3 # This file is part of Koha.
5 # Koha is free software; you can redistribute it and/or modify it under the
6 # terms of the GNU General Public License as published by the Free Software
7 # Foundation; either version 2 of the License, or (at your option) any later
10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
14 # You should have received a copy of the GNU General Public License along with
15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
16 # Suite 330, Boston, MA 02111-1307 USA
19 #use warnings; FIXME - Bug 2505
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
25 use C4::Search::PazPar2;
27 use C4::Dates qw(format_date);
28 use C4::Members qw(GetHideLostItemsPreference);
31 use C4::Reserves; # CheckReserves
40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
42 # set the version for version checking
44 $VERSION = 3.07.00.049;
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
50 C4::Search - Functions for searching the Koha catalog.
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
58 This module provides searching functions for Koha's bibliographic databases
74 &enabled_staff_search_views
78 # make all your functions, whether exported or not;
82 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
84 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
90 my $dbh = C4::Context->dbh;
91 my $result = TransformMarcToKoha( $dbh, $record, '' );
96 my ( $biblionumber, $title );
98 # search duplicate on ISBN, easy and fast..
100 if ( $result->{isbn} ) {
101 $result->{isbn} =~ s/\(.*$//;
102 $result->{isbn} =~ s/\s+$//;
103 $query = "isbn=$result->{isbn}";
106 $result->{title} =~ s /\\//g;
107 $result->{title} =~ s /\"//g;
108 $result->{title} =~ s /\(//g;
109 $result->{title} =~ s /\)//g;
111 # FIXME: instead of removing operators, could just do
112 # quotes around the value
113 $result->{title} =~ s/(and|or|not)//g;
114 $query = "ti,ext=$result->{title}";
115 if ( $result->{author} ) {
116 $result->{author} =~ s /\\//g;
117 $result->{author} =~ s /\"//g;
118 $result->{author} =~ s /\(//g;
119 $result->{author} =~ s /\)//g;
121 # remove valid operators
122 $result->{author} =~ s/(and|or|not)//g;
123 $query .= " and au,ext=$result->{author}";
127 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
129 if (!defined $error) {
130 foreach my $possible_duplicate_record (@{$searchresults}) {
132 MARC::Record->new_from_usmarc($possible_duplicate_record);
133 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
135 # FIXME :: why 2 $biblionumber ?
137 push @results, $result->{'biblionumber'};
138 push @results, $result->{'title'};
147 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
149 This function provides a simple search API on the bibliographic catalog
155 * $query can be a simple keyword or a complete CCL query
156 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
157 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
158 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
163 Returns an array consisting of three elements
164 * $error is undefined unless an error is detected
165 * $results is a reference to an array of records.
166 * $total_hits is the number of hits that would have been returned with no limit
168 If an error is returned the two other return elements are undefined. If error itself is undefined
169 the other two elements are always defined
171 =item C<usage in the script:>
175 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
177 if (defined $error) {
178 $template->param(query_error => $error);
179 warn "error: ".$error;
180 output_html_with_http_headers $input, $cookie, $template->output;
184 my $hits = @{$marcresults};
187 for my $r ( @{$marcresults} ) {
188 my $marcrecord = MARC::File::USMARC::decode($r);
189 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
191 #build the iarray of hashs for the template.
193 title => $biblio->{'title'},
194 subtitle => $biblio->{'subtitle'},
195 biblionumber => $biblio->{'biblionumber'},
196 author => $biblio->{'author'},
197 publishercode => $biblio->{'publishercode'},
198 publicationyear => $biblio->{'publicationyear'},
203 $template->param(result=>\@results);
208 my ( $query, $offset, $max_results, $servers ) = @_;
210 if ( C4::Context->preference('NoZebra') ) {
211 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
214 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
215 return ( undef, $search_result, scalar($result->{hits}) );
218 return ( 'No query entered', undef, undef ) unless $query;
219 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
220 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
227 # Initialize & Search Zebra
228 for ( my $i = 0 ; $i < @servers ; $i++ ) {
230 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
231 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
232 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
236 $zconns[$i]->errmsg() . " ("
237 . $zconns[$i]->errcode() . ") "
238 . $zconns[$i]->addinfo() . " "
239 . $zconns[$i]->diagset();
241 return ( $error, undef, undef ) if $zconns[$i]->errcode();
245 # caught a ZOOM::Exception
249 . $@->addinfo() . " "
251 warn $error." for query: $query";
252 return ( $error, undef, undef );
261 my $first_record = defined($offset) ? $offset + 1 : 1;
262 my $hits = $tmpresults[ $i - 1 ]->size();
263 $total_hits += $hits;
264 my $last_record = $hits;
265 if ( defined $max_results && $offset + $max_results < $hits ) {
266 $last_record = $offset + $max_results;
269 for my $j ( $first_record .. $last_record ) {
271 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
273 push @{$results}, $record;
278 foreach my $zoom_query (@zoom_queries) {
279 $zoom_query->destroy();
282 return ( undef, $results, $total_hits );
288 ( undef, $results_hashref, \@facets_loop ) = getRecords (
290 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
291 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
295 The all singing, all dancing, multi-server, asynchronous, scanning,
296 searching, record nabbing, facet-building
298 See verbse embedded documentation.
304 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
305 $results_per_page, $offset, $expanded_facet, $branches,
306 $itemtypes, $query_type, $scan, $opac
309 my @servers = @$servers_ref;
310 my @sort_by = @$sort_by_ref;
312 # Initialize variables for the ZOOM connection and results object
316 my $results_hashref = ();
318 # Initialize variables for the faceted results objects
319 my $facets_counter = ();
320 my $facets_info = ();
321 my $facets = getFacets();
322 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
324 my @facets_loop; # stores the ref to array of hashes for template facets loop
326 ### LOOP THROUGH THE SERVERS
327 for ( my $i = 0 ; $i < @servers ; $i++ ) {
328 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
330 # perform the search, create the results objects
331 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
332 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
334 #$query_to_use = $simple_query if $scan;
335 warn $simple_query if ( $scan and $DEBUG );
337 # Check if we've got a query_type defined, if so, use it
340 if ($query_type =~ /^ccl/) {
341 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
342 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
343 } elsif ($query_type =~ /^cql/) {
344 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
345 } elsif ($query_type =~ /^pqf/) {
346 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
348 warn "Unknown query_type '$query_type'. Results undetermined.";
351 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
353 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
357 warn "WARNING: query problem with $query_to_use " . $@;
360 # Concatenate the sort_by limits and pass them to the results object
361 # Note: sort will override rank
363 foreach my $sort (@sort_by) {
364 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
365 $sort_by .= "1=1003 <i ";
367 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
368 $sort_by .= "1=1003 >i ";
370 elsif ( $sort eq "popularity_asc" ) {
371 $sort_by .= "1=9003 <i ";
373 elsif ( $sort eq "popularity_dsc" ) {
374 $sort_by .= "1=9003 >i ";
376 elsif ( $sort eq "call_number_asc" ) {
377 $sort_by .= "1=8007 <i ";
379 elsif ( $sort eq "call_number_dsc" ) {
380 $sort_by .= "1=8007 >i ";
382 elsif ( $sort eq "pubdate_asc" ) {
383 $sort_by .= "1=31 <i ";
385 elsif ( $sort eq "pubdate_dsc" ) {
386 $sort_by .= "1=31 >i ";
388 elsif ( $sort eq "acqdate_asc" ) {
389 $sort_by .= "1=32 <i ";
391 elsif ( $sort eq "acqdate_dsc" ) {
392 $sort_by .= "1=32 >i ";
394 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
395 $sort_by .= "1=4 <i ";
397 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
398 $sort_by .= "1=4 >i ";
401 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
404 if ($sort_by && !$scan) {
405 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
406 warn "WARNING sort $sort_by failed";
409 } # finished looping through servers
411 # The big moment: asynchronously retrieve results from all servers
416 my ( $i, $size ) = @_;
419 # loop through the results
420 $results_hash->{'hits'} = $size;
