3 # This file is part of Koha.
5 # Koha is free software; you can redistribute it and/or modify it under the
6 # terms of the GNU General Public License as published by the Free Software
7 # Foundation; either version 2 of the License, or (at your option) any later
10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
14 # You should have received a copy of the GNU General Public License along with
15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
16 # Suite 330, Boston, MA 02111-1307 USA
21 use C4::Biblio; # GetMarcFromKohaField
22 use C4::Koha; # getFacets
25 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);
27 # set the version for version checking
28 $VERSION = do { my @v = '$Revision$' =~ /\d+/g;
29 shift(@v) . "." . join( "_", map { sprintf "%03d", $_ } @v );
34 C4::Search - Functions for searching the Koha catalog.
38 see opac/opac-search.pl or catalogue/search.pl for example of usage
42 This module provides the searching facilities for the Koha into a zebra catalog.
60 # make all your functions, whether exported or not;
62 =head2 findseealso($dbh,$fields);
64 C<$dbh> is a link to the DB handler.
67 my $dbh =C4::Context->dbh;
69 C<$fields> is a reference to the fields array
71 This function modify the @$fields array and add related fields to search on.
76 my ( $dbh, $fields ) = @_;
77 my $tagslib = GetMarcStructure( 1 );
78 for ( my $i = 0 ; $i <= $#{$fields} ; $i++ ) {
79 my ($tag) = substr( @$fields[$i], 1, 3 );
80 my ($subfield) = substr( @$fields[$i], 4, 1 );
81 @$fields[$i] .= ',' . $tagslib->{$tag}->{$subfield}->{seealso}
82 if ( $tagslib->{$tag}->{$subfield}->{seealso} );
88 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
95 my $dbh = C4::Context->dbh;
96 my $result = TransformMarcToKoha( $dbh, $record, '' );
101 my ( $biblionumber, $title );
103 # search duplicate on ISBN, easy and fast..
104 #$search->{'avoidquerylog'}=1;
105 if ( $result->{isbn} ) {
106 $query = "isbn=$result->{isbn}";
109 $result->{title} =~ s /\\//g;
110 $result->{title} =~ s /\"//g;
111 $result->{title} =~ s /\(//g;
112 $result->{title} =~ s /\)//g;
113 $query = "ti,ext=$result->{title}";
115 my ($possible_duplicate_record) =
116 C4::Biblio::getRecord( "biblioserver", $query, "usmarc" ); # FIXME :: hardcoded !
117 if ($possible_duplicate_record) {
119 MARC::Record->new_from_usmarc($possible_duplicate_record);
120 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
122 # FIXME :: why 2 $biblionumber ?
123 return $result->{'biblionumber'}, $result->{'biblionumber'},
131 ($error,$results) = SimpleSearch($query,@servers);
133 this function performs a simple search on the catalog using zoom.
139 * $query could be a simple keyword or a complete CCL query wich is depending on your ccl file.
140 * @servers is optionnal. default one is read on koha.xml
143 * $error is a string which containt the description error if there is one. Else it's empty.
144 * \@results is an array of marc record.
146 =item C<usage in the script:>
150 my ($error, $marcresults) = SimpleSearch($query);
152 if (defined $error) {
153 $template->param(query_error => $error);
154 warn "error: ".$error;
155 output_html_with_http_headers $input, $cookie, $template->output;
159 my $hits = scalar @$marcresults;
162 for(my $i=0;$i<$hits;$i++) {
164 my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
165 my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
167 #build the hash for the template.
168 $resultsloop{highlight} = ($i % 2)?(1):(0);
169 $resultsloop{title} = $biblio->{'title'};
170 $resultsloop{subtitle} = $biblio->{'subtitle'};
171 $resultsloop{biblionumber} = $biblio->{'biblionumber'};
172 $resultsloop{author} = $biblio->{'author'};
173 $resultsloop{publishercode} = $biblio->{'publishercode'};
174 $resultsloop{publicationyear} = $biblio->{'publicationyear'};
176 push @results, \%resultsloop;
178 $template->param(result=>\@results);
184 if (C4::Context->preference('NoZebra')) {
185 my $result = NZorder(NZanalyse($query))->{'biblioserver'}->{'RECORDS'};
186 return (undef,$result);
192 return ( "No query entered", undef ) unless $query;
194 #@servers = (C4::Context->config("biblioserver")) unless @servers;
196 ("biblioserver") unless @servers
197 ; # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
200 for ( my $i = 0 ; $i < @servers ; $i++ ) {
201 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
204 ->search( new ZOOM::Query::CCL2RPN( $query, $zconns[$i] ) );
206 # getting error message if one occured.
