort_by.="1=1003 >i ";
351 elsif ($sort eq "popularity_asc") {
352 $sort_by.="1=9003 <i ";
354 elsif ($sort eq "popularity_dsc") {
355 $sort_by.="1=9003 >i ";
357 elsif ($sort eq "call_number_asc") {
358 $sort_by.="1=20 <i ";
360 elsif ($sort eq "call_number_dsc") {
361 $sort_by.="1=20 >i ";
363 elsif ($sort eq "pubdate_asc") {
364 $sort_by.="1=31 <i ";
366 elsif ($sort eq "pubdate_dsc") {
367 $sort_by.="1=31 >i ";
369 elsif ($sort eq "acqdate_asc") {
370 $sort_by.="1=32 <i ";
372 elsif ($sort eq "acqdate_dsc") {
373 $sort_by.="1=32 >i ";
375 elsif ($sort eq "title_az") {
378 elsif ($sort eq "title_za") {
383 if ( $results[$i]->sort( "yaz", $sort_by ) < 0) {
384 warn "WARNING sort $sort_by failed";
388 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
389 my $ev = $zconns[ $i - 1 ]->last_event();
390 if ( $ev == ZOOM::Event::ZEND ) {
391 my $size = $results[ $i - 1 ]->size();
394 #$results_hash->{'server'} = $servers[$i-1];
395 # loop through the results
396 $results_hash->{'hits'} = $size;
398 if ( $offset + $results_per_page <= $size ) {
399 $times = $offset + $results_per_page;
404 for ( my $j = $offset ; $j < $times ; $j++ )
405 { #(($offset+$count<=$size) ? ($offset+$count):$size) ; $j++){
409 ## This is just an index scan
411 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
412 # here we create a minimal MARC record and hand it off to the
413 # template just like a normal result ... perhaps not ideal, but
415 my $tmprecord = MARC::Record->new();
416 $tmprecord->encoding('UTF-8');
419 # srote the minimal record in author/title (depending on MARC flavour)
420 if ( C4::Context->preference("marcflavour") eq
423 $tmptitle = MARC::Field->new(
430 $tmptitle = MARC::Field->new(
436 $tmprecord->append_fields($tmptitle);
437 $results_hash->{'RECORDS'}[$j] =
438 $tmprecord->as_usmarc();
441 $record = $results[ $i - 1 ]->record($j)->raw();
443 #warn "RECORD $j:".$record;
444 $results_hash->{'RECORDS'}[$j] =
445 $record; # making a reference to a hash
446 # Fill the facets while we're looping
447 $facet_record = MARC::Record->new_from_usmarc($record);
449 #warn $servers[$i-1].$facet_record->title();
450 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
451 if ( $facets->[$k] ) {
453 for my $tag ( @{ $facets->[$k]->{'tags'} } ) {
454 push @fields, $facet_record->field($tag);
456 for my $field (@fields) {
457 my @subfields = $field->subfields();
458 for my $subfield (@subfields) {
459 my ( $code, $data ) = @$subfield;
461 $facets->[$k]->{'subfield'} )
463 $facets_counter->{ $facets->[$k]
464 ->{'link_value'} }->{$data}++;
468 $facets_info->{ $facets->[$k]->{'link_value'} }
470 $facets->[$k]->{'label_value'};
471 $facets_info->{ $facets->[$k]->{'link_value'} }
472 ->{'expanded'} = $facets->[$k]->{'expanded'};
477 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
480 #print "connection ", $i-1, ": $size hits";
481 #print $results[$i-1]->record(0)->render() if $size > 0;
484 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
485 keys %$facets_counter
489 my $number_of_facets;
490 my @this_facets_array;
493 $facets_counter->{$link_value}
494 ->{$b} <=> $facets_counter->{$link_value}->{$a}
495 } keys %{ $facets_counter->{$link_value} }
499 if ( ( $number_of_facets < 6 )
500 || ( $expanded_facet eq $link_value )
501 || ( $facets_info->{$link_value}->{'expanded'} ) )
504 # sanitize the link value ), ( will cause errors with CCL
505 my $facet_link_value = $one_facet;
506 $facet_link_value =~ s/(\(|\))/ /g;
508 # fix the length that will display in the label
509 my $facet_label_value = $one_facet;
510 $facet_label_value = substr( $one_facet, 0, 20 ) . "..."
511 unless length($facet_label_value) <= 20;
513 # well, if it's a branch, label by the name, not the code
514 if ( $link_value =~ /branch/ ) {
516 $branches->{$one_facet}->{'branchname'};
519 # but we're down with the whole label being in the link's title
520 my $facet_title_value = $one_facet;
522 push @this_facets_array,
526 $facets_counter->{$link_value}->{$one_facet},
527 facet_label_value => $facet_label_value,
528 facet_title_value => $facet_title_value,
529 facet_link_value => $facet_link_value,
530 type_link_value => $link_value,
535 unless ( $facets_info->{$link_value}->{'expanded'} ) {
537 if ( ( $number_of_facets > 6 )
538 && ( $expanded_facet ne $link_value ) );
543 type_link_value => $link_value,
544 type_id => $link_value . "_id",
546 $facets_info->{$link_value}->{'label_value'},
547 facets => \@this_facets_array,
548 expandable => $expandable,
549 expand => $link_value,
555 return ( undef, $results_hashref, \@facets_loop );
559 sub _remove_stopwords {
560 my ($operand,$index) = @_;
561 my @stopwords_removed;
562 # phrase and exact-qualified indexes shouldn't have stopwords removed
563 if ($index!~m/phr|ext/){
564 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
565 # we use IsAlpha unicode definition, to deal correctly with diacritics.