422 if ( $offset + $results_per_page <= $size ) {
423 $times = $offset + $results_per_page;
428 for ( my $j = $offset ; $j < $times ; $j++ ) {
432 ## Check if it's an index scan
434 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
436 # here we create a minimal MARC record and hand it off to the
437 # template just like a normal result ... perhaps not ideal, but
439 my $tmprecord = MARC::Record->new();
440 $tmprecord->encoding('UTF-8');
444 # the minimal record in author/title (depending on MARC flavour)
445 if ( C4::Context->preference("marcflavour") eq
448 $tmptitle = MARC::Field->new(
453 $tmprecord->append_fields($tmptitle);
457 MARC::Field->new( '245', ' ', ' ', a => $term, );
459 MARC::Field->new( '100', ' ', ' ', a => $occ, );
460 $tmprecord->append_fields($tmptitle);
461 $tmprecord->append_fields($tmpauthor);
463 $results_hash->{'RECORDS'}[$j] =
464 $tmprecord->as_usmarc();
469 $record = $results[ $i - 1 ]->record($j)->raw();
471 # warn "RECORD $j:".$record;
472 $results_hash->{'RECORDS'}[$j] = $record;
476 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
478 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
479 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
482 $size > $facets_maxrecs ? $facets_maxrecs : $size;
483 for my $facet (@$facets) {
484 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
486 $results[ $i - 1 ]->record($j)->render();
488 foreach my $tag ( @{ $facet->{tags} } ) {
491 my $tag_num = substr( $tag, 0, 3 );
492 my $letters = substr( $tag, 3 );
494 '\n' . $tag_num . ' ([^z][^\n]+)';
495 $field_pattern = '\n' . $tag_num . ' ([^\n]+)'
496 if ( int($tag_num) < 10 );
498 ( $render_record =~ /$field_pattern/g );
499 foreach my $field_token (@field_tokens) {
500 my @subf = ( $field_token =~
501 /\$([a-zA-Z0-9]) ([^\$]+)/g );
503 for ( my $i = 0 ; $i < @subf ; $i += 2 ) {
504 if ( $letters =~ $subf[$i] ) {
505 my $value = $subf[ $i + 1 ];
508 push @values, $value;
511 my $data = join( $facet->{sep}, @values );
512 unless ( $data ~~ @used_datas ) {
513 $facets_counter->{ $facet->{idx} }
515 push @used_datas, $data;
520 $facets_info->{ $facet->{idx} }->{label_value} =
522 $facets_info->{ $facet->{idx} }->{expanded} =
527 # warn "connection ", $i-1, ": $size hits";
528 # warn $results[$i-1]->record(0)->render() if $size > 0;
531 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
533 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
534 keys %$facets_counter
538 my $number_of_facets;
539 my @this_facets_array;
542 $facets_counter->{$link_value}
543 ->{$b} <=> $facets_counter->{$link_value}
545 } keys %{ $facets_counter->{$link_value} }
549 if ( ( $number_of_facets < 6 )
550 || ( $expanded_facet eq $link_value )
551 || ( $facets_info->{$link_value}->{'expanded'} )
555 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
556 my $facet_link_value = $one_facet;
557 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
559 # fix the length that will display in the label,
560 my $facet_label_value = $one_facet;
561 my $facet_max_length = C4::Context->preference(
562 'FacetLabelTruncationLength')
565 substr( $one_facet, 0, $facet_max_length )
567 if length($facet_label_value) >
570 # if it's a branch, label by the name, not the code,
571 if ( $link_value =~ /branch/ ) {
572 if ( defined $branches
573 && ref($branches) eq "HASH"
574 && defined $branches->{$one_facet}
575 && ref( $branches->{$one_facet} ) eq
579 $branches->{$one_facet}
583 $facet_label_value = "*";
587 # if it's a itemtype, label by the name, not the code,
588 if ( $link_value =~ /itype/ ) {
589 if ( defined $itemtypes
590 && ref($itemtypes) eq "HASH"
591 && defined $itemtypes->{$one_facet}
592 && ref( $itemtypes->{$one_facet} ) eq
596 $itemtypes->{$one_facet}
601 # also, if it's a location code, use the name instead of the code
602 if ( $link_value =~ /location/ ) {
604 GetKohaAuthorisedValueLib( 'LOC',
608 # but we're down with the whole label being in the link's title.
609 push @this_facets_array,
612 $facets_counter->{$link_value}
614 facet_label_value => $facet_label_value,
615 facet_title_value => $one_facet,
616 facet_link_value => $facet_link_value,
617 type_link_value => $link_value,
619 if ($facet_label_value);
623 # handle expanded option
624 unless ( $facets_info->{$link_value}->{'expanded'} ) {
626 if ( ( $number_of_facets > 6 )
627 && ( $expanded_facet ne $link_value ) );
631 type_link_value => $link_value,
632 type_id => $link_value . "_id",
634 . $facets_info->{$link_value}->{'label_value'} =>
636 facets => \@this_facets_array,
637 expandable => $expandable,
638 expand => $link_value,
642 $facets_info->{$link_value}->{'label_value'} =~
645 and ( C4::Context->preference('singleBranchMode') )
651 return ( undef, $results_hashref, \@facets_loop );
656 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
657 $results_per_page, $offset, $expanded_facet, $branches,
661 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
663 $paz->search($simple_query);
664 sleep 1; # FIXME: WHY?
667 my $results_hashref = {};
668 my $stats = XMLin($paz->stat);
669 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
671 # for a grouped search result, the number of hits
672 # is the number of groups returned; 'bib_hits' will have
673 # the total number of bibs.
674 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
675 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
677 HIT: foreach my $hit (@{ $results->{'hit'} }) {
678 my $recid = $hit->{recid}->[0];
680 my $work_title = $hit->{'md-work-title'}->[0];
682 if (exists $hit->{'md-work-author'}) {
683 $work_author = $hit->{'md-work-author'}->[0];
685 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
687 my $result_group = {};
688 $result_group->{'group_label'} = $group_label;
689 $result_group->{'group_merge_key'} = $recid;
692 if (exists $hit->{count}) {
693 $count = $hit->{count}->[0];
695 $result_group->{'group_count'} = $count;
697 for (my $i = 0; $i < $count; $i++) {
698 # FIXME -- may need to worry about diacritics here
699 my $rec = $paz->record($recid, $i);
700 push @{ $result_group->{'RECORDS'} }, $rec;
703 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
706 # pass through facets
707 my $termlist_xml = $paz->termlist('author,subject');
708 my $terms = XMLin($termlist_xml, forcearray => 1);
709 my @facets_loop = ();
710 #die Dumper($results);
711 # foreach my $list (sort keys %{ $terms->{'list'} }) {
713 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
715 # facet_label_value => $facet->{'name'}->[0],
718 # push @facets_loop, ( {
719 # type_label => $list,
720 # facets => \@facets,
724 return ( undef, $results_hashref, \@facets_loop );
728 sub _remove_stopwords {
729 my ( $operand, $index ) = @_;
730 my @stopwords_removed;
732 # phrase and exact-qualified indexes shouldn't have stopwords removed
733 if ( $index !~ m/phr|ext/ ) {
735 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
736 # we use IsAlpha unicode definition, to deal correctly with diacritics.
737 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
738 # is a stopword, we'd get "çon" and wouldn't find anything...
740 foreach ( keys %{ C4::Context->stopwords } ) {
741 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
742 if ( my ($matched) = ($operand =~
743 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
745 $operand =~ s/\Q$matched\E/ /gi;
746 push @stopwords_removed, $_;
750 return ( $operand, \@stopwords_removed );
754 sub _detect_truncation {
755 my ( $operand, $index ) = @_;
756 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
759 my @wordlist = split( /\s/, $operand );
760 foreach my $word (@wordlist) {
761 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
762 push @rightlefttruncated, $word;
764 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
765 push @lefttruncated, $word;
767 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
768 push @righttruncated, $word;
770 elsif ( index( $word, "*" ) < 0 ) {
771 push @nontruncated, $word;
774 push @regexpr, $word;
778 \@nontruncated, \@righttruncated, \@lefttruncated,
779 \@rightlefttruncated, \@regexpr
784 sub _build_stemmed_operand {
785 my ($operand,$lang) = @_;
786 require Lingua::Stem::Snowball ;
787 my $stemmed_operand=q{};
789 # If operand contains a digit, it is almost certainly an identifier, and should
790 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
791 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
792 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
793 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
794 return $operand if $operand =~ /\d/;
796 # FIXME: the locale should be set based on the user's language and/or search choice
798 # Make sure we only use the first two letters from the language code
799 $lang = lc(substr($lang, 0, 2));
800 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
801 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
802 if ($lang eq 'nb' || $lang eq 'nn') {
805 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
806 encoding => "UTF-8" );
808 my @words = split( / /, $operand );
809 my @stems = $stemmer->stem(\@words);
810 for my $stem (@stems) {
811 $stemmed_operand .= "$stem";
812 $stemmed_operand .= "?"
