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
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
25 use C4::Search::PazPar2;
27 use C4::Dates qw(format_date);
30 use C4::Reserves; # CheckReserves
35 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
37 # set the version for version checking
40 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
45 C4::Search - Functions for searching the Koha catalog.
49 See opac/opac-search.pl or catalogue/search.pl for example of usage
53 This module provides searching functions for Koha's bibliographic databases
71 #FIXME: i had to add BiblioAddAuthorities here because in Biblios.pm it caused circular dependencies (C4::Search uses C4::Biblio, and BiblioAddAuthorities uses SimpleSearch from C4::Search)
73 # make all your functions, whether exported or not;
77 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
79 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
85 my $dbh = C4::Context->dbh;
86 my $result = TransformMarcToKoha( $dbh, $record, '' );
91 my ( $biblionumber, $title );
93 # search duplicate on ISBN, easy and fast..
95 if ( $result->{isbn} ) {
96 $result->{isbn} =~ s/\(.*$//;
97 $result->{isbn} =~ s/\s+$//;
98 $query = "isbn=$result->{isbn}";
101 $result->{title} =~ s /\\//g;
102 $result->{title} =~ s /\"//g;
103 $result->{title} =~ s /\(//g;
104 $result->{title} =~ s /\)//g;
106 # FIXME: instead of removing operators, could just do
107 # quotes around the value
108 $result->{title} =~ s/(and|or|not)//g;
109 $query = "ti,ext=$result->{title}";
110 $query .= " and itemtype=$result->{itemtype}"
111 if ( $result->{itemtype} );
112 if ( $result->{author} ) {
113 $result->{author} =~ s /\\//g;
114 $result->{author} =~ s /\"//g;
115 $result->{author} =~ s /\(//g;
116 $result->{author} =~ s /\)//g;
118 # remove valid operators
119 $result->{author} =~ s/(and|or|not)//g;
120 $query .= " and au,ext=$result->{author}";
124 # FIXME: add error handling
125 my ( $error, $searchresults ) = SimpleSearch($query); # FIXME :: hardcoded !
127 foreach my $possible_duplicate_record (@$searchresults) {
129 MARC::Record->new_from_usmarc($possible_duplicate_record);
130 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
132 # FIXME :: why 2 $biblionumber ?
134 push @results, $result->{'biblionumber'};
135 push @results, $result->{'title'};
143 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
145 This function provides a simple search API on the bibliographic catalog
151 * $query can be a simple keyword or a complete CCL query
152 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
153 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
154 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
159 * $error is a empty unless an error is detected
160 * \@results is an array of records.
161 * $total_hits is the number of hits that would have been returned with no limit
163 =item C<usage in the script:>
167 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
169 if (defined $error) {
170 $template->param(query_error => $error);
171 warn "error: ".$error;
172 output_html_with_http_headers $input, $cookie, $template->output;
176 my $hits = scalar @$marcresults;
179 for my $i (0..$hits) {
181 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
182 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
184 #build the hash for the template.
185 $resultsloop{title} = $biblio->{'title'};
186 $resultsloop{subtitle} = $biblio->{'subtitle'};
187 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
188 $resultsloop{author} = $biblio->{'author'};
189 $resultsloop{publishercode} = $biblio->{'publishercode'};
190 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
192 push @results, \%resultsloop;
195 $template->param(result=>\@results);
200 my ( $query, $offset, $max_results, $servers ) = @_;
202 if ( C4::Context->preference('NoZebra') ) {
203 my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
206 && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
207 return ( undef, $search_result, scalar($result->{hits}) );
210 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
211 my @servers = defined ( $servers ) ? @$servers : ( "biblioserver" );
217 return ( "No query entered", undef, undef ) unless $query;
219 # Initialize & Search Zebra
220 for ( my $i = 0 ; $i < @servers ; $i++ ) {
222 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
223 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
224 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
228 $zconns[$i]->errmsg() . " ("
229 . $zconns[$i]->errcode() . ") "
230 . $zconns[$i]->addinfo() . " "
231 . $zconns[$i]->diagset();
233 return ( $error, undef, undef ) if $zconns[$i]->errcode();
237 # caught a ZOOM::Exception
241 . $@->addinfo() . " "
244 return ( $error, undef, undef );
247 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
248 my $event = $zconns[ $i - 1 ]->last_event();
249 if ( $event == ZOOM::Event::ZEND ) {
251 my $first_record = defined( $offset ) ? $offset+1 : 1;
252 my $hits = $tmpresults[ $i - 1 ]->size();
253 $total_hits += $hits;
254 my $last_record = $hits;
255 if ( defined $max_results && $offset + $max_results < $hits ) {
256 $last_record = $offset + $max_results;
259 for my $j ( $first_record..$last_record ) {
260 my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
261 push @results, $record;
266 foreach my $result (@tmpresults) {
269 foreach my $zoom_query (@zoom_queries) {
270 $zoom_query->destroy();
273 return ( undef, \@results, $total_hits );
279 ( undef, $results_hashref, \@facets_loop ) = getRecords (
281 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
282 $results_per_page, $offset, $expanded_facet, $branches,
286 The all singing, all dancing, multi-server, asynchronous, scanning,
287 searching, record nabbing, facet-building
289 See verbse embedded documentation.
295 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
296 $results_per_page, $offset, $expanded_facet, $branches,
300 my @servers = @$servers_ref;
301 my @sort_by = @$sort_by_ref;
303 # Initialize variables for the ZOOM connection and results object
307 my $results_hashref = ();
309 # Initialize variables for the faceted results objects
310 my $facets_counter = ();
311 my $facets_info = ();
312 my $facets = getFacets();
314 my @facets_loop; # stores the ref to array of hashes for template facets loop
316 ### LOOP THROUGH THE SERVERS
317 for ( my $i = 0 ; $i < @servers ; $i++ ) {
318 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
320 # perform the search, create the results objects
321 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
322 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
324 #$query_to_use = $simple_query if $scan;
325 warn $simple_query if ( $scan and $DEBUG );
327 # Check if we've got a query_type defined, if so, use it
330 if ($query_type =~ /^ccl/) {
331 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
332 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
333 } elsif ($query_type =~ /^cql/) {
334 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
335 } elsif ($query_type =~ /^pqf/) {
336 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
338 warn "Unknown query_type '$query_type'. Results undetermined.";
341 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
343 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
347 warn "WARNING: query problem with $query_to_use " . $@;
350 # Concatenate the sort_by limits and pass them to the results object
351 # Note: sort will override rank
353 foreach my $sort (@sort_by) {
354 if ( $sort eq "author_az" ) {
355 $sort_by .= "1=1003 <i ";
357 elsif ( $sort eq "author_za" ) {
358 $sort_by .= "1=1003 >i ";
360 elsif ( $sort eq "popularity_asc" ) {
361 $sort_by .= "1=9003 <i ";
363 elsif ( $sort eq "popularity_dsc" ) {
364 $sort_by .= "1=9003 >i ";
366 elsif ( $sort eq "call_number_asc" ) {
367 $sort_by .= "1=20 <i ";
369 elsif ( $sort eq "call_number_dsc" ) {
370 $sort_by .= "1=20 >i ";
372 elsif ( $sort eq "pubdate_asc" ) {
373 $sort_by .= "1=31 <i ";
375 elsif ( $sort eq "pubdate_dsc" ) {
376 $sort_by .= "1=31 >i ";
378 elsif ( $sort eq "acqdate_asc" ) {
379 $sort_by .= "1=32 <i ";
381 elsif ( $sort eq "acqdate_dsc" ) {
382 $sort_by .= "1=32 >i ";
384 elsif ( $sort eq "title_az" ) {
385 $sort_by .= "1=4 <i ";
387 elsif ( $sort eq "title_za" ) {
388 $sort_by .= "1=4 >i ";
391 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
395 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
396 warn "WARNING sort $sort_by failed";
399 } # finished looping through servers