566 # otherwise, a French word like "leçon" woudl be split into "le" "çon", le
567 # is an empty word, we'd get "çon" and wouldn't find anything...
568 foreach (keys %{C4::Context->stopwords}) {
569 next if ($_ =~/(and|or|not)/); # don't remove operators
570 if ($operand =~ /(\P{IsAlpha}$_\P{IsAlpha}|^$_\P{IsAlpha}|\P{IsAlpha}$_$)/) {
571 $operand=~ s/\P{IsAlpha}$_\P{IsAlpha}/ /gi;
572 $operand=~ s/^$_\P{IsAlpha}/ /gi;
573 $operand=~ s/\P{IsAlpha}$_$/ /gi;
574 push @stopwords_removed, $_;
578 return ($operand, \@stopwords_removed);
582 sub _detect_truncation {
583 my ($operand,$index) = @_;
584 my (@nontruncated,@righttruncated,@lefttruncated,@rightlefttruncated,@regexpr);
586 my @wordlist= split (/\s/,$operand);
587 foreach my $word (@wordlist){
588 if ($word=~s/^\*([^\*]+)\*$/$1/){
589 push @rightlefttruncated,$word;
591 elsif($word=~s/^\*([^\*]+)$/$1/){
592 push @lefttruncated,$word;
594 elsif ($word=~s/^([^\*]+)\*$/$1/){
595 push @righttruncated,$word;
597 elsif (index($word,"*")<0){
598 push @nontruncated,$word;
604 return (\@nontruncated,\@righttruncated,\@lefttruncated,\@rightlefttruncated,\@regexpr);
607 sub _build_stemmed_operand {
610 # FIXME: the locale should be set based on the user's language and/or search choice
611 my $stemmer = Lingua::Stem->new( -locale => 'EN-US' );
612 # FIXME: these should be stored in the db so the librarian can modify the behavior
613 $stemmer->add_exceptions(
621 my @words = split( / /, $operand );
622 my $stems = $stemmer->stem(@words);
623 for my $stem (@$stems) {
624 $stemmed_operand .= "$stem";
625 $stemmed_operand .= "?" unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
626 $stemmed_operand .= " ";
628 #warn "STEMMED OPERAND: $stemmed_operand";
629 return $stemmed_operand;
632 sub _build_weighted_query {
633 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
634 # pretty well but will work much better when we have an actual query parser
635 my ($operand,$stemmed_operand,$index) = @_;
636 my $stemming = C4::Context->preference("QueryStemming") || 0;
637 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
638 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
640 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
642 # Keyword, or, no index specified
643 if ( ( $index eq 'kw' ) || ( !$index ) ) {
644 $weighted_query .= "Title-cover,ext,r1=\"$operand\""; # exact title-cover
645 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
646 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
647 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
648 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
649 $weighted_query .= " or wrd,fuzzy,r8=\"$operand\"" if $fuzzy_enabled; # add fuzzy, word list
650 $weighted_query .= " or wrd,right-Truncation,r9=\"$stemmed_operand\"" if ($stemming and $stemmed_operand); # add stemming, right truncation
651 # embedded sorting: 0 a-z; 1 z-a
652 # $weighted_query .= ") or (sort1,aut=1";
654 elsif ( $index eq 'bc' ) {
655 $weighted_query .= "bc=\"$operand\"";
657 # if the index already has more than one qualifier, just wrap the operand
658 # in quotes and pass it back
659 elsif ($index =~ ',') {
660 $weighted_query .=" $index=\"$operand\"";
662 #TODO: build better cases based on specific search indexes
664 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
665 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
666 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
667 $weighted_query .= " or $index,rt,wrd,r3=\"$operand\""; # word list index
669 $weighted_query .= "))"; # close rank specification
670 return $weighted_query;
673 # build the query itself
675 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan) = @_;
677 my @operators = @$operators if $operators;
678 my @indexes = @$indexes if $indexes;
679 my @operands = @$operands if $operands;
680 my @limits = @$limits if $limits;
681 my @sort_by = @$sort_by if $sort_by;
683 my $stemming = C4::Context->preference("QueryStemming") || 0;
684 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
685 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
686 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
687 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
689 my $query = $operands[0];
690 my $simple_query = $operands[0];
699 my $stopwords_removed;
701 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
703 if ( $query =~ /^ccl=/ ) {
704 return ( undef, $', $', $', $', '', '', '', '', 'ccl' );
706 if ( $query =~ /^cql=/ ) {
707 return ( undef, $', $', $', $', '', '', '', '', 'cql' );
709 if ( $query =~ /^pqf=/ ) {
710 return ( undef, $', $', $', $', '', '', '', '', 'pqf' );
713 # pass nested queries directly
714 if ( $query =~ /(\(|\))/ ) {
715 return ( undef, $query, $simple_query, $query_cgi, $query, $limit, $limit_cgi, $limit_desc, $stopwords_removed, 'ccl' );