813 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
814 $stemmed_operand .= " ";
816 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
817 return $stemmed_operand;
821 sub _build_weighted_query {
823 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
824 # pretty well but could work much better if we had a smarter query parser
825 my ( $operand, $stemmed_operand, $index ) = @_;
826 my $stemming = C4::Context->preference("QueryStemming") || 0;
827 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
828 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
830 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
832 # Keyword, or, no index specified
833 if ( ( $index eq 'kw' ) || ( !$index ) ) {
835 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
836 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
837 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
838 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
839 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
840 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
841 if $fuzzy_enabled; # add fuzzy, word list
842 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
843 if ( $stemming and $stemmed_operand )
844 ; # add stemming, right truncation
845 $weighted_query .= " or wrdl,r9=\"$operand\"";
847 # embedded sorting: 0 a-z; 1 z-a
848 # $weighted_query .= ") or (sort1,aut=1";
851 # Barcode searches should skip this process
852 elsif ( $index eq 'bc' ) {
853 $weighted_query .= "bc=\"$operand\"";
856 # Authority-number searches should skip this process
857 elsif ( $index eq 'an' ) {
858 $weighted_query .= "an=\"$operand\"";
861 # If the index already has more than one qualifier, wrap the operand
862 # in quotes and pass it back (assumption is that the user knows what they
863 # are doing and won't appreciate us mucking up their query
864 elsif ( $index =~ ',' ) {
865 $weighted_query .= " $index=\"$operand\"";
868 #TODO: build better cases based on specific search indexes
870 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
871 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
872 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
874 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
877 $weighted_query .= "))"; # close rank specification
878 return $weighted_query;
883 Return an array with available indexes.
905 'Author-personal-bibliography',
915 'Chronological-subdivision',
925 'Conference-name-heading',
926 'Conference-name-see',
927 'Conference-name-seealso',
932 'Corporate-name-heading',
933 'Corporate-name-see',
934 'Corporate-name-seealso',
936 'date-entered-on-file',
937 'Date-of-acquisition',
938 'Date-of-publication',
939 'Dewey-classification',
946 'Geographic-subdivision',
949 'Heading-use-main-or-added-entry',
950 'Heading-use-series-added-entry ',
951 'Heading-use-subject-added-entry',
969 'Local-classification',
972 'Match-heading-see-from',
980 'Name-geographic-heading',
981 'Name-geographic-see',
982 'Name-geographic-seealso',
990 'Personal-name-heading',
992 'Personal-name-seealso',
999 'Record-control-number',
1010 'Subject-heading-thesaurus',
1011 'Subject-name-personal',
1012 'Subject-subdivision',
1022 'Term-genre-form-heading',
1023 'Term-genre-form-see',
1024 'Term-genre-form-seealso',
1031 'Title-uniform-heading',
1032 'Title-uniform-see',
1033 'Title-uniform-seealso',
1043 'classification-source',
1045 'coded-location-qualifier',
1056 'Local-classification',
1059 'materials-specified',
1068 'replacementpricedate',
1083 =head2 _handle_exploding_index
1085 my $query = _handle_exploding_index($index, $term)
1087 Callback routine to generate the search for "exploding" indexes (i.e.
1088 those indexes which are turned into multiple or-connected searches based
1093 sub _handle_exploding_index {
1094 my ( $index, $term ) = @_;
1096 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1098 my $marcflavour = C4::Context->preference('marcflavour');
1100 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1101 my $wantedcodes = '';
1102 my @subqueries = ( "(su=\"$term\")");
1103 my ($error, $results, $total_hits) = SimpleSearch( "Heading,wrdl=$term", undef, undef, [ "authorityserver" ] );
1104 foreach my $auth (@$results) {
1105 my $record = MARC::Record->new_from_usmarc($auth);
1106 my @references = $record->field('5..');
1108 if ($index eq 'su-br') {
1110 } elsif ($index eq 'su-na') {
1112 } elsif ($index eq 'su-rl') {
1115 foreach my $reference (@references) {
1116 my $codes = $reference->subfield($codesubfield);
1117 push @subqueries, '(su="' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '")' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1121 return join(' or ', @subqueries);
1126 ( $operators, $operands, $indexes, $limits,
1127 $sort_by, $scan, $lang ) =
1128 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1130 Shim function to ease the transition from buildQuery to a new QueryParser.
1131 This function is called at the beginning of buildQuery, and modifies
1132 buildQuery's input. If it can handle the input, it returns a query that
1133 buildQuery will not try to parse.
1137 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1139 my @operators = $operators ? @$operators : ();
1140 my @indexes = $indexes ? @$indexes : ();
1141 my @operands = $operands ? @$operands : ();
1142 my @limits = $limits ? @$limits : ();
1143 my @sort_by = $sort_by ? @$sort_by : ();
1145 my $query = $operands[0];
1149 # TODO: once we are using QueryParser, all this special case code for
1150 # exploded search indexes will be replaced by a callback to
1151 # _handle_exploding_index
1152 if ( $query =~ m/^(.*)\b(su-br|su-na|su-rl)[:=](\w.*)$/ ) {
1158 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1159 if ($operands[$i] && $indexes[$i] =~ m/(su-br|su-na|su-rl)/) {
1160 $index = $indexes[$i];
1161 $term = $operands[$i];
1162 } elsif ($operands[$i]) {
1163 $query .= $operators[$i] eq 'or' ? ' or ' : ' and ' if ($query);
1164 $query .= "($indexes[$i]:$operands[$i])";
1170 my $queryPart = _handle_exploding_index($index, $term);
1172 $query .= "($queryPart)";
1175 $operands[0] = "ccl=$query";
1178 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang);
1184 $simple_query, $query_cgi,
1185 $query_desc, $limit,
1186 $limit_cgi, $limit_desc,
1187 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1189 Build queries and limits in CCL, CGI, Human,
1190 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1192 See verbose embedded documentation.
1198 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1200 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1202 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1205 my @operators = $operators ? @$operators : ();
1206 my @indexes = $indexes ? @$indexes : ();
1207 my @operands = $operands ? @$operands : ();
1208 my @limits = $limits ? @$limits : ();
1209 my @sort_by = $sort_by ? @$sort_by : ();
1211 my $stemming = C4::Context->preference("QueryStemming") || 0;
1212 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1213 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1214 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1215 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1217 # no stemming/weight/fuzzy in NoZebra
1218 if ( C4::Context->preference("NoZebra") ) {
1222 $auto_truncation = 0;
1225 my $query = $operands[0];
1226 my $simple_query = $operands[0];
1228 # initialize the variables we're passing back
1237 my $stopwords_removed; # flag to determine if stopwords have been removed
1240 my $cclindexes = getIndexes();
1241 if ( $query !~ /\s*ccl=/ ) {
1242 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1244 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1246 $query = "ccl=$query" if $cclq;
1249 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1251 if ( $query =~ /^ccl=/ ) {
1253 # This is needed otherwise ccl= and &limit won't work together, and
1254 # this happens when selecting a subject on the opac-detail page
1255 @limits = grep {!/^$/} @limits;
1257 $q .= ' and '.join(' and ', @limits);
1259 return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1261 if ( $query =~ /^cql=/ ) {
1262 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1264 if ( $query =~ /^pqf=/ ) {
1265 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1268 # pass nested queries directly
1269 # FIXME: need better handling of some of these variables in this case
1270 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1271 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1272 # if ( $query =~ /(\(|\))/ ) {
1274 # undef, $query, $simple_query, $query_cgi,
1275 # $query, $limit, $limit_cgi, $limit_desc,
1276 # $stopwords_removed, 'ccl'
1280 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1281 # query operands and indexes and add stemming, truncation, field weighting, etc.