208 $zconns[$i]->errmsg() . " ("
209 . $zconns[$i]->errcode() . ") "
210 . $zconns[$i]->addinfo() . " "
211 . $zconns[$i]->diagset();
213 return ( $error, undef ) if $zconns[$i]->errcode();
217 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
218 $ev = $zconns[ $i - 1 ]->last_event();
219 if ( $ev == ZOOM::Event::ZEND ) {
220 $hits = $tmpresults[ $i - 1 ]->size();
223 for ( my $j = 0 ; $j < $hits ; $j++ ) {
224 my $record = $tmpresults[ $i - 1 ]->record($j)->raw();
225 push @results, $record;
229 return ( undef, \@results );
233 # performs the search
236 $koha_query, $federated_query, $sort_by_ref,
237 $servers_ref, $results_per_page, $offset,
238 $expanded_facet, $branches, $query_type,
242 my @servers = @$servers_ref;
243 my @sort_by = @$sort_by_ref;
245 # create the zoom connection and query object
249 my $results_hashref = ();
252 my $facets_counter = ();
253 my $facets_info = ();
254 my $facets = getFacets();
256 #### INITIALIZE SOME VARS USED CREATE THE FACETED RESULTS
257 my @facets_loop; # stores the ref to array of hashes for template
258 for ( my $i = 0 ; $i < @servers ; $i++ ) {
259 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
261 # perform the search, create the results objects
262 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
264 if ( $servers[$i] =~ /biblioserver/ ) {
265 $query_to_use = $koha_query;
268 $query_to_use = $federated_query;
271 # check if we've got a query_type defined
275 if ( $query_type =~ /^ccl/ ) {
277 s/\:/\=/g; # change : to = last minute (FIXME)
279 # warn "CCL : $query_to_use";
282 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
285 elsif ( $query_type =~ /^cql/ ) {
287 # warn "CQL : $query_to_use";
290 new ZOOM::Query::CQL( $query_to_use, $zconns[$i] ) );
292 elsif ( $query_type =~ /^pqf/ ) {
294 # warn "PQF : $query_to_use";
297 new ZOOM::Query::PQF( $query_to_use, $zconns[$i] ) );
303 # warn "preparing to scan";
306 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
311 # warn "LAST : $query_to_use";
314 new ZOOM::Query::CCL2RPN( $query_to_use, $zconns[$i] )
320 warn "prob with query toto $query_to_use " . $@;
323 # concatenate the sort_by limits and pass them to the results object
325 foreach my $sort (@sort_by) {
326 $sort_by .= $sort . " "; # used to be $sort,
328 $results[$i]->sort( "yaz", $sort_by ) if $sort_by;
330 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
331 my $ev = $zconns[ $i - 1 ]->last_event();
332 if ( $ev == ZOOM::Event::ZEND ) {
333 my $size = $results[ $i - 1 ]->size();
336 #$results_hash->{'server'} = $servers[$i-1];
337 # loop through the results
338 $results_hash->{'hits'} = $size;
340 if ( $offset + $results_per_page <= $size ) {
341 $times = $offset + $results_per_page;
346 for ( my $j = $offset ; $j < $times ; $j++ )
347 { #(($offset+$count<=$size) ? ($offset+$count):$size) ; $j++){
351 ## This is just an index scan
353 my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
355 # here we create a minimal MARC record and hand it off to the
356 # template just like a normal result ... perhaps not ideal, but
358 my $tmprecord = MARC::Record->new();
359 $tmprecord->encoding('UTF-8');
362 # srote the minimal record in author/title (depending on MARC flavour)
363 if ( C4::Context->preference("marcflavour") eq
366 $tmptitle = MARC::Field->new(
373 $tmptitle = MARC::Field->new(
379 $tmprecord->append_fields($tmptitle);
380 $results_hash->{'RECORDS'}[$j] =
381 $tmprecord->as_usmarc();
384 $record = $results[ $i - 1 ]->record($j)->raw();
386 #warn "RECORD $j:".$record;
387 $results_hash->{'RECORDS'}[$j] =
388 $record; # making a reference to a hash
389 # Fill the facets while we're looping
390 $facet_record = MARC::Record->new_from_usmarc($record);
392 #warn $servers[$i-1].$facet_record->title();
393 for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
394 if ( $facets->[$k] ) {
396 for my $tag ( @{ $facets->[$k]->{'tags'} } ) {
397 push @fields, $facet_record->field($tag);
399 for my $field (@fields) {
400 my @subfields = $field->subfields();
401 for my $subfield (@subfields) {
402 my ( $code, $data ) = @$subfield;
404 $facets->[$k]->{'subfield'} )
406 $facets_counter->{ $facets->[$k]
407 ->{'link_value'} }->{$data}++;
411 $facets_info->{ $facets->[$k]->{'link_value'} }
413 $facets->[$k]->{'label_value'};
414 $facets_info->{ $facets->[$k]->{'link_value'} }
415 ->{'expanded'} = $facets->[$k]->{'expanded'};
420 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
423 #print "connection ", $i-1, ": $size hits";
424 #print $results[$i-1]->record(0)->render() if $size > 0;
427 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
428 keys %$facets_counter
432 my $number_of_facets;
433 my @this_facets_array;
436 $facets_counter->{$link_value}
437 ->{$b} <=> $facets_counter->{$link_value}->{$a}
438 } keys %{ $facets_counter->{$link_value} }
442 if ( ( $number_of_facets < 6 )
443 || ( $expanded_facet eq $link_value )
444 || ( $facets_info->{$link_value}->{'expanded'} ) )
447 # sanitize the link value ), ( will cause errors with CCL
448 my $facet_link_value = $one_facet;
449 $facet_link_value =~ s/(\(|\))/ /g;
451 # fix the length that will display in the label
452 my $facet_label_value = $one_facet;
453 $facet_label_value = substr( $one_facet, 0, 20 ) . "..."