718 # form-based queries are limited to non-nested at a specific depth, so we can easily
719 # modify the incoming query operands and indexes to do stemming and field weighting
720 # Once we do so, we'll end up with a value in $query, just like if we had an
721 # incoming $query from the user
723 $query = ""; # clear it out so we can populate properly with field-weighted stemmed query
724 my $previous_operand; # a flag used to keep track if there was a previous query
725 # if there was, we can apply the current operator
727 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
729 # COMBINE OPERANDS, INDEXES AND OPERATORS
730 if ( $operands[$i] ) {
732 # a flag to determine whether or not to add the index to the query
734 # if the user is sophisticated enough to specify an index, turn off some defaults
735 if ($operands[$i] =~ /(:|=)/ || $scan) {
738 $remove_stopwords = 0;
740 my $operand = $operands[$i];
741 my $index = $indexes[$i];
743 # some helpful index modifs
744 my $index_plus = "$index:" if $index;
745 my $index_plus_comma="$index," if $index;
748 if ($remove_stopwords) {
749 ($operand, $stopwords_removed) = _remove_stopwords($operand,$index);
750 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
751 warn "REMOVED STOPWORDS: @$stopwords_removed" if ($stopwords_removed && $DEBUG);
755 my ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr);
756 my $truncated_operand;
757 ($nontruncated,$righttruncated,$lefttruncated,$rightlefttruncated,$regexpr) = _detect_truncation($operand,$index);
758 warn "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<" if $DEBUG;
761 if (scalar(@$righttruncated)+scalar(@$lefttruncated)+scalar(@$rightlefttruncated)>0){
762 # don't field weight or add the index to the query, we do it here
764 undef $weight_fields;
765 my $previous_truncation_operand;
766 if (scalar(@$nontruncated)>0) {
767 $truncated_operand.= "$index_plus @$nontruncated ";
768 $previous_truncation_operand = 1;
770 if (scalar(@$righttruncated)>0){
771 $truncated_operand .= "and " if $previous_truncation_operand;
772 $truncated_operand .= "$index_plus_comma"."rtrn:@$righttruncated ";
773 $previous_truncation_operand = 1;
775 if (scalar(@$lefttruncated)>0){
776 $truncated_operand .= "and " if $previous_truncation_operand;
777 $truncated_operand .= "$index_plus_comma"."ltrn:@$lefttruncated ";
778 $previous_truncation_operand = 1;
780 if (scalar(@$rightlefttruncated)>0){
781 $truncated_operand .= "and " if $previous_truncation_operand;
782 $truncated_operand .= "$index_plus_comma"."rltrn:@$rightlefttruncated ";
783 $previous_truncation_operand = 1;
786 $operand = $truncated_operand if $truncated_operand;
787 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
791 $stemmed_operand = _build_stemmed_operand($operand) if $stemming;
792 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
794 # Handle Field Weighting
795 my $weighted_operand;
796 $weighted_operand = _build_weighted_query($operand,$stemmed_operand,$index) if $weight_fields;
797 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
798 $operand = $weighted_operand if $weight_fields;
799 $indexes_set = 1 if $weight_fields;
801 # If there's a previous operand, we need to add an operator
802 if ($previous_operand) {
804 # user-specified operator
805 if ( $operators[$i-1] ) {
806 $query .= " $operators[$i-1] ";
807 $query .= " $index_plus " unless $indexes_set;
808 $query .= " $operand";
809 $query_cgi .="&op=$operators[$i-1]";
810 $query_cgi .="&idx=$index" if $index;
811 $query_cgi .="&q=$operands[$i]" if $operands[$i];
812 $query_desc .=" $operators[$i-1] $index_plus $operands[$i]";
815 # the default operator is and
818 $query .= "$index_plus " unless $indexes_set;
819 $query .= "$operand";
820 $query_cgi .="&op=and&idx=$index" if $index;
821 $query_cgi .="&q=$operands[$i]" if $operands[$i];
822 $query_desc .= " and $index_plus $operands[$i]";
826 # there isn't a pervious operand, don't need an operator
828 # field-weighted queries already have indexes set
829 $query .=" $index_plus " unless $indexes_set;
831 $query_desc .= " $index_plus $operands[$i]";
832 $query_cgi.="&idx=$index" if $index;
833 $query_cgi.="&q=$operands[$i]" if $operands[$i];
835 $previous_operand = 1;
840 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
844 my $availability_limit;
845 foreach my $this_limit (@limits) {
846 if ( $this_limit =~ /available/ ) {
847 # available is defined as (items.notloan is NULL) and (items.itemlost > 0 or NULL) (last clause handles NULL values for lost in zebra)
848 $availability_limit .="( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and ((lost,st-numeric gt 0) or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
849 $limit_cgi .= "&limit=available";
853 # these are treated as OR
854 elsif ( $this_limit =~ /mc/ ) {
855 $group_OR_limits .= " or " if $group_OR_limits;