1282 # Once we do so, we'll end up with a value in $query, just like if we had an
1283 # incoming $query from the user
1286 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1287 my $previous_operand
1288 ; # a flag used to keep track if there was a previous query
1289 # if there was, we can apply the current operator
1291 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1293 # COMBINE OPERANDS, INDEXES AND OPERATORS
1294 if ( $operands[$i] ) {
1295 $operands[$i]=~s/^\s+//;
1297 # A flag to determine whether or not to add the index to the query
1300 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1301 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1304 $remove_stopwords = 0;
1306 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1308 my $operand = $operands[$i];
1309 my $index = $indexes[$i];
1311 # Add index-specific attributes
1312 # Date of Publication
1313 if ( $index eq 'yr' ) {
1314 $index .= ",st-numeric";
1316 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1319 # Date of Acquisition
1320 elsif ( $index eq 'acqdate' ) {
1321 $index .= ",st-date-normalized";
1323 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1325 # ISBN,ISSN,Standard Number, don't need special treatment
1326 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1328 $stemming, $auto_truncation,
1329 $weight_fields, $fuzzy_enabled,
1331 ) = ( 0, 0, 0, 0, 0 );
1339 # Set default structure attribute (word list)
1340 my $struct_attr = q{};
1341 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1342 $struct_attr = ",wrdl";
1345 # Some helpful index variants
1346 my $index_plus = $index . $struct_attr . ':';
1347 my $index_plus_comma = $index . $struct_attr . ',';
1350 if ($remove_stopwords) {
1351 ( $operand, $stopwords_removed ) =
1352 _remove_stopwords( $operand, $index );
1353 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1354 warn "REMOVED STOPWORDS: @$stopwords_removed"
1355 if ( $stopwords_removed && $DEBUG );
1358 if ($auto_truncation){
1359 unless ( $index =~ /(st-|phr|ext)/ ) {
1360 #FIXME only valid with LTR scripts
1361 $operand=join(" ",map{
1362 (index($_,"*")>0?"$_":"$_*")
1363 }split (/\s+/,$operand));
1364 warn $operand if $DEBUG;
1369 my $truncated_operand;
1370 my( $nontruncated, $righttruncated, $lefttruncated,
1371 $rightlefttruncated, $regexpr
1372 ) = _detect_truncation( $operand, $index );
1374 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1379 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1380 scalar(@$rightlefttruncated) > 0 )
1383 # Don't field weight or add the index to the query, we do it here
1385 undef $weight_fields;
1386 my $previous_truncation_operand;
1387 if (scalar @$nontruncated) {
1388 $truncated_operand .= "$index_plus @$nontruncated ";
1389 $previous_truncation_operand = 1;
1391 if (scalar @$righttruncated) {
1392 $truncated_operand .= "and " if $previous_truncation_operand;
1393 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1394 $previous_truncation_operand = 1;
1396 if (scalar @$lefttruncated) {
1397 $truncated_operand .= "and " if $previous_truncation_operand;
1398 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1399 $previous_truncation_operand = 1;
1401 if (scalar @$rightlefttruncated) {
1402 $truncated_operand .= "and " if $previous_truncation_operand;
1403 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1404 $previous_truncation_operand = 1;
1407 $operand = $truncated_operand if $truncated_operand;
1408 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1411 my $stemmed_operand;
1412 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1415 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1417 # Handle Field Weighting
1418 my $weighted_operand;
1419 if ($weight_fields) {
1420 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1421 $operand = $weighted_operand;
1425 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1427 # If there's a previous operand, we need to add an operator
1428 if ($previous_operand) {
1430 # User-specified operator
1431 if ( $operators[ $i - 1 ] ) {
1432 $query .= " $operators[$i-1] ";
1433 $query .= " $index_plus " unless $indexes_set;
1434 $query .= " $operand";
1435 $query_cgi .= "&op=$operators[$i-1]";
1436 $query_cgi .= "&idx=$index" if $index;
1437 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1439 " $operators[$i-1] $index_plus $operands[$i]";
1442 # Default operator is and
1445 $query .= "$index_plus " unless $indexes_set;
1446 $query .= "$operand";
1447 $query_cgi .= "&op=and&idx=$index" if $index;
1448 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1449 $query_desc .= " and $index_plus $operands[$i]";
1453 # There isn't a pervious operand, don't need an operator
1456 # Field-weighted queries already have indexes set
1457 $query .= " $index_plus " unless $indexes_set;
1459 $query_desc .= " $index_plus $operands[$i]";
1460 $query_cgi .= "&idx=$index" if $index;
1461 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1462 $previous_operand = 1;
1467 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1470 my %group_OR_limits;
1471 my $availability_limit;
1472 foreach my $this_limit (@limits) {
1473 next unless $this_limit;
1474 if ( $this_limit =~ /available/ ) {
1476 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1478 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1479 $availability_limit .=
1480 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1481 $limit_cgi .= "&limit=available";
1485 # group_OR_limits, prefixed by mc-
1486 # OR every member of the group
1487 elsif ( $this_limit =~ /mc/ ) {
1488 my ($k,$v) = split(/:/, $this_limit,2);
1489 if ( $k !~ /mc-i(tem)?type/ ) {
1490 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1491 $this_limit =~ tr/"//d;
1492 $this_limit = $k.":\"".$v."\"";
1495 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1496 $limit_desc .= " or " if $group_OR_limits{$k};
1497 $group_OR_limits{$k} .= "$this_limit";
1498 $limit_cgi .= "&limit=$this_limit";
1499 $limit_desc .= " $this_limit";
1502 # Regular old limits
1504 $limit .= " and " if $limit || $query;
1505 $limit .= "$this_limit";
1506 $limit_cgi .= "&limit=$this_limit";
1507 if ($this_limit =~ /^branch:(.+)/) {
1508 my $branchcode = $1;
1509 my $branchname = GetBranchName($branchcode);
1510 if (defined $branchname) {
1511 $limit_desc .= " branch:$branchname";
1513 $limit_desc .= " $this_limit";
1516 $limit_desc .= " $this_limit";
1520 foreach my $k (keys (%group_OR_limits)) {
1521 $limit .= " and " if ( $query || $limit );
1522 $limit .= "($group_OR_limits{$k})";
1524 if ($availability_limit) {
1525 $limit .= " and " if ( $query || $limit );
1526 $limit .= "($availability_limit)";
1529 # Normalize the query and limit strings
1530 # This is flawed , means we can't search anything with : in it
1531 # if user wants to do ccl or cql, start the query with that
1532 # $query =~ s/:/=/g;
1533 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1534 $query =~ s/(?<=(wrdl)):/=/g;
1535 $query =~ s/(?<=(trn|phr)):/=/g;
1537 for ( $query, $query_desc, $limit, $limit_desc ) {
1538 s/ +/ /g; # remove extra spaces
1539 s/^ //g; # remove any beginning spaces
1540 s/ $//g; # remove any ending spaces
1541 s/==/=/g; # remove double == from query
1543 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1545 for ($query_cgi,$simple_query) {
1548 # append the limit to the query
1549 $query .= " " . $limit;
1553 warn "QUERY:" . $query;
1554 warn "QUERY CGI:" . $query_cgi;
1555 warn "QUERY DESC:" . $query_desc;
1556 warn "LIMIT:" . $limit;
1557 warn "LIMIT CGI:" . $limit_cgi;
1558 warn "LIMIT DESC:" . $limit_desc;
1559 warn "---------\nLeave buildQuery\n---------";
1562 undef, $query, $simple_query, $query_cgi,
1563 $query_desc, $limit, $limit_cgi, $limit_desc,
1564 $stopwords_removed, $query_type
1568 =head2 searchResults
1570 my @search_results = searchResults($search_context, $searchdesc, $hits,
1571 $results_per_page, $offset, $scan,
1574 Format results in a form suitable for passing to the template
1578 # IMO this subroutine is pretty messy still -- it's responsible for
1579 # building the HTML output for the template
1581 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1582 my $dbh = C4::Context->dbh;
1587 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1588 my ($is_opac, $hidelostitems);
1589 if ($search_context eq 'opac') {
1590 $hidelostitems = C4::Context->preference('hidelostitems');
1594 #Build branchnames hash
1596 #get branch information.....