454 unless length($facet_label_value) <= 20;
456 # well, if it's a branch, label by the name, not the code
457 if ( $link_value =~ /branch/ ) {
459 $branches->{$one_facet}->{'branchname'};
462 # but we're down with the whole label being in the link's title
463 my $facet_title_value = $one_facet;
465 push @this_facets_array,
469 $facets_counter->{$link_value}->{$one_facet},
470 facet_label_value => $facet_label_value,
471 facet_title_value => $facet_title_value,
472 facet_link_value => $facet_link_value,
473 type_link_value => $link_value,
478 unless ( $facets_info->{$link_value}->{'expanded'} ) {
480 if ( ( $number_of_facets > 6 )
481 && ( $expanded_facet ne $link_value ) );
486 type_link_value => $link_value,
487 type_id => $link_value . "_id",
489 $facets_info->{$link_value}->{'label_value'},
490 facets => \@this_facets_array,
491 expandable => $expandable,
492 expand => $link_value,
498 return ( undef, $results_hashref, \@facets_loop );
501 # build the query itself
503 my ( $query, $operators, $operands, $indexes, $limits, $sort_by ) = @_;
505 my @operators = @$operators if $operators;
506 my @indexes = @$indexes if $indexes;
507 my @operands = @$operands if $operands;
508 my @limits = @$limits if $limits;
509 my @sort_by = @$sort_by if $sort_by;
511 my $human_search_desc; # a human-readable query
512 my $machine_search_desc; #a machine-readable query
513 # FIXME: the locale should be set based on the syspref
514 my $stemmer = Lingua::Stem->new( -locale => 'EN-US' );
516 # FIXME: these should be stored in the db so the librarian can modify the behavior
517 $stemmer->add_exceptions(
525 # STEP I: determine if this is a form-based / simple query or if it's complex (if complex,
526 # we can't handle field weighting, stemming until a formal query parser is written
527 # I'll work on this soon -- JF
528 #if (!$query) { # form-based
529 # check if this is a known query language query, if it is, return immediately:
530 if ( $query =~ /^ccl=/ ) {
531 return ( undef, $', $', $', 'ccl' );
533 if ( $query =~ /^cql=/ ) {
534 return ( undef, $', $', $', 'cql' );
536 if ( $query =~ /^pqf=/ ) {
537 return ( undef, $', $', $', 'pqf' );
539 if ( $query =~ /(\(|\))/ ) { # sorry, too complex
540 return ( undef, $query, $query, $query, 'ccl' );
543 # form-based queries are limited to non-nested a specific depth, so we can easily
544 # modify the incoming query operands and indexes to do stemming and field weighting
545 # Once we do so, we'll end up with a value in $query, just like if we had an
546 # incoming $query from the user
549 ; # clear it out so we can populate properly with field-weighted stemmed query
551 ; # a flag used to keep track if there was a previous query
552 # if there was, we can apply the current operator
553 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
554 my $operand = $operands[$i];
555 my $index = $indexes[$i];
557 my $stemming = C4::Context->parameters("Stemming") || 0;
558 my $weight_fields = C4::Context->parameters("WeightFields") || 0;
560 if ( $operands[$i] ) {
561 $operand =~ s/^(and |or |not )//i;
563 # STEMMING FIXME: need to refine the field weighting so stemmed operands don't disrupt the query ranking
565 my @words = split( / /, $operands[$i] );
566 my $stems = $stemmer->stem(@words);
567 foreach my $stem (@$stems) {
568 $stemmed_operand .= "$stem";
569 $stemmed_operand .= "?"
570 unless ( $stem =~ /(and$|or$|not$)/ )
571 || ( length($stem) < 3 );
572 $stemmed_operand .= " ";
574 #warn "STEM: $stemmed_operand";
577 #$operand = $stemmed_operand;
580 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
581 # pretty well but will work much better when we have an actual query parser
583 if ($weight_fields) {
585 " rk=("; # Specifies that we're applying rank
586 # keyword has different weight properties
587 if ( ( $index =~ /kw/ ) || ( !$index ) )
588 { # FIXME: do I need to add right-truncation in the case of stemming?
589 # a simple way to find out if this query uses an index
590 if ( $operand =~ /(\=|\:)/ ) {
591 $weighted_query .= " $operand";
595 " Title-cover,ext,r1=\"$operand\""
596 ; # index label as exact
598 " or ti,ext,r2=$operand"; # index as exact
599 #$weighted_query .= " or ti,phr,r3=$operand"; # index as phrase
600 #$weighted_query .= " or any,ext,r4=$operand"; # index as exact
602 " or kw,wrdl,r5=$operand"; # index as exact
603 $weighted_query .= " or wrd,fuzzy,r9=$operand";
604 $weighted_query .= " or wrd=$stemmed_operand"
608 elsif ( $index =~ /au/ ) {
610 " $index,ext,r1=$operand"; # index label as exact
611 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
613 " or $index,phr,r3=$operand"; # index as phrase
614 $weighted_query .= " or $index,rt,wrd,r3=$operand";
616 elsif ( $index =~ /ti/ ) {
618 " Title-cover,ext,r1=$operand"; # index label as exact
619 $weighted_query .= " or Title-series,ext,r2=$operand";
621 #$weighted_query .= " or ti,ext,r2=$operand";