856 $limit_desc .=" or " if $group_OR_limits;
857 $group_OR_limits .= "$this_limit";
858 $limit_cgi .="&limit=$this_limit";
859 $limit_desc .= "$this_limit";
864 $limit .= " and " if $limit || $query;
865 $limit .= "$this_limit";
866 $limit_cgi .="&limit=$this_limit";
867 $limit_desc .=" and $this_limit";
870 if ($group_OR_limits) {
871 $limit.=" and " if ($query || $limit );
872 $limit.="($group_OR_limits)";
874 if ($availability_limit) {
875 $limit.=" not " if ($query || $limit );
876 $limit.="$availability_limit";
878 # normalize the strings
881 for ($query, $query_desc, $limit, $limit_desc) {
882 $_ =~ s/ / /g; # remove extra spaces
883 $_ =~ s/^ //g; # remove any beginning spaces
884 $_ =~ s/ $//g; # remove any ending spaces
885 $_ =~ s/==/=/g; # remove double == from query
888 $query_cgi =~ s/^&//;
890 # append the limit to the query
891 $query .= " ".$limit;
893 warn "QUERY:".$query if $DEBUG;
894 warn "QUERY CGI:".$query_cgi if $DEBUG;
895 warn "QUERY DESC:".$query_desc if $DEBUG;
896 warn "LIMIT:".$limit if $DEBUG;
897 warn "LIMIT CGI:".$limit_cgi if $DEBUG;
898 warn "LIMIT DESC:".$limit_desc if $DEBUG;
900 return ( undef, $query,$simple_query,$query_cgi,$query_desc,$limit,$limit_cgi,$limit_desc,$stopwords_removed,$query_type );
903 # IMO this subroutine is pretty messy still -- it's responsible for
904 # building the HTML output for the template
906 my ( $searchdesc, $hits, $results_per_page, $offset, @marcresults ) = @_;
908 my $dbh = C4::Context->dbh;
912 my $span_terms_hashref;
913 for my $span_term ( split( / /, $searchdesc ) ) {
914 $span_term =~ s/(.*=|\)|\(|\+|\.)//g;
915 $span_terms_hashref->{$span_term}++;
918 #Build brancnames hash
920 #get branch information.....
923 $dbh->prepare("SELECT branchcode,branchname FROM branches")
924 ; # FIXME : use C4::Koha::GetBranches
926 while ( my $bdata = $bsth->fetchrow_hashref ) {
927 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
931 #find itemtype & itemtype image
934 $dbh->prepare("SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes");
936 while ( my $bdata = $bsth->fetchrow_hashref ) {
937 $itemtypes{ $bdata->{'itemtype'} }->{description} =
938 $bdata->{'description'};
939 $itemtypes{ $bdata->{'itemtype'} }->{imageurl} = $bdata->{'imageurl'};
940 $itemtypes{ $bdata->{'itemtype'} }->{summary} = $bdata->{'summary'};
941 $itemtypes{ $bdata->{'itemtype'} }->{notforloan} = $bdata->{'notforloan'};
944 #search item field code
947 "select tagfield from marc_subfield_structure where kohafield like 'items.itemnumber'"
950 my ($itemtag) = $sth->fetchrow;
952 ## find column names of items related to MARC
953 my $sth2 = $dbh->prepare("SHOW COLUMNS from items");
955 my %subfieldstosearch;
956 while ( ( my $column ) = $sth2->fetchrow ) {
957 my ( $tagfield, $tagsubfield ) =
958 &GetMarcFromKohaField( "items." . $column, "" );
959 $subfieldstosearch{$column} = $tagsubfield;
963 if ( $hits && $offset + $results_per_page <= $hits ) {
964 $times = $offset + $results_per_page;
970 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
972 $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
973 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, '' );
974 $oldbiblio->{result_number} = $i+1;
975 # add image url if there is one
976 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} =~ /^http:/ ) {
977 $oldbiblio->{imageurl} =
978 $itemtypes{ $oldbiblio->{itemtype} }->{imageurl};
979 $oldbiblio->{description} =
980 $itemtypes{ $oldbiblio->{itemtype} }->{description};
983 $oldbiblio->{imageurl} =
984 getitemtypeimagesrc() . "/"
985 . $itemtypes{ $oldbiblio->{itemtype} }->{imageurl}
986 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
987 $oldbiblio->{description} =
988 $itemtypes{ $oldbiblio->{itemtype} }->{description};
991 # build summary if there is one (the summary is defined in itemtypes table
993 if ($itemtypes{ $oldbiblio->{itemtype} }->{summary}) {
994 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
995 my @fields = $marcrecord->fields();
996 foreach my $field (@fields) {
997 my $tag = $field->tag();
998 my $tagvalue = $field->as_string();
999 $summary =~ s/\[(.?.?.?.?)$tag\*(.*?)]/$1$tagvalue$2\[$1$tag$2]/g;
1001 my @subf = $field->subfields;
1002 for my $i (0..$#subf) {
1003 my $subfieldcode = $subf[$i][0];
1004 my $subfieldvalue = $subf[$i][1];
1005 my $tagsubf = $tag.$subfieldcode;
1006 $summary =~ s/\[(.?.?.?.?)$tagsubf(.*?)]/$1$subfieldvalue$2\[$1$tagsubf$2]/g;
1010 $summary =~ s/\[(.*?)]//g;
1011 $summary =~ s/\n/<br>/g;
1012 $oldbiblio->{summary} = $summary;
1014 # add spans to search term in results for search term highlighting
1015 foreach my $term ( keys %$span_terms_hashref ) {
1016 my $old_term = $term;
1017 if ( length($term) > 3 ) {
1018 $term =~ s/(.*=|\)|\(|\+|\.|\?|\[|\])//g;
1022 #FIXME: is there a better way to do this?