1598 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1600 while ( my $bdata = $bsth->fetchrow_hashref ) {
1601 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1603 # FIXME - We build an authorised values hash here, using the default framework
1604 # though it is possible to have different authvals for different fws.
1606 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1608 # get notforloan authorised value list (see $shelflocations FIXME)
1609 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1611 #Build itemtype hash
1612 #find itemtype & itemtype image
1616 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1619 while ( my $bdata = $bsth->fetchrow_hashref ) {
1620 foreach (qw(description imageurl summary notforloan)) {
1621 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1625 #search item field code
1626 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1628 ## find column names of items related to MARC
1629 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1631 my %subfieldstosearch;
1632 while ( ( my $column ) = $sth2->fetchrow ) {
1633 my ( $tagfield, $tagsubfield ) =
1634 &GetMarcFromKohaField( "items." . $column, "" );
1635 $subfieldstosearch{$column} = $tagsubfield;
1638 # handle which records to actually retrieve
1640 if ( $hits && $offset + $results_per_page <= $hits ) {
1641 $times = $offset + $results_per_page;
1644 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1647 my $marcflavour = C4::Context->preference("marcflavour");
1648 # We get the biblionumber position in MARC
1649 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1651 # loop through all of the records we've retrieved
1652 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1653 my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1657 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1658 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1659 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1660 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1661 $oldbiblio->{result_number} = $i + 1;
1663 # add imageurl to itemtype if there is one
1664 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1666 $oldbiblio->{'authorised_value_images'} = ($search_context eq 'opac' && C4::Context->preference('AuthorisedValueImages')) || ($search_context eq 'intranet' && C4::Context->preference('StaffAuthorisedValueImages')) ? C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) ) : [];
1667 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1668 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1669 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1670 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1671 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1673 # edition information, if any
1674 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1675 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1676 # Build summary if there is one (the summary is defined in the itemtypes table)
1677 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1678 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1679 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1680 my @fields = $marcrecord->fields();
1683 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1685 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1686 $tag =~ /(.{3})(.)/;
1687 if($marcrecord->field($1)){
1688 my @abc = $marcrecord->field($1)->subfield($2);
1689 $tags->{$tag} = $#abc + 1 ;
1693 # We catch how many times to repeat this line
1695 foreach my $tag (keys(%$tags)){
1696 $max = $tags->{$tag} if($tags->{$tag} > $max);
1699 # we replace, and repeat each line
1700 for (my $i = 0 ; $i < $max ; $i++){
1701 my $newline = $line;
1703 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1704 $tag =~ /(.{3})(.)/;
1706 if($marcrecord->field($1)){
1707 my @repl = $marcrecord->field($1)->subfield($2);
1708 my $subfieldvalue = $repl[$i];
1710 if (! utf8::is_utf8($subfieldvalue)) {
1711 utf8::decode($subfieldvalue);
1714 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1717 $newsummary .= "$newline\n";
1721 $newsummary =~ s/\[(.*?)]//g;
1722 $newsummary =~ s/\n/<br\/>/g;
1723 $oldbiblio->{summary} = $newsummary;
1726 # Pull out the items fields
1727 my @fields = $marcrecord->field($itemtag);
1728 my $marcflavor = C4::Context->preference("marcflavour");
1729 # adding linked items that belong to host records
1730 my $analyticsfield = '773';
1731 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1732 $analyticsfield = '773';
1733 } elsif ($marcflavor eq 'UNIMARC') {
1734 $analyticsfield = '461';
1736 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1737 my $hostbiblionumber = $hostfield->subfield("0");
1738 my $linkeditemnumber = $hostfield->subfield("9");
1739 if(!$hostbiblionumber eq undef){
1740 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1741 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1742 if(!$hostbiblio eq undef){
1743 my @hostitems = $hostbiblio->field($itemfield);
1744 foreach my $hostitem (@hostitems){
1745 if ($hostitem->subfield("9") eq $linkeditemnumber){
1746 my $linkeditem =$hostitem;
1747 # append linked items if they exist
1748 if (!$linkeditem eq undef){
1749 push (@fields, $linkeditem);}
1756 # Setting item statuses for display
1757 my @available_items_loop;
1758 my @onloan_items_loop;
1759 my @other_items_loop;
1761 my $available_items;
1765 my $ordered_count = 0;
1766 my $available_count = 0;
1767 my $onloan_count = 0;
1768 my $longoverdue_count = 0;
1769 my $other_count = 0;
1770 my $wthdrawn_count = 0;
1771 my $itemlost_count = 0;
1772 my $hideatopac_count = 0;
1773 my $itembinding_count = 0;
1774 my $itemdamaged_count = 0;
1775 my $item_in_transit_count = 0;
1776 my $can_place_holds = 0;
1777 my $item_onhold_count = 0;
1778 my $items_count = scalar(@fields);
1779 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1780 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1781 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
1783 # loop through every item
1784 foreach my $field (@fields) {
1787 # populate the items hash
1788 foreach my $code ( keys %subfieldstosearch ) {
1789 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1791 $item->{description} = $itemtypes{ $item->{itype} }{description};
1795 # hidden because lost
1796 if ($hidelostitems && $item->{itemlost}) {
1797 $hideatopac_count++;
1800 # hidden based on OpacHiddenItems syspref
1801 my @hi = C4::Items::GetHiddenItemnumbers($item);
1803 push @hiddenitems, @hi;
1804 $hideatopac_count++;
1809 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1810 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1812 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1813 if ($item->{$hbranch}) {
1814 $item->{'branchname'} = $branches{$item->{$hbranch}};
1816 elsif ($item->{$otherbranch}) { # Last resort
1817 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1820 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1821 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1822 my $userenv = C4::Context->userenv;
1823 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1825 my $key = $prefix . $item->{onloan} . $item->{barcode};
1826 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1827 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1828 $onloan_items->{$key}->{branchname} = $item->{branchname};
1829 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1830 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1831 $onloan_items->{$key}->{description} = $item->{description};
1832 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1833 # if something's checked out and lost, mark it as 'long overdue'
1834 if ( $item->{itemlost} ) {
1835 $onloan_items->{$prefix}->{longoverdue}++;
1836 $longoverdue_count++;
1837 } else { # can place holds as long as item isn't lost
1838 $can_place_holds = 1;
1842 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1846 if ( $item->{notforloan} < 0 ) {
1850 # is item in transit?
1851 my $transfertwhen = '';
1852 my ($transfertfrom, $transfertto);
1854 # is item on the reserve shelf?
1855 my $reservestatus = '';
1858 unless ($item->{wthdrawn}
1859 || $item->{itemlost}
1861 || $item->{notforloan}
1862 || $items_count > 20) {
1864 # A couple heuristics to limit how many times
1865 # we query the database for item transfer information, sacrificing
1866 # accuracy in some cases for speed;
1868 # 1. don't query if item has one of the other statuses
1869 # 2. don't check transit status if the bib has
1870 # more than 20 items
1872 # FIXME: to avoid having the query the database like this, and to make
1873 # the in transit status count as unavailable for search limiting,
1874 # should map transit status to record indexed in Zebra.
1876 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1877 ($reservestatus, $reserveitem, undef) = C4::Reserves::CheckReserves($item->{itemnumber});
1880 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1881 if ( $item->{wthdrawn}
1882 || $item->{itemlost}
1884 || $item->{notforloan}
1885 || $reservestatus eq 'Waiting'
1886 || ($transfertwhen ne ''))
1888 $wthdrawn_count++ if $item->{wthdrawn};
1889 $itemlost_count++ if $item->{itemlost};
1890 $itemdamaged_count++ if $item->{damaged};
1891 $item_in_transit_count++ if $transfertwhen ne '';
1892 $item_onhold_count++ if $reservestatus eq 'Waiting';
1893 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1895 # can place hold on item ?