622 #$weighted_query .= " or ti,phr,r3=$operand";
623 #$weighted_query .= " or ti,wrd,r3=$operand";
625 " or (title-sort-az=0 or Title-cover,startswithnt,st-word,r3=$operand #)";
627 " or (title-sort-az=0 or Title-cover,phr,r6=$operand)";
629 #$weighted_query .= " or Title-cover,wrd,r5=$operand";
630 #$weighted_query .= " or ti,ext,r6=$operand";
631 #$weighted_query .= " or ti,startswith,phr,r7=$operand";
632 #$weighted_query .= " or ti,phr,r8=$operand";
633 #$weighted_query .= " or ti,wrd,r9=$operand";
635 #$weighted_query .= " or ti,ext,r2=$operand"; # index as exact
636 #$weighted_query .= " or ti,phr,r3=$operand"; # index as phrase
637 #$weighted_query .= " or any,ext,r4=$operand"; # index as exact
638 #$weighted_query .= " or kw,wrd,r5=$operand"; # index as exact
642 " $index,ext,r1=$operand"; # index label as exact
643 #$weighted_query .= " or $index,ext,r2=$operand"; # index as exact
645 " or $index,phr,r3=$operand"; # index as phrase
646 $weighted_query .= " or $index,rt,wrd,r3=$operand";
648 " or $index,wrd,r5=$operand"
649 ; # index as word right-truncated
650 $weighted_query .= " or $index,wrd,fuzzy,r8=$operand";
652 $weighted_query .= ")"; # close rank specification
653 $operand = $weighted_query;
656 # only add an operator if there is a previous operand
657 if ($previous_operand) {
658 if ( $operators[ $i - 1 ] ) {
659 $query .= " $operators[$i-1] $index: $operand";
661 $human_search_desc .=
662 " $operators[$i-1] $operands[$i]";
665 $human_search_desc .=
666 " $operators[$i-1] $index: $operands[$i]";
670 # the default operator is and
672 $query .= " and $index: $operand";
673 $human_search_desc .= " and $index: $operands[$i]";
678 $query .= " $operand";
679 $human_search_desc .= " $operands[$i]";
682 $query .= " $index: $operand";
683 $human_search_desc .= " $index: $operands[$i]";
685 $previous_operand = 1;
693 my $limit_search_desc;
694 foreach my $limit (@limits) {
696 # FIXME: not quite right yet ... will work on this soon -- JF
697 my $type = $1 if $limit =~ m/([^:]+):([^:]*)/;
698 if ( $limit =~ /available/ ) {
700 " (($query and datedue=0000-00-00) or ($query and datedue=0000-00-00 not lost=1) or ($query and datedue=0000-00-00 not lost=2))";
702 #$limit_search_desc.=" and available";
704 elsif ( ($limit_query) && ( index( $limit_query, $type, 0 ) > 0 ) ) {
705 if ( $limit_query !~ /\(/ ) {
707 substr( $limit_query, 0, index( $limit_query, $type, 0 ) )
709 . substr( $limit_query, index( $limit_query, $type, 0 ) )
713 substr( $limit_search_desc, 0,
714 index( $limit_search_desc, $type, 0 ) )
716 . substr( $limit_search_desc,
717 index( $limit_search_desc, $type, 0 ) )
723 chop $limit_search_desc;
724 $limit_query .= " or $limit )" if $limit;
725 $limit_search_desc .= " or $limit )" if $limit;
728 elsif ( ($limit_query) && ( $limit =~ /mc/ ) ) {
729 $limit_query .= " or $limit" if $limit;
730 $limit_search_desc .= " or $limit" if $limit;
733 # these are treated as AND
734 elsif ($limit_query) {
735 if ($limit =~ /branch/){
736 $limit_query .= " ) and ( $limit" if $limit;
737 $limit_search_desc .= " ) and ( $limit" if $limit;
739 $limit_query .= " or $limit" if $limit;
740 $limit_search_desc .= " or $limit" if $limit;
744 # otherwise, there is nothing but the limit
746 $limit_query .= "$limit" if $limit;
747 $limit_search_desc .= "$limit" if $limit;
751 # if there's also a query, we need to AND the limits to it
752 if ( ($limit_query) && ($query) ) {
753 $limit_query = " and (" . $limit_query . ")";
754 $limit_search_desc = " and ($limit_search_desc)" if $limit_search_desc;
757 $query .= $limit_query;
758 $human_search_desc .= $limit_search_desc;
760 # now normalize the strings
761 $query =~ s/ / /g; # remove extra spaces
762 $query =~ s/^ //g; # remove any beginning spaces
763 $query =~ s/:/=/g; # causes probs for server
764 $query =~ s/==/=/g; # remove double == from query
766 my $federated_query = $human_search_desc;
767 $federated_query =~ s/ / /g;
768 $federated_query =~ s/^ //g;
769 $federated_query =~ s/:/=/g;
770 my $federated_query_opensearch = $federated_query;
772 # my $federated_query_RPN = new ZOOM::Query::CCL2RPN( $query , C4::Context->ZConn('biblioserver'));
774 $human_search_desc =~ s/ / /g;
775 $human_search_desc =~ s/^ //g;
776 my $koha_query = $query;
778 #warn "QUERY:".$koha_query;
779 #warn "SEARCHDESC:".$human_search_desc;
780 #warn "FEDERATED QUERY:".$federated_query;
781 return ( undef, $human_search_desc, $koha_query, $federated_query );
784 # IMO this subroutine is pretty messy still -- it's responsible for
785 # building the HTML output for the template
787 my ( $searchdesc, $hits, $results_per_page, $offset, @marcresults ) = @_;
789 my $dbh = C4::Context->dbh;
793 my $span_terms_hashref;
794 for my $span_term ( split( / /, $searchdesc ) ) {
795 $span_term =~ s/(.*=|\)|\(|\+|\.)//g;
796 $span_terms_hashref->{$span_term}++;
799 #Build brancnames hash
801 #get branch information.....