401 # The big moment: asynchronously retrieve results from all servers
402 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
403 my $ev = $zconns[ $i - 1 ]->last_event();
404 if ( $ev == ZOOM::Event::ZEND ) {
405 next unless $results[ $i - 1 ];
406 my $size = $results[ $i - 1 ]->size();
410 # loop through the results
411 $results_hash->{'hits'} = $size;
413 if ( $offset + $results_per_page <= $size ) {
414 $times = $offset + $results_per_page;
419 for ( my $j = $offset ; $j < $times ; $j++ ) {
424 ## Check if it's an index scan
426 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
428 # here we create a minimal MARC record and hand it off to the
429 # template just like a normal result ... perhaps not ideal, but
431 my $tmprecord = MARC::Record->new();
432 $tmprecord->encoding('UTF-8');
436 # the minimal record in author/title (depending on MARC flavour)
437 if (C4::Context->preference("marcflavour") eq "UNIMARC") {
438 $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
439 $tmprecord->append_fields($tmptitle);
441 $tmptitle = MARC::Field->new('245',' ',' ', a => $term,);
442 $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
443 $tmprecord->append_fields($tmptitle);
444 $tmprecord->append_fields($tmpauthor);
446 $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
451 $record = $results[ $i - 1 ]->record($j)->raw();
453 # warn "RECORD $j:".$record;
454 $results_hash->{'RECORDS'}[$j] = $record;
456 # Fill the facets while we're looping, but only for the biblioserver
457 $facet_record = MARC::Record->new_from_usmarc($record)
458 if $servers[ $i - 1 ] =~ /biblioserver/;
460 #warn $servers[$i-1]."\n".$record; #.$facet_record->title();
462 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
463 ($facets->[$k]) or next;
464 my @fields = map {$facet_record->field($_)} @{$facets->[$k]->{'tags'}} ;
465 for my $field (@fields) {
466 my @subfields = $field->subfields();
467 for my $subfield (@subfields) {
468 my ( $code, $data ) = @$subfield;
469 ($code eq $facets->[$k]->{'subfield'}) or next;
470 $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
473 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} =
474 $facets->[$k]->{'label_value'};
475 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} =
476 $facets->[$k]->{'expanded'};
481 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
484 # warn "connection ", $i-1, ": $size hits";
485 # warn $results[$i-1]->record(0)->render() if $size > 0;
488 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
490 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
491 keys %$facets_counter )
494 my $number_of_facets;
495 my @this_facets_array;
498 $facets_counter->{$link_value}->{$b}
499 <=> $facets_counter->{$link_value}->{$a}
500 } keys %{ $facets_counter->{$link_value} }
504 if ( ( $number_of_facets < 6 )
505 || ( $expanded_facet eq $link_value )
506 || ( $facets_info->{$link_value}->{'expanded'} ) )
509 # Sanitize the link value ), ( will cause errors with CCL,
510 my $facet_link_value = $one_facet;
511 $facet_link_value =~ s/(\(|\))/ /g;
513 # fix the length that will display in the label,
514 my $facet_label_value = $one_facet;
516 substr( $one_facet, 0, 20 ) . "..."
517 unless length($facet_label_value) <= 20;
519 # if it's a branch, label by the name, not the code,
520 if ( $link_value =~ /branch/ ) {
521 if (defined $branches
522 && ref($branches) eq "HASH"
523 && defined $branches->{$one_facet}
524 && ref ($branches->{$one_facet}) eq "HASH")
527 $branches->{$one_facet}->{'branchname'};
530 $facet_label_value = "*";
534 # but we're down with the whole label being in the link's title.
535 push @this_facets_array, {
536 facet_count => $facets_counter->{$link_value}->{$one_facet},
537 facet_label_value => $facet_label_value,
538 facet_title_value => $one_facet,
539 facet_link_value => $facet_link_value,
540 type_link_value => $link_value,
545 # handle expanded option
546 unless ( $facets_info->{$link_value}->{'expanded'} ) {
548 if ( ( $number_of_facets > 6 )
549 && ( $expanded_facet ne $link_value ) );
552 type_link_value => $link_value,
553 type_id => $link_value . "_id",
554 "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
555 facets => \@this_facets_array,
556 expandable => $expandable,
557 expand => $link_value,
558 } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
563 return ( undef, $results_hashref, \@facets_loop );
568 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
569 $results_per_page, $offset, $expanded_facet, $branches,
573 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
575 $paz->search($simple_query);
576 sleep 1; # FIXME: WHY?
579 my $results_hashref = {};
580 my $stats = XMLin($paz->stat);
581 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
583 # for a grouped search result, the number of hits
584 # is the number of groups returned; 'bib_hits' will have
585 # the total number of bibs.
586 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
587 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
589 HIT: foreach my $hit (@{ $results->{'hit'} }) {
590 my $recid = $hit->{recid}->[0];
592 my $work_title = $hit->{'md-work-title'}->[0];
594 if (exists $hit->{'md-work-author'}) {
595 $work_author = $hit->{'md-work-author'}->[0];
597 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
599 my $result_group = {};
600 $result_group->{'group_label'} = $group_label;
601 $result_group->{'group_merge_key'} = $recid;
604 if (exists $hit->{count}) {
605 $count = $hit->{count}->[0];
607 $result_group->{'group_count'} = $count;
609 for (my $i = 0; $i < $count; $i++) {
610 # FIXME -- may need to worry about diacritics here
611 my $rec = $paz->record($recid, $i);
612 push @{ $result_group->{'RECORDS'} }, $rec;
615 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
618 # pass through facets
619 my $termlist_xml = $paz->termlist('author,subject');
620 my $terms = XMLin($termlist_xml, forcearray => 1);
621 my @facets_loop = ();
622 #die Dumper($results);
623 # foreach my $list (sort keys %{ $terms->{'list'} }) {
625 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
627 # facet_label_value => $facet->{'name'}->[0],
630 # push @facets_loop, ( {
631 # type_label => $list,
632 # facets => \@facets,
636 return ( undef, $results_hashref, \@facets_loop );
640 sub _remove_stopwords {
641 my ( $operand, $index ) = @_;
642 my @stopwords_removed;
644 # phrase and exact-qualified indexes shouldn't have stopwords removed
645 if ( $index !~ m/phr|ext/ ) {
647 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
648 # we use IsAlpha unicode definition, to deal correctly with diacritics.
649 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
650 # is a stopword, we'd get "çon" and wouldn't find anything...
652 foreach ( keys %{ C4::Context->stopwords } ) {
653 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
654 $debug && warn "$_ Dump($operand)";
655 if ( my ($matched) = ($operand =~
656 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
658 $operand =~ s/\Q$matched\E/ /gi;
659 push @stopwords_removed, $_;
663 return ( $operand, \@stopwords_removed );
667 sub _detect_truncation {
668 my ( $operand, $index ) = @_;
669 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
672 my @wordlist = split( /\s/, $operand );
673 foreach my $word (@wordlist) {
674 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
675 push @rightlefttruncated, $word;
677 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
678 push @lefttruncated, $word;
680 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
681 push @righttruncated, $word;
683 elsif ( index( $word, "*" ) < 0 ) {
684 push @nontruncated, $word;
687 push @regexpr, $word;
691 \@nontruncated, \@righttruncated, \@lefttruncated,
692 \@rightlefttruncated, \@regexpr
697 sub _build_stemmed_operand {
698 my ($operand,$lang) = @_;
699 require Lingua::Stem::Snowball ;
702 # If operand contains a digit, it is almost certainly an identifier, and should
703 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
704 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
705 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
706 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
707 return $operand if $operand =~ /\d/;
709 # FIXME: the locale should be set based on the user's language and/or search choice
711 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
712 encoding => "UTF-8" );
714 my @words = split( / /, $operand );
715 my @stems = $stemmer->stem(\@words);
716 for my $stem (@stems) {
717 $stemmed_operand .= "$stem";
718 $stemmed_operand .= "?"