1896 if ( !$item->{itemlost} ) {
1897 if ( !$item->{wthdrawn} ){
1898 if ( $item->{damaged} ){
1899 if ( C4::Context->preference('AllowHoldsOnDamagedItems') ){
1900 # can place a hold on a damaged item if AllowHoldsOnDamagedItems is true
1901 if ( ( !$item->{notforloan} || $item->{notforloan} < 0 ) ){
1902 # item is either for loan or has notforloan < 0
1903 $can_place_holds = 1;
1906 } elsif ( $item->{notforloan} < 0 ) {
1907 # item is not damaged and notforloan is < 0
1908 $can_place_holds = 1;
1915 my $key = $prefix . $item->{status};
1916 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1917 $other_items->{$key}->{$_} = $item->{$_};
1919 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1920 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1921 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
1922 $other_items->{$key}->{count}++ if $item->{$hbranch};
1923 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1924 $other_items->{$key}->{description} = $item->{description};
1925 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1929 $can_place_holds = 1;
1931 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1932 foreach (qw(branchname itemcallnumber description)) {
1933 $available_items->{$prefix}->{$_} = $item->{$_};
1935 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1936 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1939 } # notforloan, item level and biblioitem level
1941 # if all items are hidden, do not show the record
1942 if ($items_count > 0 && $hideatopac_count == $items_count) {
1946 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1947 for my $key ( sort keys %$onloan_items ) {
1948 (++$onloanitemscount > $maxitems) and last;
1949 push @onloan_items_loop, $onloan_items->{$key};
1951 for my $key ( sort keys %$other_items ) {
1952 (++$otheritemscount > $maxitems) and last;
1953 push @other_items_loop, $other_items->{$key};
1955 for my $key ( sort keys %$available_items ) {
1956 (++$availableitemscount > $maxitems) and last;
1957 push @available_items_loop, $available_items->{$key}
1960 # XSLT processing of some stuff
1962 SetUTF8Flag($marcrecord);
1963 warn $marcrecord->as_formatted if $DEBUG;
1964 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1965 if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
1966 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
1967 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1970 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1971 if (!C4::Context->preference("item-level_itypes")) {
1972 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1973 $can_place_holds = 0;
1976 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1977 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1978 $oldbiblio->{items_count} = $items_count;
1979 $oldbiblio->{available_items_loop} = \@available_items_loop;
1980 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1981 $oldbiblio->{other_items_loop} = \@other_items_loop;
1982 $oldbiblio->{availablecount} = $available_count;
1983 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1984 $oldbiblio->{onloancount} = $onloan_count;
1985 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1986 $oldbiblio->{othercount} = $other_count;
1987 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1988 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1989 $oldbiblio->{itemlostcount} = $itemlost_count;
1990 $oldbiblio->{damagedcount} = $itemdamaged_count;
1991 $oldbiblio->{intransitcount} = $item_in_transit_count;
1992 $oldbiblio->{onholdcount} = $item_onhold_count;
1993 $oldbiblio->{orderedcount} = $ordered_count;
1995 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1996 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1997 my $subfields = substr $fieldspec, 3;
1998 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1999 my @alternateholdingsinfo = ();
2000 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2001 my $alternateholdingscount = 0;
2003 for my $field (@holdingsfields) {
2004 my %holding = ( holding => '' );
2005 my $havesubfield = 0;
2006 for my $subfield ($field->subfields()) {
2007 if ((index $subfields, $$subfield[0]) >= 0) {
2008 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2009 $holding{'holding'} .= $$subfield[1];
2013 if ($havesubfield) {
2014 push(@alternateholdingsinfo, \%holding);
2015 $alternateholdingscount++;
2019 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2020 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2023 push( @newresults, $oldbiblio );
2029 =head2 SearchAcquisitions
2030 Search for acquisitions
2033 sub SearchAcquisitions{
2034 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2036 my $dbh=C4::Context->dbh;
2037 # Variable initialization
2041 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2042 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2043 WHERE dateaccessioned BETWEEN ? AND ?
2046 my (@params,@loopcriteria);
2048 push @params, $datebegin->output("iso");
2049 push @params, $dateend->output("iso");
2051 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2052 if(C4::Context->preference("item-level_itypes")){
2053 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2055 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2057 push @params, @$itemtypes;
2060 if ($criteria =~/itemtype/){
2061 if(C4::Context->preference("item-level_itypes")){
2062 $str .= "AND items.itype=? ";
2064 $str .= "AND biblioitems.itemtype=? ";
2067 if(scalar(@$itemtypes) == 0){
2068 my $itypes = GetItemTypes();
2069 for my $key (keys %$itypes){
2070 push @$itemtypes, $key;
2074 @loopcriteria= @$itemtypes;
2075 }elsif ($criteria=~/itemcallnumber/){
2076 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2077 OR items.itemcallnumber is NULL
2078 OR items.itemcallnumber = '')";
2080 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2082 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2083 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2086 if ($orderby =~ /date_desc/){
2087 $str.=" ORDER BY dateaccessioned DESC";
2089 $str.=" ORDER BY title";
2092 my $qdataacquisitions=$dbh->prepare($str);
2094 my @loopacquisitions;
2095 foreach my $value(@loopcriteria){
2096 push @params,$value;
2098 $cell{"title"}=$value;
2099 $cell{"titlecode"}=$value;
2101 eval{$qdataacquisitions->execute(@params);};
2103 if ($@){ warn "recentacquisitions Error :$@";}
2106 while (my $data=$qdataacquisitions->fetchrow_hashref){
2107 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2109 $cell{"loopdata"}=\@loopdata;
2111 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2114 $qdataacquisitions->finish;
2115 return \@loopacquisitions;
2117 #----------------------------------------------------------------------
2119 # Non-Zebra GetRecords#
2120 #----------------------------------------------------------------------
2124 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
2130 $query, $simple_query, $sort_by_ref, $servers_ref,
2131 $results_per_page, $offset, $expanded_facet, $branches,
2134 warn "query =$query" if $DEBUG;
2135 my $result = NZanalyse($query);
2136 warn "results =$result" if $DEBUG;
2138 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
2144 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
2145 the list is built from an inverted index in the nozebra SQL table
2146 note that title is here only for convenience : the sorting will be very fast when requested on title
2147 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
2152 my ( $string, $server ) = @_;
2153 # warn "---------" if $DEBUG;
2154 warn " NZanalyse" if $DEBUG;
2155 # warn "---------" if $DEBUG;
2157 # $server contains biblioserver or authorities, depending on what we search on.
2158 #warn "querying : $string on $server";
2159 $server = 'biblioserver' unless $server;
2161 # if we have a ", replace the content to discard temporarily any and/or/not inside
2163 if ( $string =~ /"/ ) {
2164 $string =~ s/"(.*?)"/__X__/;
2166 warn "commacontent : $commacontent" if $DEBUG;
2169 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
2170 # then, call again NZanalyse with $left and $right
2171 # (recursive until we find a leaf (=> something without and/or/not)
2172 # delete repeated operator... Would then go in infinite loop
2173 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
2176 #process parenthesis before.