804 $dbh->prepare("SELECT branchcode,branchname FROM branches")
805 ; # FIXME : use C4::Koha::GetBranches
807 while ( my $bdata = $bsth->fetchrow_hashref ) {
808 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
812 #find itemtype & itemtype image
815 $dbh->prepare("SELECT itemtype,description,imageurl,summary FROM itemtypes");
817 while ( my $bdata = $bsth->fetchrow_hashref ) {
818 $itemtypes{ $bdata->{'itemtype'} }->{description} =
819 $bdata->{'description'};
820 $itemtypes{ $bdata->{'itemtype'} }->{imageurl} = $bdata->{'imageurl'};
821 $itemtypes{ $bdata->{'itemtype'} }->{summary} = $bdata->{'summary'};
824 #search item field code
827 "select tagfield from marc_subfield_structure where kohafield like 'items.itemnumber'"
830 my ($itemtag) = $sth->fetchrow;
832 ## find column names of items related to MARC
833 my $sth2 = $dbh->prepare("SHOW COLUMNS from items");
835 my %subfieldstosearch;
836 while ( ( my $column ) = $sth2->fetchrow ) {
837 my ( $tagfield, $tagsubfield ) =
838 &GetMarcFromKohaField( "items." . $column, "" );
839 $subfieldstosearch{$column} = $tagsubfield;
843 if ( $hits && $offset + $results_per_page <= $hits ) {
844 $times = $offset + $results_per_page;
850 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
852 $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
854 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, '' );
856 # add image url if there is one
857 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} =~ /^http:/ ) {
858 $oldbiblio->{imageurl} =
859 $itemtypes{ $oldbiblio->{itemtype} }->{imageurl};
860 $oldbiblio->{description} =
861 $itemtypes{ $oldbiblio->{itemtype} }->{description};
864 $oldbiblio->{imageurl} =
865 getitemtypeimagesrc() . "/"
866 . $itemtypes{ $oldbiblio->{itemtype} }->{imageurl}
867 if ( $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
868 $oldbiblio->{description} =
869 $itemtypes{ $oldbiblio->{itemtype} }->{description};
872 # build summary if there is one (the summary is defined in itemtypes table
874 if ($itemtypes{ $oldbiblio->{itemtype} }->{summary}) {
875 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
876 my @fields = $marcrecord->fields();
877 foreach my $field (@fields) {
878 my $tag = $field->tag();
879 my $tagvalue = $field->as_string();
880 $summary =~ s/\[(.?.?.?.?)$tag\*(.*?)]/$1$tagvalue$2\[$1$tag$2]/g;
882 my @subf = $field->subfields;
883 for my $i (0..$#subf) {
884 my $subfieldcode = $subf[$i][0];
885 my $subfieldvalue = $subf[$i][1];
886 my $tagsubf = $tag.$subfieldcode;
887 $summary =~ s/\[(.?.?.?.?)$tagsubf(.*?)]/$1$subfieldvalue$2\[$1$tagsubf$2]/g;
891 $summary =~ s/\[(.*?)]//g;
892 $summary =~ s/\n/<br>/g;
893 $oldbiblio->{summary} = $summary;
895 # add spans to search term in results
896 foreach my $term ( keys %$span_terms_hashref ) {
899 my $old_term = $term;
900 if ( length($term) > 3 ) {
901 $term =~ s/(.*=|\)|\(|\+|\.|\?)//g;
903 #FIXME: is there a better way to do this?
904 $oldbiblio->{'title'} =~ s/$term/<span class=term>$&<\/span>/gi;
905 $oldbiblio->{'subtitle'} =~
906 s/$term/<span class=term>$&<\/span>/gi;
908 $oldbiblio->{'author'} =~ s/$term/<span class=term>$&<\/span>/gi;
909 $oldbiblio->{'publishercode'} =~ s/$term/<span class=term>$&<\/span>/gi;
910 $oldbiblio->{'place'} =~ s/$term/<span class=term>$&<\/span>/gi;
911 $oldbiblio->{'pages'} =~ s/$term/<span class=term>$&<\/span>/gi;
912 $oldbiblio->{'notes'} =~ s/$term/<span class=term>$&<\/span>/gi;
913 $oldbiblio->{'size'} =~ s/$term/<span class=term>$&<\/span>/gi;
923 $oldbiblio->{'toggle'} = $toggle;
924 my @fields = $marcrecord->field($itemtag);
927 my $ordered_count = 0;
928 my $onloan_count = 0;
929 my $wthdrawn_count = 0;
930 my $itemlost_count = 0;
931 my $itembinding_count = 0;
934 foreach my $field (@fields) {
936 foreach my $code ( keys %subfieldstosearch ) {
937 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
939 if ( $item->{wthdrawn} ) {
942 elsif ( $item->{notforloan} == -1 ) {
946 elsif ( $item->{itemlost} ) {
949 elsif ( $item->{binding} ) {
950 $itembinding_count++;
952 elsif ( ( $item->{onloan} ) && ( $item->{onloan} != '0000-00-00' ) )
959 if ( $item->{'homebranch'} ) {
960 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{count}++;
964 elsif ( $item->{'holdingbranch'} ) {
965 $items->{ $item->{'homebranch'} }->{count}++;
967 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{itemcallnumber} = $item->{itemcallnumber};
968 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{location} = $item->{location};
969 $items->{ $item->{'homebranch'}.'--'.$item->{'itemcallnumber'} }->{branchcode} = $item->{homebranch};
971 } # notforloan, item level and biblioitem level
972 for my $key ( keys %$items ) {
976 branchname => $branches{$items->{$key}->{branchcode}},
977 branchcode => $items->{$key}->{branchcode},
978 count => $items->{$key}->{count},
979 itemcallnumber => $items->{$key}->{itemcallnumber},
980 location => $items->{$key}->{location},
982 push @items_loop, $this_item;