1023 $oldbiblio->{'title'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1024 $oldbiblio->{'subtitle'} =~
1025 s/$term/<span class=\"term\">$&<\/span>/gi;
1027 $oldbiblio->{'author'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1028 $oldbiblio->{'publishercode'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1029 $oldbiblio->{'place'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1030 $oldbiblio->{'pages'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1031 $oldbiblio->{'notes'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1032 $oldbiblio->{'size'} =~ s/$term/<span class=\"term\">$&<\/span>/gi;
1037 $toggle = "#ffffcc";
1042 $oldbiblio->{'toggle'} = $toggle;
1043 my @fields = $marcrecord->field($itemtag);
1046 my $ordered_count = 0;
1047 my $onloan_count = 0;
1048 my $wthdrawn_count = 0;
1049 my $itemlost_count = 0;
1053 # check the loan status of the item :
1054 # it is not stored in the MARC record, for pref (zebra reindexing)
1055 # reason. Thus, we have to get the status from a specific SQL query
1057 my $sth_issue = $dbh->prepare("
1058 SELECT date_due,returndate
1060 WHERE itemnumber=? AND returndate IS NULL");
1061 my $items_count=scalar(@fields);
1062 foreach my $field (@fields) {
1064 foreach my $code ( keys %subfieldstosearch ) {
1065 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1067 $sth_issue->execute($item->{itemnumber});
1068 $item->{due_date} = format_date($sth_issue->fetchrow);
1069 $item->{onloan} = 1 if $item->{due_date};
1070 # at least one item can be reserved : suppose no
1072 if ( $item->{wthdrawn} ) {
1074 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1075 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{wthdrawn}=1;
1077 elsif ( $item->{itemlost} ) {
1079 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1080 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemlost}=1;
1082 unless ( $item->{notforloan}) {
1083 # OK, this one can be issued, so at least one can be reserved
1086 if ( ( $item->{onloan} ) && ( $item->{onloan} != '0000-00-00' ) )
1088 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{unavailable}=1;
1089 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{onloancount} = 1;
1090 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{due_date} = $item->{due_date};
1093 if ( $item->{'homebranch'} ) {
1094 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{count}++;
1098 elsif ( $item->{'holdingbranch'} ) {
1099 $items->{ $item->{'holdingbranch'} }->{count}++;
1101 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemcallnumber} = $item->{itemcallnumber};
1102 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{location} = $item->{location};
1103 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{branchcode} = $item->{homebranch};
1104 } # notforloan, item level and biblioitem level
1106 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1107 $norequests = 1 if $itemtypes{$oldbiblio->{itemtype}}->{notforloan};
1109 for my $key ( sort keys %$items ) {
1112 branchname => $branches{$items->{$key}->{branchcode}},
1113 branchcode => $items->{$key}->{branchcode},
1114 count => $items->{$key}->{count},
1115 itemcallnumber => $items->{$key}->{itemcallnumber},
1116 location => $items->{$key}->{location},
1117 onloancount => $items->{$key}->{onloancount},
1118 due_date => $items->{$key}->{due_date},
1119 wthdrawn => $items->{$key}->{wthdrawn},
1120 lost => $items->{$key}->{itemlost},
1122 # only show the number specified by the user
1123 my $maxitems = (C4::Context->preference('maxItemsinSearchResults')) ? C4::Context->preference('maxItemsinSearchResults')- 1 : 1;
1124 push @items_loop, $this_item unless $itemscount > $maxitems;;
1126 $oldbiblio->{norequests} = $norequests;
1127 $oldbiblio->{items_count} = $items_count;
1128 $oldbiblio->{items_loop} = \@items_loop;
1129 $oldbiblio->{onloancount} = $onloan_count;
1130 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1131 $oldbiblio->{itemlostcount} = $itemlost_count;
1132 $oldbiblio->{orderedcount} = $ordered_count;
1133 $oldbiblio->{isbn} =~ s/-//g; # deleting - in isbn to enable amazon content
1134 push( @newresults, $oldbiblio );
1141 #----------------------------------------------------------------------
1143 # Non-Zebra GetRecords#
1144 #----------------------------------------------------------------------
1148 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1152 my ($query,$simple_query,$sort_by_ref,$servers_ref,$results_per_page,$offset,$expanded_facet,$branches,$query_type,$scan) = @_;
1153 my $result = NZanalyse($query);
1154 return (undef,NZorder($result,@$sort_by_ref[0],$results_per_page,$offset),undef);
1159 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1160 the list is built from an inverted index in the nozebra SQL table
1161 note that title is here only for convenience : the sorting will be very fast when requested on title
1162 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1167 my ($string,$server) = @_;
1168 # $server contains biblioserver or authorities, depending on what we search on.