2177 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
2180 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
2182 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
2184 my $leftresult = NZanalyse( $left, $server );
2186 my $rightresult = NZanalyse( $right, $server );
2188 # OK, we have the results for right and left part of the query
2189 # depending of operand, intersect, union or exclude both lists
2190 # to get a result list
2191 if ( $operator eq ' and ' ) {
2192 return NZoperatorAND($leftresult,$rightresult);
2194 elsif ( $operator eq ' or ' ) {
2196 # just merge the 2 strings
2197 return $leftresult . $rightresult;
2199 elsif ( $operator eq ' not ' ) {
2200 return NZoperatorNOT($leftresult,$rightresult);
2204 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2208 warn "string :" . $string if $DEBUG;
2212 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2215 $operator = lc($2); # FIXME: and/or/not are operators, not operands
2217 warn "no parenthesis. left : $left operator: $operator right: $right"
2220 # it's not a leaf, we have a and/or/not
2223 # reintroduce comma content if needed
2224 $right =~ s/__X__/"$commacontent"/ if $commacontent;
2225 $left =~ s/__X__/"$commacontent"/ if $commacontent;
2226 warn "node : $left / $operator / $right\n" if $DEBUG;
2227 my $leftresult = NZanalyse( $left, $server );
2228 my $rightresult = NZanalyse( $right, $server );
2229 warn " leftresult : $leftresult" if $DEBUG;
2230 warn " rightresult : $rightresult" if $DEBUG;
2231 # OK, we have the results for right and left part of the query
2232 # depending of operand, intersect, union or exclude both lists
2233 # to get a result list
2234 if ( $operator eq ' and ' ) {
2235 return NZoperatorAND($leftresult,$rightresult);
2237 elsif ( $operator eq ' or ' ) {
2239 # just merge the 2 strings
2240 return $leftresult . $rightresult;
2242 elsif ( $operator eq ' not ' ) {
2243 return NZoperatorNOT($leftresult,$rightresult);
2247 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2248 die "error : operand unknown : $operator for $string";
2251 # it's a leaf, do the real SQL query and return the result
2254 $string =~ s/__X__/"$commacontent"/ if $commacontent;
2255 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2256 #remove trailing blank at the beginning
2258 warn "leaf:$string" if $DEBUG;
2260 # parse the string in in operator/operand/value again
2264 if ($string =~ /(.*)(>=|<=)(.*)/) {
2271 # warn "handling leaf... left:$left operator:$operator right:$right"
2273 unless ($operator) {
2274 if ($string =~ /(.*)(>|<|=)(.*)/) {
2279 "handling unless (operator)... left:$left operator:$operator right:$right"
2287 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2290 # automatic replace for short operators
2291 $left = 'title' if $left =~ '^ti$';
2292 $left = 'author' if $left =~ '^au$';
2293 $left = 'publisher' if $left =~ '^pb$';
2294 $left = 'subject' if $left =~ '^su$';
2295 $left = 'koha-Auth-Number' if $left =~ '^an$';
2296 $left = 'keyword' if $left =~ '^kw$';
2297 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2298 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2299 my $dbh = C4::Context->dbh;
2300 if ( $operator && $left ne 'keyword' ) {
2301 #do a specific search
2302 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2303 my $sth = $dbh->prepare(
2304 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2306 warn "$left / $operator / $right\n" if $DEBUG;
2308 # split each word, query the DB and build the biblionumbers result
2309 #sanitizing leftpart
2310 $left =~ s/^\s+|\s+$//;
2311 foreach ( split / /, $right ) {
2313 $_ =~ s/^\s+|\s+$//;
2315 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2316 $sth->execute( $server, $left, $_ )
2317 or warn "execute failed: $!";
2318 while ( my ( $line, $value ) = $sth->fetchrow ) {
2320 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2321 # otherwise, fill the result
2322 $biblionumbers .= $line
2323 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2324 warn "result : $value "
2325 . ( $right =~ /\d/ ) . "=="
2326 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2329 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2331 warn "NZAND" if $DEBUG;
2332 $results = NZoperatorAND($biblionumbers,$results);
2334 $results = $biblionumbers;
2339 #do a complete search (all indexes), if index='kw' do complete search too.
2340 my $sth = $dbh->prepare(
2341 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2344 # split each word, query the DB and build the biblionumbers result
2345 foreach ( split / /, $string ) {
2346 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2347 warn "search on all indexes on $_" if $DEBUG;
2350 $sth->execute( $server, $_ );
2351 while ( my $line = $sth->fetchrow ) {
2352 $biblionumbers .= $line;
2355 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2357 $results = NZoperatorAND($biblionumbers,$results);
2360 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2361 $results = $biblionumbers;
2365 warn "return : $results for LEAF : $string" if $DEBUG;
2368 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2372 my ($rightresult, $leftresult)=@_;
2374 my @leftresult = split /;/, $leftresult;
2375 warn " @leftresult / $rightresult \n" if $DEBUG;
2377 # my @rightresult = split /;/,$leftresult;
2380 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2381 # the result is stored twice, to have the same weight for AND than OR.
2382 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2383 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2384 foreach (@leftresult) {
2387 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2388 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2389 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2391 "$value-$countvalue;$value-$countvalue;";
2394 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2395 return $finalresult;
2399 my ($rightresult, $leftresult)=@_;
2400 return $rightresult.$leftresult;
2404 my ($leftresult, $rightresult)=@_;
2406 my @leftresult = split /;/, $leftresult;
2408 # my @rightresult = split /;/,$leftresult;
2410 foreach (@leftresult) {
2412 $value=$1 if $value=~m/(.*)-\d+$/;
2413 unless ($rightresult =~ "$value-") {
2414 $finalresult .= "$_;";
2417 return $finalresult;
2422 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2429 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2430 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2432 # order title asc by default
2433 # $ordering = '1=36 <i' unless $ordering;
2434 $results_per_page = 20 unless $results_per_page;
2435 $offset = 0 unless $offset;
2436 my $dbh = C4::Context->dbh;
2439 # order by POPULARITY
2441 if ( $ordering =~ /popularity/ ) {
2445 # popularity is not in MARC record, it's builded from a specific query
2447 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2448 foreach ( split /;/, $biblionumbers ) {
2449 my ( $biblionumber, $title ) = split /,/, $_;
2450 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2451 $sth->execute($biblionumber);
2452 my $popularity = $sth->fetchrow || 0;
2454 # hint : the key is popularity.title because we can have
2455 # many results with the same popularity. In this case, sub-ordering is done by title
2456 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2457 # (un-frequent, I agree, but we won't forget anything that way ;-)
2458 $popularity{ sprintf( "%10d", $popularity ) . $title
2459 . $biblionumber } = $biblionumber;
2462 # sort the hash and return the same structure as GetRecords (Zebra querying)
2465 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2466 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2467 $result_hash->{'RECORDS'}[ $numbers++ ] =
2468 $result{ $popularity{$key} }->as_usmarc();
2471 else { # sort popularity ASC
2472 foreach my $key ( sort ( keys %popularity ) ) {
2473 $result_hash->{'RECORDS'}[ $numbers++ ] =
2474 $result{ $popularity{$key} }->as_usmarc();
2477 my $finalresult = ();
2478 $result_hash->{'hits'} = $numbers;
2479 $finalresult->{'biblioserver'} = $result_hash;
2480 return $finalresult;
2486 elsif ( $ordering =~ /author/ ) {
2488 foreach ( split /;/, $biblionumbers ) {
2489 my ( $biblionumber, $title ) = split /,/, $_;
2490 my $record = GetMarcBiblio($biblionumber);
2492 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2493 $author = $record->subfield( '200', 'f' );
2494 $author = $record->subfield( '700', 'a' ) unless $author;
2497 $author = $record->subfield( '100', 'a' );
2500 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2501 # and we don't want to get only 1 result for each of them !!!
2502 $result{ $author . $biblionumber } = $record;
2505 # sort the hash and return the same structure as GetRecords (Zebra querying)
2508 if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) { # sort by author desc
2509 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2510 $result_hash->{'RECORDS'}[ $numbers++ ] =
2511 $result{$key}->as_usmarc();
2514 else { # sort by author ASC
2515 foreach my $key ( sort ( keys %result ) ) {
2516 $result_hash->{'RECORDS'}[ $numbers++ ] =
2517 $result{$key}->as_usmarc();
2520 my $finalresult = ();
2521 $result_hash->{'hits'} = $numbers;
2522 $finalresult->{'biblioserver'} = $result_hash;
2523 return $finalresult;
2526 # ORDER BY callnumber
2529 elsif ( $ordering =~ /callnumber/ ) {
2531 foreach ( split /;/, $biblionumbers ) {
2532 my ( $biblionumber, $title ) = split /,/, $_;
2533 my $record = GetMarcBiblio($biblionumber);
2535 my $frameworkcode = GetFrameworkCode($biblionumber);
2536 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2537 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2538 unless $callnumber_tag;
2539 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2540 $callnumber = $record->subfield( '200', 'f' );
2542 $callnumber = $record->subfield( '100', 'a' );
2545 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2546 # and we don't want to get only 1 result for each of them !!!