984 $oldbiblio->{norequests} = $norequests;
985 $oldbiblio->{items_loop} = \@items_loop;
986 $oldbiblio->{onloancount} = $onloan_count;
987 $oldbiblio->{wthdrawncount} = $wthdrawn_count;
988 $oldbiblio->{itemlostcount} = $itemlost_count;
989 $oldbiblio->{bindingcount} = $itembinding_count;
990 $oldbiblio->{orderedcount} = $ordered_count;
993 # Ugh ... this is ugly, I'll re-write it better above then delete it
994 # my $norequests = 1;
998 # foreach my $itm (@items) {
999 # $norequests = 0 unless $itm->{'itemnotforloan'};
1002 # $oldbiblio->{'noitems'} = $noitems;
1003 # $oldbiblio->{'norequests'} = $norequests;
1004 # $oldbiblio->{'even'} = $even = not $even;
1005 # $oldbiblio->{'itemcount'} = $counts{'total'};
1006 # my $totalitemcounts = 0;
1007 # foreach my $key (keys %counts){
1008 # if ($key ne 'total'){
1009 # $totalitemcounts+= $counts{$key};
1010 # $oldbiblio->{'locationhash'}->{$key}=$counts{$key};
1013 # my ($locationtext, $locationtextonly, $notavailabletext) = ('','','');
1014 # foreach (sort keys %{$oldbiblio->{'locationhash'}}) {
1015 # if ($_ eq 'notavailable') {
1016 # $notavailabletext="Not available";
1017 # my $c=$oldbiblio->{'locationhash'}->{$_};
1018 # $oldbiblio->{'not-available-p'}=$c;
1020 # $locationtext.="$_";
1021 # my $c=$oldbiblio->{'locationhash'}->{$_};
1022 # if ($_ eq 'Item Lost') {
1023 # $oldbiblio->{'lost-p'} = $c;
1024 # } elsif ($_ eq 'Withdrawn') {
1025 # $oldbiblio->{'withdrawn-p'} = $c;
1026 # } elsif ($_ eq 'On Loan') {
1027 # $oldbiblio->{'on-loan-p'} = $c;
1029 # $locationtextonly.= $_;
1030 # $locationtextonly.= " ($c)<br/> " if $totalitemcounts > 1;
1032 # if ($totalitemcounts>1) {
1033 # $locationtext.=" ($c)<br/> ";
1037 # if ($notavailabletext) {
1038 # $locationtext.= $notavailabletext;
1040 # $locationtext=~s/, $//;
1042 # $oldbiblio->{'location'} = $locationtext;
1043 # $oldbiblio->{'location-only'} = $locationtextonly;
1044 # $oldbiblio->{'use-location-flags-p'} = 1;
1046 push( @newresults, $oldbiblio );
1054 ($countchanged,$listunchanged) = EditBiblios($listbiblios, $tagsubfield,$initvalue,$targetvalue,$test);
1056 this function changes all the values $initvalue in subfield $tag$subfield in any record in $listbiblios
1057 test parameter if set donot perform change to records in database.
1063 * $listbiblios is an array ref to marcrecords to be changed
1064 * $tagsubfield is the reference of the subfield to change.
1065 * $initvalue is the value to search the record for
1066 * $targetvalue is the value to set the subfield to
1067 * $test is to be set only not to perform changes in database.
1069 =item C<Output arg:>
1070 * $countchanged counts all the changes performed.
1071 * $listunchanged contains the list of all the biblionumbers of records unchanged.
1073 =item C<usage in the script:>
1077 my ($countchanged, $listunchanged) = EditBiblios($results->{RECORD}, $tagsubfield,$initvalue,$targetvalue);;
1078 #If one wants to display unchanged records, you should get biblios foreach @$listunchanged
1079 $template->param(countchanged => $countchanged, loopunchanged=>$listunchanged);
1083 my ($listbiblios,$tagsubfield,$initvalue,$targetvalue,$test)=@_;
1086 my ($tag,$subfield)=($1,$2) if ($tagsubfield=~/^(\d{1,3})(.)$/);
1087 my ($bntag,$bnsubf) = GetMarcFromKohaField('biblio.biblionumber');
1089 foreach my $usmarc (@$listbiblios){
1090 my $record=MARC::Record->new_from_usmarc($usmarc);
1093 $biblionumber = $record->subfield($bntag,$bnsubf);
1095 $biblionumber=$record->field($bntag)->data;
1097 #GetBiblionumber is to be written.
1098 #Could be replaced by TransformMarcToKoha (But Would be longer)
1099 if ($record->field($tag)){
1100 foreach my $field ($record->field($tag)){
1101 if ($field->delete_subfield('code' =>$subfield,'match'=>qr($initvalue))){
1103 $field->update($subfield,$targetvalue) if ($targetvalue);
1106 # warn $record->as_formatted;
1107 ModBiblio($record,$biblionumber,GetFrameworkCode($biblionumber)) unless ($test);
1109 push @unmatched, $biblionumber;
1112 return ($countmatched,\@unmatched);
1115 #----------------------------------------------------------------------
1117 # Non-Zebra GetRecords#
1118 #----------------------------------------------------------------------
1121 NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1126 $koha_query, $federated_query, $sort_by_ref,
1127 $servers_ref, $results_per_page, $offset,
1128 $expanded_facet, $branches, $query_type,
1131 my $result = NZanalyse($koha_query);
1133 # warn "==========".@$sort_by_ref[0];
1134 return (undef,NZorder($result,@$sort_by_ref[0],$results_per_page,$offset),undef);
1139 NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1140 the list is builded from inverted index in nozebra SQL table
1141 note that title is here only for convenience : the sorting will be very fast when requested on title
1142 if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1147 my ($string,$server) = @_;
1148 # $server contains biblioserver or authorities, depending on what we search on.