1169 #warn "querying : $string on $server";
1170 $server='biblioserver' unless $server;
1172 # if we have a ", replace the content to discard temporarily any and/or/not inside
1174 if ($string =~/"/) {
1175 $string =~ s/"(.*?)"/__X__/;
1177 warn "commacontent : $commacontent" if $DEBUG;
1179 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1180 # then, call again NZanalyse with $left and $right
1181 # (recursive until we find a leaf (=> something without and/or/not)
1182 $string =~ /(.*)( and | or | not | AND | OR | NOT )(.*)/;
1185 my $operand = lc($2); # FIXME: and/or/not are operators, not operands
1186 # it's not a leaf, we have a and/or/not
1188 # reintroduce comma content if needed
1189 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1190 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1191 warn "node : $left / $operand / $right\n" if $DEBUG;
1192 my $leftresult = NZanalyse($left,$server);
1193 my $rightresult = NZanalyse($right,$server);
1194 # OK, we have the results for right and left part of the query
1195 # depending of operand, intersect, union or exclude both lists
1196 # to get a result list
1197 if ($operand eq ' and ') {
1198 my @leftresult = split /;/, $leftresult;
1199 # my @rightresult = split /;/,$leftresult;
1201 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1202 # the result is stored twice, to have the same weight for AND than OR.
1203 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1204 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1205 foreach (@leftresult) {
1206 if ($rightresult =~ "$_;") {
1207 $finalresult .= "$_;$_;";
1210 return $finalresult;
1211 } elsif ($operand eq ' or ') {
1212 # just merge the 2 strings
1213 return $leftresult.$rightresult;
1214 } elsif ($operand eq ' not ') {
1215 my @leftresult = split /;/, $leftresult;
1216 # my @rightresult = split /;/,$leftresult;
1218 foreach (@leftresult) {
1219 unless ($rightresult =~ "$_;") {
1220 $finalresult .= "$_;";
1223 return $finalresult;
1225 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1226 die "error : operand unknown : $operand for $string";
1228 # it's a leaf, do the real SQL query and return the result
1230 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1231 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1232 warn "leaf : $string\n" if $DEBUG;
1233 # parse the string in in operator/operand/value again
1234 $string =~ /(.*)(>=|<=)(.*)/;
1238 unless ($operator) {
1239 $string =~ /(.*)(>|<|=)(.*)/;
1245 # automatic replace for short operators
1246 $left='title' if $left =~ '^ti';
1247 $left='author' if $left =~ '^au';
1248 $left='publisher' if $left =~ '^pb';
1249 $left='subject' if $left =~ '^su';
1250 $left='koha-Auth-Number' if $left =~ '^an';
1251 $left='keyword' if $left =~ '^kw';
1253 #do a specific search
1254 my $dbh = C4::Context->dbh;
1255 $operator='LIKE' if $operator eq '=' and $right=~ /%/;
1256 my $sth = $dbh->prepare("SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?");
1257 warn "$left / $operator / $right\n";
1258 # split each word, query the DB and build the biblionumbers result
1259 foreach (split / /,$right) {
1260 my ($biblionumbers,$value);
1262 warn "EXECUTE : $server, $left, $_";
1263 $sth->execute($server, $left, $_) or warn "execute failed: $!";
1264 while (my ($line,$value) = $sth->fetchrow) {
1265 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1266 # otherwise, fill the result
1267 $biblionumbers .= $line unless ($right =~ /\d/ && $value =~ /\D/);
1268 warn "result : $value ". ($right =~ /\d/) . "==".(!$value =~ /\d/) ;#= $line";
1270 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1272 my @leftresult = split /;/, $biblionumbers;
1274 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1275 # remove weight at the end
1276 my $cleaned = $entry;
1277 $cleaned =~ s/-\d*$//;
1278 # if the entry already in the hash, take it & increase weight
1279 warn "===== $cleaned =====" if $DEBUG;
1280 if ($results =~ "$cleaned") {
1281 $temp .= "$entry;$entry;";
1282 warn "INCLUDING $entry" if $DEBUG;
1287 $results = $biblionumbers;
1291 #do a complete search (all indexes)
1292 my $dbh = C4::Context->dbh;
1293 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?");
1294 # split each word, query the DB and build the biblionumbers result
1295 foreach (split / /,$string) {
1296 next if C4::Context->stopwords->{uc($_)}; # skip if stopword
1297 warn "search on all indexes on $_" if $DEBUG;
1300 $sth->execute($server, $_);
1301 while (my $line = $sth->fetchrow) {
1302 $biblionumbers .= $line;
1304 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1306 warn "RES for $_ = $biblionumbers" if $DEBUG;
1307 my @leftresult = split /;/, $biblionumbers;
1309 foreach my $entry (@leftresult) { # $_ contains biblionumber,title-weight
1310 # remove weight at the end
1311 my $cleaned = $entry;
1312 $cleaned =~ s/-\d*$//;
1313 # if the entry already in the hash, take it & increase weight
1314 warn "===== $cleaned =====" if $DEBUG;
1315 if ($results =~ "$cleaned") {
1316 $temp .= "$entry;$entry;";
1317 warn "INCLUDING $entry" if $DEBUG;
1322 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
1323 $results = $biblionumbers;
1327 warn "return : $results for LEAF : $string" if $DEBUG;
1334 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