719 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
720 $stemmed_operand .= " ";
722 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
723 return $stemmed_operand;
727 sub _build_weighted_query {
729 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
730 # pretty well but could work much better if we had a smarter query parser
731 my ( $operand, $stemmed_operand, $index ) = @_;
732 my $stemming = C4::Context->preference("QueryStemming") || 0;
733 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
734 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
736 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
738 # Keyword, or, no index specified
739 if ( ( $index eq 'kw' ) || ( !$index ) ) {
741 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
742 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
743 $weighted_query .= " or ti,phr,r3=\"$operand\""; # phrase title
744 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
745 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
746 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
747 if $fuzzy_enabled; # add fuzzy, word list
748 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
749 if ( $stemming and $stemmed_operand )
750 ; # add stemming, right truncation
751 $weighted_query .= " or wrdl,r9=\"$operand\"";
753 # embedded sorting: 0 a-z; 1 z-a
754 # $weighted_query .= ") or (sort1,aut=1";
757 # Barcode searches should skip this process
758 elsif ( $index eq 'bc' ) {
759 $weighted_query .= "bc=\"$operand\"";
762 # Authority-number searches should skip this process
763 elsif ( $index eq 'an' ) {
764 $weighted_query .= "an=\"$operand\"";
767 # If the index already has more than one qualifier, wrap the operand
768 # in quotes and pass it back (assumption is that the user knows what they
769 # are doing and won't appreciate us mucking up their query
770 elsif ( $index =~ ',' ) {
771 $weighted_query .= " $index=\"$operand\"";
774 #TODO: build better cases based on specific search indexes
776 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
777 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
778 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
780 " or $index,rt,wrdl,r3=\"$operand\""; # word list index
783 $weighted_query .= "))"; # close rank specification
784 return $weighted_query;
789 Return an array with available indexes.
811 'Author-personal-bibliography',
820 'Chronological-subdivision',
830 'Conference-name-heading',
831 'Conference-name-see',
832 'Conference-name-seealso',
837 'Corporate-name-heading',
838 'Corporate-name-see',
839 'Corporate-name-seealso',
841 'date-entered-on-file',
842 'Date-of-acquisition',
843 'Date-of-publication',
844 'Dewey-classification',
850 'Geographic-subdivision',
853 'Heading-use-main-or-added-entry',
854 'Heading-use-series-added-entry ',
855 'Heading-use-subject-added-entry',
871 'Local-classification',
874 'Match-heading-see-from',
880 'Name-geographic-heading',
881 'Name-geographic-see',
882 'Name-geographic-seealso',
889 'Personal-name-heading',
891 'Personal-name-seealso',
907 'Subject-heading-thesaurus',
908 'Subject-name-personal',
909 'Subject-subdivision',
918 'Term-genre-form-heading',
919 'Term-genre-form-see',
920 'Term-genre-form-seealso',
926 'Title-uniform-heading',
928 'Title-uniform-seealso',
938 'classification-source',
940 'coded-location-qualifier',
951 'Local-classification',
954 'materials-specified',
963 'replacementpricedate',
979 $simple_query, $query_cgi,
981 $limit_cgi, $limit_desc,
982 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
984 Build queries and limits in CCL, CGI, Human,
985 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
987 See verbose embedded documentation.
993 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
995 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
998 my @operators = $operators ? @$operators : ();
999 my @indexes = $indexes ? @$indexes : ();
1000 my @operands = $operands ? @$operands : ();
1001 my @limits = $limits ? @$limits : ();
1002 my @sort_by = $sort_by ? @$sort_by : ();
1004 my $stemming = C4::Context->preference("QueryStemming") || 0;
1005 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1006 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1007 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1008 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1010 # no stemming/weight/fuzzy in NoZebra
1011 if ( C4::Context->preference("NoZebra") ) {
1015 $auto_truncation = 0;
1018 my $query = $operands[0];
1019 my $simple_query = $operands[0];
1021 # initialize the variables we're passing back
1030 my $stopwords_removed; # flag to determine if stopwords have been removed
1033 my $cclindexes = getIndexes();
1034 if( $query !~ /\s*ccl=/ ){
1035 for my $index (@$cclindexes){
1036 if($query =~ /($index)(,?\w)*[:=]/){
1040 $query = "ccl=$query" if($cclq);
1043 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1045 if ( $query =~ /^ccl=/ ) {
1046 return ( undef, $', $', "q=ccl=$'", $', '', '', '', '', 'ccl' );
1048 if ( $query =~ /^cql=/ ) {
1049 return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1051 if ( $query =~ /^pqf=/ ) {
1052 return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1055 # pass nested queries directly
1056 # FIXME: need better handling of some of these variables in this case
1057 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1058 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1059 # if ( $query =~ /(\(|\))/ ) {
1061 # undef, $query, $simple_query, $query_cgi,
1062 # $query, $limit, $limit_cgi, $limit_desc,
1063 # $stopwords_removed, 'ccl'
1067 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1068 # query operands and indexes and add stemming, truncation, field weighting, etc.