1149 warn "querying : $string on $server";
1150 $server='biblioserver' unless $server;
1151 # if we have a ", replace the content to discard temporarily any and/or/not inside
1153 if ($string =~/"/) {
1154 $string =~ s/"(.*?)"/__X__/;
1156 # print "commacontent : $commacontent\n";
1158 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1159 # then, call again NZanalyse with $left and $right
1160 # (recursive until we find a leaf (=> something without and/or/not)
1161 $string =~ /(.*)( and | or | not | AND | OR | NOT )(.*)/;
1164 my $operand = lc($2);
1165 # it's not a leaf, we have a and/or/not
1167 # reintroduce comma content if needed
1168 $right =~ s/__X__/"$commacontent"/ if $commacontent;
1169 $left =~ s/__X__/"$commacontent"/ if $commacontent;
1170 # warn "node : $left / $operand / $right\n";
1171 my $leftresult = NZanalyse($left,$server);
1172 my $rightresult = NZanalyse($right,$server);
1173 # OK, we have the results for right and left part of the query
1174 # depending of operand, intersect, union or exclude both lists
1175 # to get a result list
1176 if ($operand eq ' and ') {
1177 my @leftresult = split /;/, $leftresult;
1178 # my @rightresult = split /;/,$leftresult;
1180 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
1181 # the result is stored twice, to have the same weight for AND than OR.
1182 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
1183 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
1184 foreach (@leftresult) {
1185 if ($rightresult =~ "$_;") {
1186 $finalresult .= "$_;$_;";
1189 return $finalresult;
1190 } elsif ($operand eq ' or ') {
1191 # just merge the 2 strings
1192 return $leftresult.$rightresult;
1193 } elsif ($operand eq ' not ') {
1194 my @leftresult = split /;/, $leftresult;
1195 # my @rightresult = split /;/,$leftresult;
1197 foreach (@leftresult) {
1198 unless ($rightresult =~ "$_;") {
1199 $finalresult .= "$_;";
1202 return $finalresult;
1204 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1205 die "error : operand unknown : $operand for $string";
1207 # it's a leaf, do the real SQL query and return the result
1209 $string =~ s/__X__/"$commacontent"/ if $commacontent;
1210 $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|<|>|&|\+|\*|\// /g;
1211 # warn "leaf : $string\n";
1212 # parse the string in in operator/operand/value again
1213 $string =~ /(.*)(=|>|>=|<|<=)(.*)/;
1218 # automatic replace for short operators
1219 $left='title' if $left eq 'ti';
1220 $left='author' if $left eq 'au';
1221 $left='publisher' if $left eq 'pb';
1222 $left='subject' if $left eq 'su';
1223 $left='koha-Auth-Number' if $left eq 'an';
1225 #do a specific search
1226 my $dbh = C4::Context->dbh;
1227 $operator='LIKE' if $operator eq '=' and $right=~ /%/;
1228 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND indexname=? AND value $operator ?");
1229 warn "$left / $operator / $right\n";
1230 # split each word, query the DB and build the biblionumbers result
1231 foreach (split / /,$right) {
1234 # warn "EXECUTE : $server, $left, $_";
1235 $sth->execute($server, $left, $_);
1236 while (my $line = $sth->fetchrow) {
1237 $biblionumbers .= $line;
1238 # warn "result : $line";
1240 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1242 my @leftresult = split /;/, $biblionumbers;
1244 foreach (@leftresult) {
1245 if ($results =~ "$_;") {
1251 $results = $biblionumbers;
1255 #do a complete search (all indexes)
1256 my $dbh = C4::Context->dbh;
1257 my $sth = $dbh->prepare("SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?");
1258 # split each word, query the DB and build the biblionumbers result
1259 foreach (split / /,$string) {
1260 # warn "search on all indexes on $_";
1263 $sth->execute($server, $_);
1264 while (my $line = $sth->fetchrow) {
1265 $biblionumbers .= $line;
1267 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
1269 my @leftresult = split /;/, $biblionumbers;
1271 foreach (@leftresult) {
1272 if ($results =~ "$_;") {
1278 $results = $biblionumbers;
1282 # warn "return : $results for LEAF : $string";
1288 my ($biblionumbers, $ordering,$results_per_page,$offset) = @_;
1289 # order title asc by default
1290 # $ordering = '1=36 <i' unless $ordering;
1291 $results_per_page=20 unless $results_per_page;
1292 $offset = 0 unless $offset;
1293 my $dbh = C4::Context->dbh;
1295 # order by POPULARITY
1297 if ($ordering =~ /1=9523/) {
1300 # popularity is not in MARC record, it's builded from a specific query
1301 my $sth = $dbh->prepare("select sum(issues) from items where biblionumber=?");
1302 foreach (split /;/,$biblionumbers) {
1303 my ($biblionumber,$title) = split /,/,$_;
1304 $result{$biblionumber}=GetMarcBiblio($biblionumber);
1305 $sth->execute($biblionumber);
1306 my $popularity= $sth->fetchrow ||0;
1307 # hint : the key is popularity.title because we can have
1308 # many results with the same popularity. In this cas, sub-ordering is done by title
1309 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
1310 # (un-frequent, I agree, but we won't forget anything that way ;-)
1311 $popularity{sprintf("%10d",$popularity).$title.$biblionumber} = $biblionumber;
1313 # sort the hash and return the same structure as GetRecords (Zebra querying)
1316 if ($ordering eq '1=9523 >i') { # sort popularity DESC
1317 foreach my $key (sort {$b cmp $a} (keys %popularity)) {
1318 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1320 } else { # sort popularity ASC
1321 foreach my $key (sort (keys %popularity)) {
1322 $result_hash->{'RECORDS'}[$numbers++] = $result{$popularity{$key}}->as_usmarc();
1326 $result_hash->{'hits'} = $numbers;
1327 $finalresult->{'biblioserver'} = $result_hash;
1328 return $finalresult;
1332 } elsif ($ordering eq '1=1003 <i'){
1334 foreach (split /;/,$biblionumbers) {
1335 my ($biblionumber,$title) = split /,/,$_;
1336 my $record=GetMarcBiblio($biblionumber);
1338 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1339 $author=$record->subfield('200','f');
1340 $author=$record->subfield('700','a') unless $author;
1342 $author=$record->subfield('100','a');
1344 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1345 # and we don't want to get only 1 result for each of them !!!