1342 my ($biblionumbers, $ordering,$results_per_page,$offset) = @_;
1343 # order title asc by default
1344 # $ordering = '1=36 <i' unless $ordering;
1345 $results_per_page=20 unless $results_per_page;
1346 $offset = 0 unless $offset;
1347 my $dbh = C4::Context->dbh;
1349 # order by POPULARITY
1351 if ($ordering =~ /popularity/) {
1354 # popularity is not in MARC record, it's builded from a specific query
1355 my $sth = $dbh->prepare("select sum(issues) from items where biblionumber=?");
1356 foreach (split /;/,$biblionumbers) {
1357 my ($biblionumber,$title) = split /,/,$_;
1358 $result{$biblionumber}=GetMarcBiblio($biblionumber);
1359 $sth->execute($biblionumber);
1360 my $popularity= $sth->fetchrow ||0;
1361 # hint : the key is popularity.title because we can have
1362 # many results with the same popularity. In this cas, sub-ordering is done by title
1363 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1364 # (un-frequent, I agree, but we won't forget anything that way ;-)
1365 $popularity{sprintf("%10d",$popularity).$title.$biblionumber} = $biblionumber;
1367 # sort the hash and return the same structure as GetRecords (Zebra querying)
1370 if ($ordering eq 'popularity_dsc') { # sort popularity DESC
1371 foreach my $key (sort {$b cmp $a} (keys %popularity)) {
1372 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1374 } else { # sort popularity ASC
1375 foreach my $key (sort (keys %popularity)) {
1376 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1380 $result_hash->{'hits'} = $numbers;
1381 $finalresult->{'biblioserver'} = $result_hash;
1382 return $finalresult;
1386 } elsif ($ordering =~/author/){
1388 foreach (split /;/,$biblionumbers) {
1389 my ($biblionumber,$title) = split /,/,$_;
1390 my $record=GetMarcBiblio($biblionumber);
1392 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1393 $author=$record->subfield('200','f');
1394 $author=$record->subfield('700','a') unless $author;
1396 $author=$record->subfield('100','a');
1398 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1399 # and we don't want to get only 1 result for each of them !!!
1400 $result{$author.$biblionumber}=$record;
1402 # sort the hash and return the same structure as GetRecords (Zebra querying)
1405 if ($ordering eq 'author_za') { # sort by author desc
1406 foreach my $key (sort { $b cmp $a } (keys %result)) {
1407 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1409 } else { # sort by author ASC
1410 foreach my $key (sort (keys %result)) {
1411 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1415 $result_hash->{'hits'} = $numbers;
1416 $finalresult->{'biblioserver'} = $result_hash;
1417 return $finalresult;
1419 # ORDER BY callnumber
1421 } elsif ($ordering =~/callnumber/){
1423 foreach (split /;/,$biblionumbers) {
1424 my ($biblionumber,$title) = split /,/,$_;
1425 my $record=GetMarcBiblio($biblionumber);
1427 my ($callnumber_tag,$callnumber_subfield)=GetMarcFromKohaField($dbh,'items.itemcallnumber');
1428 ($callnumber_tag,$callnumber_subfield)= GetMarcFromKohaField('biblioitems.callnumber') unless $callnumber_tag;
1429 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1430 $callnumber=$record->subfield('200','f');
1432 $callnumber=$record->subfield('100','a');
1434 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1435 # and we don't want to get only 1 result for each of them !!!
1436 $result{$callnumber.$biblionumber}=$record;
1438 # sort the hash and return the same structure as GetRecords (Zebra querying)
1441 if ($ordering eq 'call_number_dsc') { # sort by title desc
1442 foreach my $key (sort { $b cmp $a } (keys %result)) {
1443 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1445 } else { # sort by title ASC
1446 foreach my $key (sort { $a cmp $b } (keys %result)) {
1447 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1451 $result_hash->{'hits'} = $numbers;
1452 $finalresult->{'biblioserver'} = $result_hash;
1453 return $finalresult;
1454 } elsif ($ordering =~ /pubdate/){ #pub year
1456 foreach (split /;/,$biblionumbers) {
1457 my ($biblionumber,$title) = split /,/,$_;
1458 my $record=GetMarcBiblio($biblionumber);
1459 my ($publicationyear_tag,$publicationyear_subfield)=GetMarcFromKohaField('biblioitems.publicationyear','');
1460 my $publicationyear=$record->subfield($publicationyear_tag,$publicationyear_subfield);
1461 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1462 # and we don't want to get only 1 result for each of them !!!
1463 $result{$publicationyear.$biblionumber}=$record;
1465 # sort the hash and return the same structure as GetRecords (Zebra querying)
1468 if ($ordering eq 'pubdate_dsc') { # sort by pubyear desc
1469 foreach my $key (sort { $b cmp $a } (keys %result)) {
1470 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1472 } else { # sort by pub year ASC
1473 foreach my $key (sort (keys %result)) {
1474 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1478 $result_hash->{'hits'} = $numbers;
1479 $finalresult->{'biblioserver'} = $result_hash;
1480 return $finalresult;
1484 } elsif ($ordering =~ /title/) {
1485 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
1487 foreach (split /;/,$biblionumbers) {
1488 my ($biblionumber,$title) = split /,/,$_;
1489 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1490 # and we don't want to get only 1 result for each of them !!!