1069 # Once we do so, we'll end up with a value in $query, just like if we had an
1070 # incoming $query from the user
1073 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1074 my $previous_operand
1075 ; # a flag used to keep track if there was a previous query
1076 # if there was, we can apply the current operator
1078 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1080 # COMBINE OPERANDS, INDEXES AND OPERATORS
1081 if ( $operands[$i] ) {
1083 # A flag to determine whether or not to add the index to the query
1086 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1087 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1090 $remove_stopwords = 0;
1092 my $operand = $operands[$i];
1093 my $index = $indexes[$i];
1095 # Add index-specific attributes
1096 # Date of Publication
1097 if ( $index eq 'yr' ) {
1098 $index .= ",st-numeric";
1100 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1103 # Date of Acquisition
1104 elsif ( $index eq 'acqdate' ) {
1105 $index .= ",st-date-normalized";
1107 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1109 # ISBN,ISSN,Standard Number, don't need special treatment
1110 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1113 $stemming, $auto_truncation,
1114 $weight_fields, $fuzzy_enabled,
1116 ) = ( 0, 0, 0, 0, 0 );
1124 # Set default structure attribute (word list)
1126 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1127 $struct_attr = ",wrdl";
1130 # Some helpful index variants
1131 my $index_plus = $index . $struct_attr . ":" if $index;
1132 my $index_plus_comma = $index . $struct_attr . "," if $index;
1135 if ($remove_stopwords) {
1136 ( $operand, $stopwords_removed ) =
1137 _remove_stopwords( $operand, $index );
1138 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1139 warn "REMOVED STOPWORDS: @$stopwords_removed"
1140 if ( $stopwords_removed && $DEBUG );
1143 if ($auto_truncation){
1144 unless ( $index =~ /(st-|phr|ext)/ ) {
1145 #FIXME only valid with LTR scripts
1146 $operand=join(" ",map{
1147 (index($_,"*")>0?"$_":"$_*")
1148 }split (/\s+/,$operand));
1149 warn $operand if $DEBUG;
1154 my $truncated_operand;
1155 my( $nontruncated, $righttruncated, $lefttruncated,
1156 $rightlefttruncated, $regexpr
1157 ) = _detect_truncation( $operand, $index );
1159 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1164 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1165 scalar(@$rightlefttruncated) > 0 )
1168 # Don't field weight or add the index to the query, we do it here
1170 undef $weight_fields;
1171 my $previous_truncation_operand;
1172 if (scalar @$nontruncated) {
1173 $truncated_operand .= "$index_plus @$nontruncated ";
1174 $previous_truncation_operand = 1;
1176 if (scalar @$righttruncated) {
1177 $truncated_operand .= "and " if $previous_truncation_operand;
1178 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1179 $previous_truncation_operand = 1;
1181 if (scalar @$lefttruncated) {
1182 $truncated_operand .= "and " if $previous_truncation_operand;
1183 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1184 $previous_truncation_operand = 1;
1186 if (scalar @$rightlefttruncated) {
1187 $truncated_operand .= "and " if $previous_truncation_operand;
1188 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1189 $previous_truncation_operand = 1;
1192 $operand = $truncated_operand if $truncated_operand;
1193 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1196 my $stemmed_operand;
1197 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1200 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1202 # Handle Field Weighting
1203 my $weighted_operand;
1204 if ($weight_fields) {
1205 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1206 $operand = $weighted_operand;
1210 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1212 # If there's a previous operand, we need to add an operator
1213 if ($previous_operand) {
1215 # User-specified operator
1216 if ( $operators[ $i - 1 ] ) {
1217 $query .= " $operators[$i-1] ";
1218 $query .= " $index_plus " unless $indexes_set;
1219 $query .= " $operand";
1220 $query_cgi .= "&op=$operators[$i-1]";
1221 $query_cgi .= "&idx=$index" if $index;
1222 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1224 " $operators[$i-1] $index_plus $operands[$i]";
1227 # Default operator is and
1230 $query .= "$index_plus " unless $indexes_set;
1231 $query .= "$operand";
1232 $query_cgi .= "&op=and&idx=$index" if $index;
1233 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1234 $query_desc .= " and $index_plus $operands[$i]";
1238 # There isn't a pervious operand, don't need an operator
1241 # Field-weighted queries already have indexes set
1242 $query .= " $index_plus " unless $indexes_set;
1244 $query_desc .= " $index_plus $operands[$i]";
1245 $query_cgi .= "&idx=$index" if $index;
1246 $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1247 $previous_operand = 1;
1252 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1255 my $group_OR_limits;
1256 my $availability_limit;
1257 foreach my $this_limit (@limits) {
1258 # if ( $this_limit =~ /available/ ) {
1260 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1262 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1263 # $availability_limit .=
1264 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1265 # $limit_cgi .= "&limit=available";
1266 # $limit_desc .= "";
1269 # group_OR_limits, prefixed by mc-
1270 # OR every member of the group
1271 # elsif ( $this_limit =~ /mc/ ) {
1272 if ( $this_limit =~ /mc/ ) {
1273 $group_OR_limits .= " or " if $group_OR_limits;
1274 $limit_desc .= " or " if $group_OR_limits;
1275 $group_OR_limits .= "$this_limit";
1276 $limit_cgi .= "&limit=$this_limit";
1277 $limit_desc .= " $this_limit";
1280 # Regular old limits
1282 $limit .= " and " if $limit || $query;
1283 $limit .= "$this_limit";
1284 $limit_cgi .= "&limit=$this_limit";
1285 if ($this_limit =~ /^branch:(.+)/) {
1286 my $branchcode = $1;
1287 my $branchname = GetBranchName($branchcode);
1288 if (defined $branchname) {
1289 $limit_desc .= " branch:$branchname";
1291 $limit_desc .= " $this_limit";
1294 $limit_desc .= " $this_limit";
1298 if ($group_OR_limits) {
1299 $limit .= " and " if ( $query || $limit );
1300 $limit .= "($group_OR_limits)";
1302 if ($availability_limit) {
1303 $limit .= " and " if ( $query || $limit );
1304 $limit .= "($availability_limit)";
1307 # Normalize the query and limit strings
1308 # This is flawed , means we can't search anything with : in it
1309 # if user wants to do ccl or cql, start the query with that
1310 # $query =~ s/:/=/g;
1311 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc)):/=/g;
1312 $query =~ s/(?<=rtrn):/=/g;
1314 for ( $query, $query_desc, $limit, $limit_desc ) {
1315 s/ / /g; # remove extra spaces
1316 s/^ //g; # remove any beginning spaces
1317 s/ $//g; # remove any ending spaces
1318 s/==/=/g; # remove double == from query
1320 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1322 for ($query_cgi,$simple_query) {
1325 # append the limit to the query
1326 $query .= " " . $limit;
1330 warn "QUERY:" . $query;
1331 warn "QUERY CGI:" . $query_cgi;
1332 warn "QUERY DESC:" . $query_desc;
1333 warn "LIMIT:" . $limit;
1334 warn "LIMIT CGI:" . $limit_cgi;
1335 warn "LIMIT DESC:" . $limit_desc;
1336 warn "---------\nLeave buildQuery\n---------";
1339 undef, $query, $simple_query, $query_cgi,
1340 $query_desc, $limit, $limit_cgi, $limit_desc,
1341 $stopwords_removed, $query_type
1345 =head2 searchResults
1347 Format results in a form suitable for passing to the template
1351 # IMO this subroutine is pretty messy still -- it's responsible for
1352 # building the HTML output for the template
1354 my ( $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1355 my $dbh = C4::Context->dbh;
1358 #Build branchnames hash
1360 #get branch information.....
1362 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1364 while ( my $bdata = $bsth->fetchrow_hashref ) {
1365 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1367 # FIXME - We build an authorised values hash here, using the default framework
1368 # though it is possible to have different authvals for different fws.