1346 $result{$author.$biblionumber}=$record;
1348 # sort the hash and return the same structure as GetRecords (Zebra querying)
1351 if ($ordering eq '1=1003 <i') { # sort by author desc
1352 foreach my $key (sort (keys %result)) {
1353 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1355 } else { # sort by author ASC
1356 foreach my $key (sort { $a cmp $b } (keys %result)) {
1357 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1361 $result_hash->{'hits'} = $numbers;
1362 $finalresult->{'biblioserver'} = $result_hash;
1363 return $finalresult;
1365 # ORDER BY callnumber
1367 } elsif ($ordering eq '1=20 <i'){
1369 foreach (split /;/,$biblionumbers) {
1370 my ($biblionumber,$title) = split /,/,$_;
1371 my $record=GetMarcBiblio($biblionumber);
1373 my ($callnumber_tag,$callnumber_subfield)=GetMarcFromKohaField($dbh,'items.itemcallnumber');
1374 ($callnumber_tag,$callnumber_subfield)= GetMarcFromKohaField('biblioitems.callnumber') unless $callnumber_tag;
1375 if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
1376 $callnumber=$record->subfield('200','f');
1378 $callnumber=$record->subfield('100','a');
1380 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1381 # and we don't want to get only 1 result for each of them !!!
1382 $result{$callnumber.$biblionumber}=$record;
1384 # sort the hash and return the same structure as GetRecords (Zebra querying)
1387 if ($ordering eq '1=1003 <i') { # sort by title desc
1388 foreach my $key (sort (keys %result)) {
1389 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1391 } else { # sort by title ASC
1392 foreach my $key (sort { $a cmp $b } (keys %result)) {
1393 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1397 $result_hash->{'hits'} = $numbers;
1398 $finalresult->{'biblioserver'} = $result_hash;
1399 return $finalresult;
1400 } elsif ($ordering =~ /1=31/){ #pub year
1402 foreach (split /;/,$biblionumbers) {
1403 my ($biblionumber,$title) = split /,/,$_;
1404 my $record=GetMarcBiblio($biblionumber);
1405 my ($publicationyear_tag,$publicationyear_subfield)=GetMarcFromKohaField($dbh,'biblioitems.publicationyear');
1406 my $publicationyear=$record->subfield($publicationyear_tag,$publicationyear_subfield);
1407 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1408 # and we don't want to get only 1 result for each of them !!!
1409 $result{$publicationyear.$biblionumber}=$record;
1411 # sort the hash and return the same structure as GetRecords (Zebra querying)
1414 if ($ordering eq '1=31 <i') { # sort by pubyear desc
1415 foreach my $key (sort (keys %result)) {
1416 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1418 } else { # sort by pub year ASC
1419 foreach my $key (sort { $b cmp $a } (keys %result)) {
1420 $result_hash->{'RECORDS'}[$numbers++] = $result{$key}->as_usmarc();
1424 $result_hash->{'hits'} = $numbers;
1425 $finalresult->{'biblioserver'} = $result_hash;
1426 return $finalresult;
1430 } elsif ($ordering =~ /1=36/) {
1431 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
1433 foreach (split /;/,$biblionumbers) {
1434 my ($biblionumber,$title) = split /,/,$_;
1435 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
1436 # and we don't want to get only 1 result for each of them !!!
1437 # hint & speed improvement : we can order without reading the record
1438 # so order, and read records only for the requested page !
1439 $result{$title.$biblionumber}=$biblionumber;
1441 # sort the hash and return the same structure as GetRecords (Zebra querying)
1444 if ($ordering eq '1=36 <i') { # sort by title desc
1445 foreach my $key (sort (keys %result)) {
1446 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1448 } else { # sort by title ASC
1449 foreach my $key (sort { $b cmp $a } (keys %result)) {
1450 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1453 # limit the $results_per_page to result size if it's more
1454 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1455 # for the requested page, replace biblionumber by the complete record
1456 # speed improvement : avoid reading too much things
1457 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1458 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1461 $result_hash->{'hits'} = $numbers;
1462 $finalresult->{'biblioserver'} = $result_hash;
1463 return $finalresult;
1468 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
1471 foreach (split /;/,$biblionumbers) {
1472 my ($biblionumber,$title) = split /,/,$_;
1473 $title =~ /(.*)-(\d)/;
1476 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
1477 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
1478 # biblio N has ranking = 6
1479 $count_ranking{$biblionumber} += $ranking;
1481 # build the result by "inverting" the count_ranking hash
1482 # 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
1484 foreach (keys %count_ranking) {
1485 $result{sprintf("%10d",$count_ranking{$_}).'-'.$_} = $_;
1487 # sort the hash and return the same structure as GetRecords (Zebra querying)
1490 foreach my $key (sort {$b cmp $a} (keys %result)) {
1491 $result_hash->{'RECORDS'}[$numbers++] = $result{$key};
1493 # limit the $results_per_page to result size if it's more
1494 $results_per_page = $numbers-1 if $numbers < $results_per_page;
1495 # for the requested page, replace biblionumber by the complete record
1496 # speed improvement : avoid reading too much things
1497 for (my $counter=$offset;$counter<=$offset+$results_per_page;$counter++) {
1498 $result_hash->{'RECORDS'}[$counter] = GetMarcBiblio($result_hash->{'RECORDS'}[$counter])->as_usmarc;
1501 $result_hash->{'hits'} = $numbers;
1502 $finalresult->{'biblioserver'} = $result_hash;
1503 return $finalresult;
1507 END { } # module clean-up code here (global destructor)
1514 Koha Developement team <info@koha.org>