1491 # hint & speed improvement : we can order without reading the record
1492 # so order, and read records only for the requested page !
1493 $result{$title.$biblionumber}=$biblionumber;
1495 # sort the hash and return the same structure as GetRecords (Zebra querying)
1498 if ($ordering eq 'title_az') { # sort by title desc
1499 foreach my $key (sort (keys %result)) {
1500 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1502 } else { # sort by title ASC
1503 foreach my $key (sort { $b cmp $a } (keys %result)) {
1504 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1507 # limit the $results_per_page to result size if it's more
1508 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1509 # for the requested page, replace biblionumber by the complete record
1510 # speed improvement : avoid reading too much things
1511 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1512 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1515 $result_hash->{'hits'} = $numbers;
1516 $finalresult->{'biblioserver'} = $result_hash;
1517 return $finalresult;
1522 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
1525 foreach (split /;/,$biblionumbers) {
1526 my ($biblionumber,$title) = split /,/,$_;
1527 $title =~ /(.*)-(\d)/;
1530 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
1531 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
1532 # biblio N has ranking = 6
1533 $count_ranking{$biblionumber} += $ranking;
1535 # build the result by "inverting" the count_ranking hash
1536 # 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
1538 foreach (keys %count_ranking) {
1539 $result{sprintf("%10d",$count_ranking{$_}).'-'.$_} = $_;
1541 # sort the hash and return the same structure as GetRecords (Zebra querying)
1544 foreach my $key (sort {$b cmp $a} (keys %result)) {
1545 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1547 # limit the $results_per_page to result size if it's more
1548 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1549 # for the requested page, replace biblionumber by the complete record
1550 # speed improvement : avoid reading too much things
1551 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1552 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc if $result_hash->{'RECORDS'}[$counter];
1555 $result_hash->{'hits'} = $numbers;
1556 $finalresult->{'biblioserver'} = $result_hash;
1557 return $finalresult;
1562 ($countchanged,$listunchanged) = ModBiblios($listbiblios, $tagsubfield,$initvalue,$targetvalue,$test);
1564 this function changes all the values $initvalue in subfield $tag$subfield in any record in $listbiblios
1565 test parameter if set donot perform change to records in database.
1571 * $listbiblios is an array ref to marcrecords to be changed
1572 * $tagsubfield is the reference of the subfield to change.
1573 * $initvalue is the value to search the record for
1574 * $targetvalue is the value to set the subfield to
1575 * $test is to be set only not to perform changes in database.
1577 =item C<Output arg:>
1578 * $countchanged counts all the changes performed.
1579 * $listunchanged contains the list of all the biblionumbers of records unchanged.
1581 =item C<usage in the script:>
1585 my ($countchanged, $listunchanged) = EditBiblios($results->{RECORD}, $tagsubfield,$initvalue,$targetvalue);;
1586 #If one wants to display unchanged records, you should get biblios foreach @$listunchanged
1587 $template->param(countchanged => $countchanged, loopunchanged=>$listunchanged);
1592 my ($listbiblios,$tagsubfield,$initvalue,$targetvalue,$test)=@_;
1595 my ($tag,$subfield)=($1,$2) if ($tagsubfield=~/^(\d{1,3})([a-z0-9A-Z@])?$/);
1596 if ((length($tag)<3)&& $subfield=~/0-9/){
1597 $tag=$tag.$subfield;
1600 my ($bntag,$bnsubf) = GetMarcFromKohaField('biblio.biblionumber');
1601 my ($itemtag,$itemsubf) = GetMarcFromKohaField('items.itemnumber');
1602 foreach my $usmarc (@$listbiblios){
1604 $record=eval{MARC::Record->new_from_usmarc($usmarc)};
1607 # usmarc is not a valid usmarc May be a biblionumber
1608 if ($tag eq $itemtag){
1609 my $bib=GetBiblioFromItemNumber($usmarc);
1610 $record=GetMarcItem($bib->{'biblionumber'},$usmarc) ;
1611 $biblionumber=$bib->{'biblionumber'};
1613 $record=GetMarcBiblio($usmarc);
1614 $biblionumber=$usmarc;
1618 $biblionumber = $record->subfield($bntag,$bnsubf);
1620 $biblionumber=$record->field($bntag)->data;
1623 #GetBiblionumber is to be written.
1624 #Could be replaced by TransformMarcToKoha (But Would be longer)
1625 if ($record->field($tag)){
1627 foreach my $field ($record->field($tag)){
1629 if ($field->delete_subfield('code' =>$subfield,'match'=>qr($initvalue))){
1632 $field->update($subfield,$targetvalue) if ($targetvalue);
1636 if ($field->delete_field($field)){
1641 $field->data=$targetvalue if ($field->data=~qr($initvalue));
1645 # warn $record->as_formatted;
1647 ModBiblio($record,$biblionumber,GetFrameworkCode($biblionumber)) unless ($test);
1649 push @unmatched, $biblionumber;
1652 push @unmatched, $biblionumber;
1655 return ($countmatched,\@unmatched);
1658 END { } # module clean-up code here (global destructor)
1665 Koha Developement team <info@koha.org>