1370 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1372 # get notforloan authorised value list (see $shelflocations FIXME)
1373 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1375 #Build itemtype hash
1376 #find itemtype & itemtype image
1380 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1383 while ( my $bdata = $bsth->fetchrow_hashref ) {
1384 foreach (qw(description imageurl summary notforloan)) {
1385 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1389 #search item field code
1392 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1395 my ($itemtag) = $sth->fetchrow;
1397 ## find column names of items related to MARC
1398 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1400 my %subfieldstosearch;
1401 while ( ( my $column ) = $sth2->fetchrow ) {
1402 my ( $tagfield, $tagsubfield ) =
1403 &GetMarcFromKohaField( "items." . $column, "" );
1404 $subfieldstosearch{$column} = $tagsubfield;
1407 # handle which records to actually retrieve
1409 if ( $hits && $offset + $results_per_page <= $hits ) {
1410 $times = $offset + $results_per_page;
1413 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1416 my $marcflavour = C4::Context->preference("marcflavour");
1417 # We get the biblionumber position in MARC
1418 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1421 # loop through all of the records we've retrieved
1422 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1423 my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1425 $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1427 $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1430 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1431 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1432 $oldbiblio->{result_number} = $i + 1;
1434 # add imageurl to itemtype if there is one
1435 $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1437 $oldbiblio->{'authorised_value_images'} = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1438 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1439 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1440 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1441 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1442 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1444 # edition information, if any
1445 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1446 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1447 # Build summary if there is one (the summary is defined in the itemtypes table)
1448 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1449 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1450 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1451 my @fields = $marcrecord->fields();
1454 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1456 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1457 $tag =~ /(.{3})(.)/;
1458 if($marcrecord->field($1)){
1459 my @abc = $marcrecord->field($1)->subfield($2);
1460 $tags->{$tag} = $#abc + 1 ;
1464 # We catch how many times to repeat this line
1466 foreach my $tag (keys(%$tags)){
1467 $max = $tags->{$tag} if($tags->{$tag} > $max);
1470 # we replace, and repeat each line
1471 for (my $i = 0 ; $i < $max ; $i++){
1472 my $newline = $line;
1474 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1475 $tag =~ /(.{3})(.)/;
1477 if($marcrecord->field($1)){
1478 my @repl = $marcrecord->field($1)->subfield($2);
1479 my $subfieldvalue = $repl[$i];
1481 if (! utf8::is_utf8($subfieldvalue)) {
1482 utf8::decode($subfieldvalue);
1485 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1488 $newsummary .= "$newline\n";
1492 $newsummary =~ s/\[(.*?)]//g;
1493 $newsummary =~ s/\n/<br\/>/g;
1494 $oldbiblio->{summary} = $newsummary;
1497 # Pull out the items fields
1498 my @fields = $marcrecord->field($itemtag);
1500 # Setting item statuses for display
1501 my @available_items_loop;
1502 my @onloan_items_loop;
1503 my @other_items_loop;
1505 my $available_items;
1509 my $ordered_count = 0;
1510 my $available_count = 0;
1511 my $onloan_count = 0;
1512 my $longoverdue_count = 0;
1513 my $other_count = 0;
1514 my $wthdrawn_count = 0;
1515 my $itemlost_count = 0;
1516 my $itembinding_count = 0;
1517 my $itemdamaged_count = 0;
1518 my $item_in_transit_count = 0;
1519 my $can_place_holds = 0;
1520 my $item_onhold_count = 0;
1521 my $items_count = scalar(@fields);
1523 ( C4::Context->preference('maxItemsinSearchResults') )
1524 ? C4::Context->preference('maxItemsinSearchResults') - 1
1527 # loop through every item
1528 foreach my $field (@fields) {
1531 # populate the items hash
1532 foreach my $code ( keys %subfieldstosearch ) {
1533 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1536 my $hbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch' : 'holdingbranch';
1537 my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1538 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1539 if ($item->{$hbranch}) {
1540 $item->{'branchname'} = $branches{$item->{$hbranch}};
1542 elsif ($item->{$otherbranch}) { # Last resort
1543 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1546 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1547 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1548 if ( $item->{onloan} ) {
1550 my $key = $prefix . $item->{onloan} . $item->{barcode};
1551 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1552 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1553 $onloan_items->{$key}->{branchname} = $item->{branchname};
1554 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1555 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1556 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1557 # if something's checked out and lost, mark it as 'long overdue'
1558 if ( $item->{itemlost} ) {
1559 $onloan_items->{$prefix}->{longoverdue}++;
1560 $longoverdue_count++;
1561 } else { # can place holds as long as item isn't lost
1562 $can_place_holds = 1;
1566 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1570 if ( $item->{notforloan} == -1 ) {
1574 # is item in transit?
1575 my $transfertwhen = '';
1576 my ($transfertfrom, $transfertto);
1578 # is item on the reserve shelf?
1579 my $reservestatus = 0;
1582 unless ($item->{wthdrawn}
1583 || $item->{itemlost}
1585 || $item->{notforloan}
1586 || $items_count > 20) {
1588 # A couple heuristics to limit how many times
1589 # we query the database for item transfer information, sacrificing
1590 # accuracy in some cases for speed;
1592 # 1. don't query if item has one of the other statuses
1593 # 2. don't check transit status if the bib has
1594 # more than 20 items
1596 # FIXME: to avoid having the query the database like this, and to make
1597 # the in transit status count as unavailable for search limiting,
1598 # should map transit status to record indexed in Zebra.
1600 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1601 ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1604 # item is withdrawn, lost or damaged
1605 if ( $item->{wthdrawn}
1606 || $item->{itemlost}
1608 || $item->{notforloan}
1609 || $reservestatus eq 'Waiting'
1610 || ($transfertwhen ne ''))
1612 $wthdrawn_count++ if $item->{wthdrawn};
1613 $itemlost_count++ if $item->{itemlost};
1614 $itemdamaged_count++ if $item->{damaged};
1615 $item_in_transit_count++ if $transfertwhen ne '';
1616 $item_onhold_count++ if $reservestatus eq 'Waiting';
1617 $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1620 my $key = $prefix . $item->{status};
1621 foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1622 $other_items->{$key}->{$_} = $item->{$_};
1624 $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1625 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1626 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1627 $other_items->{$key}->{count}++ if $item->{$hbranch};
1628 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1629 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1633 $can_place_holds = 1;
1635 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1636 foreach (qw(branchname itemcallnumber)) {
1637 $available_items->{$prefix}->{$_} = $item->{$_};
1639 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1640 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1643 } # notforloan, item level and biblioitem level
1644 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1646 ( C4::Context->preference('maxItemsinSearchResults') )
1647 ? C4::Context->preference('maxItemsinSearchResults') - 1
1649 for my $key ( sort keys %$onloan_items ) {
1650 (++$onloanitemscount > $maxitems) and last;
1651 push @onloan_items_loop, $onloan_items->{$key};
1653 for my $key ( sort keys %$other_items ) {
1654 (++$otheritemscount > $maxitems) and last;
1655 push @other_items_loop, $other_items->{$key};
1657 for my $key ( sort keys %$available_items ) {
1658 (++$availableitemscount > $maxitems) and last;
1659 push @available_items_loop, $available_items->{$key}
1662 # XSLT processing of some stuff
1664 SetUTF8Flag($marcrecord);
1665 $debug && warn $marcrecord->as_formatted;
1666 if (C4::Context->preference("XSLTResultsDisplay") && !$scan) {
1667 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display(
1668 $oldbiblio->{biblionumber}, $marcrecord, 'Results' );
1671 # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1672 $can_place_holds = 0
1673 if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1674 $oldbiblio->{norequests} = 1 unless $can_place_holds;
1675 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
1676 $oldbiblio->{items_count} = $items_count;
1677 $oldbiblio->{available_items_loop} = \@available_items_loop;
1678 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
1679 $oldbiblio->{other_items_loop} = \@other_items_loop;
1680 $oldbiblio->{availablecount} = $available_count;
1681 $oldbiblio->{availableplural} = 1 if $available_count > 1;
1682 $oldbiblio->{onloancount} = $onloan_count;
1683 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
1684 $oldbiblio->{othercount} = $other_count;
1685 $oldbiblio->{otherplural} = 1 if $other_count > 1;
1686 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
1687 $oldbiblio->{itemlostcount} = $itemlost_count;
1688 $oldbiblio->{damagedcount} = $itemdamaged_count;
1689 $oldbiblio->{intransitcount} = $item_in_transit_count;
1690 $oldbiblio->{onholdcount} = $item_onhold_count;
1691 $oldbiblio->{orderedcount} = $ordered_count;
1692 $oldbiblio->{isbn} =~
1693 s/-//g; # deleting - in isbn to enable amazon content
1694 push( @newresults, $oldbiblio )
1695 if(not $hidelostitems
1696 or (($items_count > $itemlost_count )
1697 && $hidelostitems));
1703 =head2 SearchAcquisitions
1704 Search for acquisitions
1707 sub SearchAcquisitions{
1708 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1710 my $dbh=C4::Context->dbh;
1711 # Variable initialization
1715 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1716 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1717 WHERE dateaccessioned BETWEEN ? AND ?