2547 $result{ $callnumber . $biblionumber } = $record;
2550 # sort the hash and return the same structure as GetRecords (Zebra querying)
2553 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2554 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2555 $result_hash->{'RECORDS'}[ $numbers++ ] =
2556 $result{$key}->as_usmarc();
2559 else { # sort by title ASC
2560 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2561 $result_hash->{'RECORDS'}[ $numbers++ ] =
2562 $result{$key}->as_usmarc();
2565 my $finalresult = ();
2566 $result_hash->{'hits'} = $numbers;
2567 $finalresult->{'biblioserver'} = $result_hash;
2568 return $finalresult;
2570 elsif ( $ordering =~ /pubdate/ ) { #pub year
2572 foreach ( split /;/, $biblionumbers ) {
2573 my ( $biblionumber, $title ) = split /,/, $_;
2574 my $record = GetMarcBiblio($biblionumber);
2575 my ( $publicationyear_tag, $publicationyear_subfield ) =
2576 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2577 my $publicationyear =
2578 $record->subfield( $publicationyear_tag,
2579 $publicationyear_subfield );
2581 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2582 # and we don't want to get only 1 result for each of them !!!
2583 $result{ $publicationyear . $biblionumber } = $record;
2586 # sort the hash and return the same structure as GetRecords (Zebra querying)
2589 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2590 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2591 $result_hash->{'RECORDS'}[ $numbers++ ] =
2592 $result{$key}->as_usmarc();
2595 else { # sort by pub year ASC
2596 foreach my $key ( sort ( keys %result ) ) {
2597 $result_hash->{'RECORDS'}[ $numbers++ ] =
2598 $result{$key}->as_usmarc();
2601 my $finalresult = ();
2602 $result_hash->{'hits'} = $numbers;
2603 $finalresult->{'biblioserver'} = $result_hash;
2604 return $finalresult;
2610 elsif ( $ordering =~ /title/ ) {
2612 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2614 foreach ( split /;/, $biblionumbers ) {
2615 my ( $biblionumber, $title ) = split /,/, $_;
2617 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2618 # and we don't want to get only 1 result for each of them !!!
2619 # hint & speed improvement : we can order without reading the record
2620 # so order, and read records only for the requested page !
2621 $result{ $title . $biblionumber } = $biblionumber;
2624 # sort the hash and return the same structure as GetRecords (Zebra querying)
2627 if ( $ordering eq 'title_az' ) { # sort by title desc
2628 foreach my $key ( sort ( keys %result ) ) {
2629 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2632 else { # sort by title ASC
2633 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2634 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2638 # limit the $results_per_page to result size if it's more
2639 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2641 # for the requested page, replace biblionumber by the complete record
2642 # speed improvement : avoid reading too much things
2644 my $counter = $offset ;
2645 $counter <= $offset + $results_per_page ;
2649 $result_hash->{'RECORDS'}[$counter] =
2650 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2652 my $finalresult = ();
2653 $result_hash->{'hits'} = $numbers;
2654 $finalresult->{'biblioserver'} = $result_hash;
2655 return $finalresult;
2662 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2665 foreach ( split /;/, $biblionumbers ) {
2666 my ( $biblionumber, $title ) = split /,/, $_;
2667 $title =~ /(.*)-(\d)/;
2672 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2673 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2674 # biblio N has ranking = 6
2675 $count_ranking{$biblionumber} += $ranking;
2678 # build the result by "inverting" the count_ranking hash
2679 # 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
2681 foreach ( keys %count_ranking ) {
2682 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2685 # sort the hash and return the same structure as GetRecords (Zebra querying)
2688 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2689 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2692 # limit the $results_per_page to result size if it's more
2693 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2695 # for the requested page, replace biblionumber by the complete record
2696 # speed improvement : avoid reading too much things
2698 my $counter = $offset ;
2699 $counter <= $offset + $results_per_page ;
2703 $result_hash->{'RECORDS'}[$counter] =
2704 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2705 if $result_hash->{'RECORDS'}[$counter];
2707 my $finalresult = ();
2708 $result_hash->{'hits'} = $numbers;
2709 $finalresult->{'biblioserver'} = $result_hash;
2710 return $finalresult;
2714 =head2 enabled_staff_search_views
2716 %hash = enabled_staff_search_views()
2718 This function returns a hash that contains three flags obtained from the system
2719 preferences, used to determine whether a particular staff search results view
2724 =item C<Output arg:>
2726 * $hash{can_view_MARC} is true only if the MARC view is enabled
2727 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2728 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2730 =item C<usage in the script:>
2734 $template->param ( C4::Search::enabled_staff_search_views );
2738 sub enabled_staff_search_views
2741 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2742 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2743 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2747 sub AddSearchHistory{
2748 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2749 my $dbh = C4::Context->dbh;
2751 # Add the request the user just made
2752 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2753 my $sth = $dbh->prepare($sql);
2754 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2755 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2758 sub GetSearchHistory{
2759 my ($borrowernumber,$session)=@_;
2760 my $dbh = C4::Context->dbh;
2762 # Add the request the user just made
2763 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2764 my $sth = $dbh->prepare($query);
2765 $sth->execute($borrowernumber, $session);
2766 return $sth->fetchall_hashref({});
2769 =head2 z3950_search_args
2771 $arrayref = z3950_search_args($matchpoints)
2773 This function returns an array reference that contains the search parameters to be
2774 passed to the Z39.50 search script (z3950_search.pl). The array elements
2775 are hash refs whose keys are name, value and encvalue, and whose values are the
2776 name of a search parameter, the value of that search parameter and the URL encoded
2777 value of that parameter.
2779 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2781 The search parameter values are obtained from the bibliographic record whose
2782 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2784 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2785 a general purpose search argument. In this case, the returned array contains only
2786 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2788 If a search parameter value is undefined or empty, it is not included in the returned
2791 The returned array reference may be passed directly to the template parameters.
2795 =item C<Output arg:>
2797 * $array containing hash refs as described above
2799 =item C<usage in the script:>
2803 $data = Biblio::GetBiblioData($bibno);
2804 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2808 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2812 sub z3950_search_args {
2814 my $isbn = Business::ISBN->new($bibrec);
2816 if (defined $isbn && $isbn->is_valid)
2818 $bibrec = { isbn => $bibrec } if !ref $bibrec;
2821 $bibrec = { title => $bibrec } if !ref $bibrec;
2824 for my $field (qw/ lccn isbn issn title author dewey subject /)
2826 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2827 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2832 =head2 GetDistinctValues($field);
2834 C<$field> is a reference to the fields array
2838 sub GetDistinctValues {
2839 my ($fieldname,$string)=@_;
2840 # returns a reference to a hash of references to branches...
2841 if ($fieldname=~/\./){
2842 my ($table,$column)=split /\./, $fieldname;
2843 my $dbh = C4::Context->dbh;
2844 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2845 my $sth = $dbh->prepare("select DISTINCT($column) as value, count(*) as cnt from $table ".($string?" where $column like \"$string%\"":"")."group by value order by $column ");
2847 my $elements=$sth->fetchall_arrayref({});
2852 my @servers=qw<biblioserver authorityserver>;
2853 my (@zconns,@results);
2854 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2855 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2858 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2861 # The big moment: asynchronously retrieve results from all servers
2867 my ( $i, $size ) = @_;
2868 for ( my $j = 0 ; $j < $size ; $j++ ) {
2870 @hashscan{qw(value cnt)} =
2871 $results[ $i - 1 ]->display_term($j);
2872 push @elements, \%hashscan;
2880 =head2 _ZOOM_event_loop
2882 _ZOOM_event_loop(\@zconns, \@results, sub {
2883 my ( $i, $size ) = @_;
2887 Processes a ZOOM event loop and passes control to a closure for
2888 processing the results, and destroying the resultsets.
2892 sub _ZOOM_event_loop {
2893 my ($zconns, $results, $callback) = @_;
2894 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2895 my $ev = $zconns->[ $i - 1 ]->last_event();
2896 if ( $ev == ZOOM::Event::ZEND ) {
2897 next unless $results->[ $i - 1 ];
2898 my $size = $results->[ $i - 1 ]->size();
2900 $callback->($i, $size);
2905 foreach my $result (@$results) {
2911 END { } # module clean-up code here (global destructor)
2918 Koha Development Team <http://koha-community.org/>