1720 my (@params,@loopcriteria);
1722 push @params, $datebegin->output("iso");
1723 push @params, $dateend->output("iso");
1725 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1726 if(C4::Context->preference("item-level_itypes")){
1727 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1729 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1731 push @params, @$itemtypes;
1734 if ($criteria =~/itemtype/){
1735 if(C4::Context->preference("item-level_itypes")){
1736 $str .= "AND items.itype=? ";
1738 $str .= "AND biblioitems.itemtype=? ";
1741 if(scalar(@$itemtypes) == 0){
1742 my $itypes = GetItemTypes();
1743 for my $key (keys %$itypes){
1744 push @$itemtypes, $key;
1748 @loopcriteria= @$itemtypes;
1749 }elsif ($criteria=~/itemcallnumber/){
1750 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1751 OR items.itemcallnumber is NULL
1752 OR items.itemcallnumber = '')";
1754 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1756 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1757 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1760 if ($orderby =~ /date_desc/){
1761 $str.=" ORDER BY dateaccessioned DESC";
1763 $str.=" ORDER BY title";
1766 my $qdataacquisitions=$dbh->prepare($str);
1768 my @loopacquisitions;
1769 foreach my $value(@loopcriteria){
1770 push @params,$value;
1772 $cell{"title"}=$value;
1773 $cell{"titlecode"}=$value;
1775 eval{$qdataacquisitions->execute(@params);};
1777 if ($@){ warn "recentacquisitions Error :$@";}
1780 while (my $data=$qdataacquisitions->fetchrow_hashref){
1781 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1783 $cell{"loopdata"}=\@loopdata;
1785 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1788 $qdataacquisitions->finish;
1789 return \@loopacquisitions;
1791 #----------------------------------------------------------------------
1793 # Non-Zebra GetRecords#
1794 #----------------------------------------------------------------------
1798 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1804 $query, $simple_query, $sort_by_ref, $servers_ref,
1805 $results_per_page, $offset, $expanded_facet, $branches,
1808 warn "query =$query" if $DEBUG;
1809 my $result = NZanalyse($query);
1810 warn "results =$result" if $DEBUG;
1812 NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1818 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1819 the list is built from an inverted index in the nozebra SQL table
1820 note that title is here only for convenience : the sorting will be very fast when requested on title
1821 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1826 my ( $string, $server ) = @_;
1827 # warn "---------" if $DEBUG;
1828 warn " NZanalyse" if $DEBUG;
1829 # warn "---------" if $DEBUG;
1831 # $server contains biblioserver or authorities, depending on what we search on.
1832 #warn "querying : $string on $server";
1833 $server = 'biblioserver' unless $server;
1835 # if we have a ", replace the content to discard temporarily any and/or/not inside
1837 if ( $string =~ /"/ ) {
1838 $string =~ s/"(.*?)"/__X__/;
1840 warn "commacontent : $commacontent" if $DEBUG;
1843 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1844 # then, call again NZanalyse with $left and $right
1845 # (recursive until we find a leaf (=> something without and/or/not)
1846 # delete repeated operator... Would then go in infinite loop
1847 while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1850 #process parenthesis before.
1851 if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1854 my $operator = lc($3); # FIXME: and/or/not are operators, not operands
1856 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1858 my $leftresult = NZanalyse( $left, $server );
1860 my $rightresult = NZanalyse( $right, $server );
1862 # OK, we have the results for right and left part of the query
1863 # depending of operand, intersect, union or exclude both lists
1864 # to get a result list
1865 if ( $operator eq ' and ' ) {
1866 return NZoperatorAND($leftresult,$rightresult);
1868 elsif ( $operator eq ' or ' ) {
1870 # just merge the 2 strings
1871 return $leftresult . $rightresult;
1873 elsif ( $operator eq ' not ' ) {
1874 return NZoperatorNOT($leftresult,$rightresult);
1878 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1882 warn "string :" . $string if $DEBUG;
1886 if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1889 $operator = lc($2); # FIXME: and/or/not are operators, not operands
1891 warn "no parenthesis. left : $left operator: $operator right: $right"
1894 # it's not a leaf, we have a and/or/not
1897 # reintroduce comma content if needed
1898 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1899 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1900 warn "node : $left / $operator / $right\n" if $DEBUG;
1901 my $leftresult = NZanalyse( $left, $server );
1902 my $rightresult = NZanalyse( $right, $server );
1903 warn " leftresult : $leftresult" if $DEBUG;
1904 warn " rightresult : $rightresult" if $DEBUG;
1905 # OK, we have the results for right and left part of the query
1906 # depending of operand, intersect, union or exclude both lists
1907 # to get a result list
1908 if ( $operator eq ' and ' ) {
1910 return NZoperatorAND($leftresult,$rightresult);
1912 elsif ( $operator eq ' or ' ) {
1914 # just merge the 2 strings
1915 return $leftresult . $rightresult;
1917 elsif ( $operator eq ' not ' ) {
1918 return NZoperatorNOT($leftresult,$rightresult);
1922 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1923 die "error : operand unknown : $operator for $string";
1926 # it's a leaf, do the real SQL query and return the result
1929 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1930 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1931 #remove trailing blank at the beginning
1933 warn "leaf:$string" if $DEBUG;
1935 # parse the string in in operator/operand/value again
1939 if ($string =~ /(.*)(>=|<=)(.*)/) {
1946 # warn "handling leaf... left:$left operator:$operator right:$right"
1948 unless ($operator) {
1949 if ($string =~ /(.*)(>|<|=)(.*)/) {
1954 "handling unless (operator)... left:$left operator:$operator right:$right"
1962 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1965 # automatic replace for short operators
1966 $left = 'title' if $left =~ '^ti$';
1967 $left = 'author' if $left =~ '^au$';
1968 $left = 'publisher' if $left =~ '^pb$';
1969 $left = 'subject' if $left =~ '^su$';
1970 $left = 'koha-Auth-Number' if $left =~ '^an$';
1971 $left = 'keyword' if $left =~ '^kw$';
1972 $left = 'itemtype' if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1973 warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1974 my $dbh = C4::Context->dbh;
1975 if ( $operator && $left ne 'keyword' ) {
1976 #do a specific search
1977 $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1978 my $sth = $dbh->prepare(
1979 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1981 warn "$left / $operator / $right\n" if $DEBUG;
1983 # split each word, query the DB and build the biblionumbers result
1984 #sanitizing leftpart
1985 $left =~ s/^\s+|\s+$//;
1986 foreach ( split / /, $right ) {
1988 $_ =~ s/^\s+|\s+$//;
1990 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1991 $sth->execute( $server, $left, $_ )
1992 or warn "execute failed: $!";
1993 while ( my ( $line, $value ) = $sth->fetchrow ) {
1995 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
1996 # otherwise, fill the result
1997 $biblionumbers .= $line
1998 unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
1999 warn "result : $value "
2000 . ( $right =~ /\d/ ) . "=="
2001 . ( $value =~ /\D/?$line:"" ) if $DEBUG; #= $line";
2004 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2006 warn "NZAND" if $DEBUG;
2007 $results = NZoperatorAND($biblionumbers,$results);
2009 $results = $biblionumbers;
2014 #do a complete search (all indexes), if index='kw' do complete search too.
2015 my $sth = $dbh->prepare(
2016 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2019 # split each word, query the DB and build the biblionumbers result
2020 foreach ( split / /, $string ) {
2021 next if C4::Context->stopwords->{ uc($_) }; # skip if stopword
2022 warn "search on all indexes on $_" if $DEBUG;
2025 $sth->execute( $server, $_ );
2026 while ( my $line = $sth->fetchrow ) {
2027 $biblionumbers .= $line;
2030 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2032 $results = NZoperatorAND($biblionumbers,$results);
2035 warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2036 $results = $biblionumbers;
2040 warn "return : $results for LEAF : $string" if $DEBUG;
2043 warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2047 my ($rightresult, $leftresult)=@_;
2049 my @leftresult = split /;/, $leftresult;
2050 warn " @leftresult / $rightresult \n" if $DEBUG;
2052 # my @rightresult = split /;/,$leftresult;
2055 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2056 # the result is stored twice, to have the same weight for AND than OR.
2057 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2058 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2059 foreach (@leftresult) {
2062 ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2063 if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2064 $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2066 "$value-$countvalue;$value-$countvalue;";
2069 warn "NZAND DONE : $finalresult \n" if $DEBUG;
2070 return $finalresult;
2074 my ($rightresult, $leftresult)=@_;
2075 return $rightresult.$leftresult;
2079 my ($leftresult, $rightresult)=@_;
2081 my @leftresult = split /;/, $leftresult;
2083 # my @rightresult = split /;/,$leftresult;
2085 foreach (@leftresult) {
2087 $value=$1 if $value=~m/(.*)-\d+$/;
2088 unless ($rightresult =~ "$value-") {
2089 $finalresult .= "$_;";
2092 return $finalresult;
2097 $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2104 my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2105 warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2107 # order title asc by default
2108 # $ordering = '1=36 <i' unless $ordering;
2109 $results_per_page = 20 unless $results_per_page;
2110 $offset = 0 unless $offset;
2111 my $dbh = C4::Context->dbh;
2114 # order by POPULARITY
2116 if ( $ordering =~ /popularity/ ) {
2120 # popularity is not in MARC record, it's builded from a specific query
2122 $dbh->prepare("select sum(issues) from items where biblionumber=?");
2123 foreach ( split /;/, $biblionumbers ) {
2124 my ( $biblionumber, $title ) = split /,/, $_;
2125 $result{$biblionumber} = GetMarcBiblio($biblionumber);
2126 $sth->execute($biblionumber);
2127 my $popularity = $sth->fetchrow || 0;
2129 # hint : the key is popularity.title because we can have
2130 # many results with the same popularity. In this case, sub-ordering is done by title
2131 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2132 # (un-frequent, I agree, but we won't forget anything that way ;-)
2133 $popularity{ sprintf( "%10d", $popularity ) . $title
2134 . $biblionumber } = $biblionumber;
2137 # sort the hash and return the same structure as GetRecords (Zebra querying)
2140 if ( $ordering eq 'popularity_dsc' ) { # sort popularity DESC
2141 foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2142 $result_hash->{'RECORDS'}[ $numbers++ ] =
2143 $result{ $popularity{$key} }->as_usmarc();
2146 else { # sort popularity ASC
2147 foreach my $key ( sort ( keys %popularity ) ) {
2148 $result_hash->{'RECORDS'}[ $numbers++ ] =
2149 $result{ $popularity{$key} }->as_usmarc();
2152 my $finalresult = ();
2153 $result_hash->{'hits'} = $numbers;
2154 $finalresult->{'biblioserver'} = $result_hash;
2155 return $finalresult;
2161 elsif ( $ordering =~ /author/ ) {
2163 foreach ( split /;/, $biblionumbers ) {
2164 my ( $biblionumber, $title ) = split /,/, $_;
2165 my $record = GetMarcBiblio($biblionumber);
2167 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2168 $author = $record->subfield( '200', 'f' );
2169 $author = $record->subfield( '700', 'a' ) unless $author;
2172 $author = $record->subfield( '100', 'a' );
2175 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2176 # and we don't want to get only 1 result for each of them !!!
2177 $result{ $author . $biblionumber } = $record;
2180 # sort the hash and return the same structure as GetRecords (Zebra querying)
2183 if ( $ordering eq 'author_za' ) { # sort by author desc
2184 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2185 $result_hash->{'RECORDS'}[ $numbers++ ] =
2186 $result{$key}->as_usmarc();
2189 else { # sort by author ASC
2190 foreach my $key ( sort ( keys %result ) ) {
2191 $result_hash->{'RECORDS'}[ $numbers++ ] =
2192 $result{$key}->as_usmarc();
2195 my $finalresult = ();
2196 $result_hash->{'hits'} = $numbers;
2197 $finalresult->{'biblioserver'} = $result_hash;
2198 return $finalresult;
2201 # ORDER BY callnumber
2204 elsif ( $ordering =~ /callnumber/ ) {
2206 foreach ( split /;/, $biblionumbers ) {
2207 my ( $biblionumber, $title ) = split /,/, $_;
2208 my $record = GetMarcBiblio($biblionumber);
2210 my $frameworkcode = GetFrameworkCode($biblionumber);
2211 my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField( 'items.itemcallnumber', $frameworkcode);
2212 ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2213 unless $callnumber_tag;
2214 if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2215 $callnumber = $record->subfield( '200', 'f' );
2217 $callnumber = $record->subfield( '100', 'a' );
2220 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2221 # and we don't want to get only 1 result for each of them !!!
2222 $result{ $callnumber . $biblionumber } = $record;
2225 # sort the hash and return the same structure as GetRecords (Zebra querying)
2228 if ( $ordering eq 'call_number_dsc' ) { # sort by title desc
2229 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2230 $result_hash->{'RECORDS'}[ $numbers++ ] =
2231 $result{$key}->as_usmarc();
2234 else { # sort by title ASC
2235 foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2236 $result_hash->{'RECORDS'}[ $numbers++ ] =
2237 $result{$key}->as_usmarc();
2240 my $finalresult = ();
2241 $result_hash->{'hits'} = $numbers;
2242 $finalresult->{'biblioserver'} = $result_hash;
2243 return $finalresult;
2245 elsif ( $ordering =~ /pubdate/ ) { #pub year
2247 foreach ( split /;/, $biblionumbers ) {
2248 my ( $biblionumber, $title ) = split /,/, $_;
2249 my $record = GetMarcBiblio($biblionumber);
2250 my ( $publicationyear_tag, $publicationyear_subfield ) =
2251 GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2252 my $publicationyear =
2253 $record->subfield( $publicationyear_tag,
2254 $publicationyear_subfield );
2256 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2257 # and we don't want to get only 1 result for each of them !!!
2258 $result{ $publicationyear . $biblionumber } = $record;
2261 # sort the hash and return the same structure as GetRecords (Zebra querying)
2264 if ( $ordering eq 'pubdate_dsc' ) { # sort by pubyear desc
2265 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2266 $result_hash->{'RECORDS'}[ $numbers++ ] =
2267 $result{$key}->as_usmarc();
2270 else { # sort by pub year ASC
2271 foreach my $key ( sort ( keys %result ) ) {
2272 $result_hash->{'RECORDS'}[ $numbers++ ] =
2273 $result{$key}->as_usmarc();
2276 my $finalresult = ();
2277 $result_hash->{'hits'} = $numbers;
2278 $finalresult->{'biblioserver'} = $result_hash;
2279 return $finalresult;
2285 elsif ( $ordering =~ /title/ ) {
2287 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2289 foreach ( split /;/, $biblionumbers ) {
2290 my ( $biblionumber, $title ) = split /,/, $_;
2292 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2293 # and we don't want to get only 1 result for each of them !!!
2294 # hint & speed improvement : we can order without reading the record
2295 # so order, and read records only for the requested page !
2296 $result{ $title . $biblionumber } = $biblionumber;
2299 # sort the hash and return the same structure as GetRecords (Zebra querying)
2302 if ( $ordering eq 'title_az' ) { # sort by title desc
2303 foreach my $key ( sort ( keys %result ) ) {
2304 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2307 else { # sort by title ASC
2308 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2309 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2313 # limit the $results_per_page to result size if it's more
2314 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2316 # for the requested page, replace biblionumber by the complete record
2317 # speed improvement : avoid reading too much things
2319 my $counter = $offset ;
2320 $counter <= $offset + $results_per_page ;
2324 $result_hash->{'RECORDS'}[$counter] =
2325 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2327 my $finalresult = ();
2328 $result_hash->{'hits'} = $numbers;
2329 $finalresult->{'biblioserver'} = $result_hash;
2330 return $finalresult;
2337 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2340 foreach ( split /;/, $biblionumbers ) {
2341 my ( $biblionumber, $title ) = split /,/, $_;
2342 $title =~ /(.*)-(\d)/;
2347 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2348 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2349 # biblio N has ranking = 6
2350 $count_ranking{$biblionumber} += $ranking;
2353 # build the result by "inverting" the count_ranking hash
2354 # 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
2356 foreach ( keys %count_ranking ) {
2357 $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2360 # sort the hash and return the same structure as GetRecords (Zebra querying)
2363 foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2364 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2367 # limit the $results_per_page to result size if it's more
2368 $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2370 # for the requested page, replace biblionumber by the complete record
2371 # speed improvement : avoid reading too much things
2373 my $counter = $offset ;
2374 $counter <= $offset + $results_per_page ;
2378 $result_hash->{'RECORDS'}[$counter] =
2379 GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2380 if $result_hash->{'RECORDS'}[$counter];
2382 my $finalresult = ();
2383 $result_hash->{'hits'} = $numbers;
2384 $finalresult->{'biblioserver'} = $result_hash;
2385 return $finalresult;
2389 =head2 enabled_staff_search_views
2391 %hash = enabled_staff_search_views()
2393 This function returns a hash that contains three flags obtained from the system
2394 preferences, used to determine whether a particular staff search results view
2399 =item C<Output arg:>
2401 * $hash{can_view_MARC} is true only if the MARC view is enabled
2402 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2403 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2405 =item C<usage in the script:>
2409 $template->param ( C4::Search::enabled_staff_search_views );
2413 sub enabled_staff_search_views
2416 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2417 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2418 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2422 sub AddSearchHistory{
2423 my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2424 my $dbh = C4::Context->dbh;
2426 # Add the request the user just made
2427 my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2428 my $sth = $dbh->prepare($sql);
2429 $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2430 return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2433 sub GetSearchHistory{
2434 my ($borrowernumber,$session)=@_;
2435 my $dbh = C4::Context->dbh;
2437 # Add the request the user just made
2438 my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2439 my $sth = $dbh->prepare($query);
2440 $sth->execute($borrowernumber, $session);
2441 return $sth->fetchall_hashref({});
2444 =head2 z3950_search_args
2446 $arrayref = z3950_search_args($matchpoints)
2448 This function returns an array reference that contains the search parameters to be
2449 passed to the Z39.50 search script (z3950_search.pl). The array elements
2450 are hash refs whose keys are name, value and encvalue, and whose values are the
2451 name of a search parameter, the value of that search parameter and the URL encoded
2452 value of that parameter.
2454 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2456 The search parameter values are obtained from the bibliographic record whose
2457 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2459 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2460 a general purpose search argument. In this case, the returned array contains only
2461 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2463 If a search parameter value is undefined or empty, it is not included in the returned
2466 The returned array reference may be passed directly to the template parameters.
2470 =item C<Output arg:>
2472 * $array containing hash refs as described above
2474 =item C<usage in the script:>
2478 $data = Biblio::GetBiblioData($bibno);
2479 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2483 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2487 sub z3950_search_args {
2489 $bibrec = { title => $bibrec } if !ref $bibrec;
2491 for my $field (qw/ lccn isbn issn title author dewey subject /)
2493 my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2494 push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2499 =head2 BiblioAddAuthorities
2501 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2503 this function finds the authorities linked to the biblio
2504 * search in the authority DB for the same authid (in $9 of the biblio)
2505 * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2506 * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2507 OR adds a new authority record
2513 * $record is the MARC record in question (marc blob)
2514 * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2516 =item C<Output arg:>
2518 * $countlinked is the number of authorities records that are linked to this authority
2522 * I had to add this to Search.pm (instead of the logical Biblio.pm) because of a circular dependency (this sub uses SimpleSearch, and Search.pm uses Biblio.pm)
2529 sub BiblioAddAuthorities{
2530 my ( $record, $frameworkcode ) = @_;
2531 my $dbh=C4::Context->dbh;
2532 my $query=$dbh->prepare(qq|
2533 SELECT authtypecode,tagfield
2534 FROM marc_subfield_structure
2535 WHERE frameworkcode=?
2536 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2537 # SELECT authtypecode,tagfield
2538 # FROM marc_subfield_structure
2539 # WHERE frameworkcode=?
2540 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2541 $query->execute($frameworkcode);
2542 my ($countcreated,$countlinked);
2543 while (my $data=$query->fetchrow_hashref){
2544 foreach my $field ($record->field($data->{tagfield})){
2545 next if ($field->subfield('3')||$field->subfield('9'));
2546 # No authorities id in the tag.
2547 # Search if there is any authorities to link to.
2548 my $query='at='.$data->{authtypecode}.' ';
2549 map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)} $field->subfields();
2550 my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2551 # there is only 1 result
2553 warn "BIBLIOADDSAUTHORITIES: $error";
2556 if ($results && scalar(@$results)==1) {
2557 my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2558 $field->add_subfields('9'=>$marcrecord->field('001')->data);
2560 } elsif (scalar(@$results)>1) {
2561 #More than One result
2562 #This can comes out of a lack of a subfield.
2563 # my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2564 # $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2567 #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2568 ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2569 ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2570 my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2571 next unless $authtypedata;
2572 my $marcrecordauth=MARC::Record->new();
2573 my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2574 map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )} $field->subfields();
2575 $marcrecordauth->insert_fields_ordered($authfield);
2577 # bug 2317: ensure new authority knows it's using UTF-8; currently
2578 # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2579 # automatically for UNIMARC (by not transcoding)
2580 # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2581 # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2582 # of change to a core API just before the 3.0 release.
2583 if (C4::Context->preference('marcflavour') eq 'MARC21') {
2584 SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2587 # warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2589 my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2591 $field->add_subfields('9'=>$authid);
2595 return ($countlinked,$countcreated);
2598 =head2 GetDistinctValues($field);
2600 C<$field> is a reference to the fields array
2604 sub GetDistinctValues {
2605 my ($fieldname,$string)=@_;
2606 # returns a reference to a hash of references to branches...
2607 if ($fieldname=~/\./){
2608 my ($table,$column)=split /\./, $fieldname;
2609 my $dbh = C4::Context->dbh;
2610 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column ";
2611 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 ");
2613 my $elements=$sth->fetchall_arrayref({});
2618 my @servers=qw<biblioserver authorityserver>;
2619 my (@zconns,@results);
2620 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2621 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2624 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2627 # The big moment: asynchronously retrieve results from all servers
2629 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2630 my $ev = $zconns[ $i - 1 ]->last_event();
2631 if ( $ev == ZOOM::Event::ZEND ) {
2632 next unless $results[ $i - 1 ];
2633 my $size = $results[ $i - 1 ]->size();
2635 for (my $j=0;$j<$size;$j++){
2637 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2638 push @elements, \%hashscan;
2648 END { } # module clean-up code here (global destructor)
2655 Koha Developement team <info@koha.org>