?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 17698: (follow-up) Fixing some small issues
[koha.git] / C4 / Search.pm
blob3a3a57dc259e7e20b11242c7b17981bab75bc331
1 package C4::Search;
2
3 # This file is part of Koha.
4 #
5 # Koha is free software; you can redistribute it and/or modify it
6 # under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3 of the License, or
8 # (at your option) any later version.
9 #
10 # Koha is distributed in the hope that it will be useful, but
11 # WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
14 #
15 # You should have received a copy of the GNU General Public License
16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
17
18 use strict;
19 #use warnings; FIXME - Bug 2505
20 require Exporter;
21 use C4::Context;
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
24 use Koha::DateUtils;
25 use Koha::Libraries;
26 use Lingua::Stem;
27 use C4::Search::PazPar2;
28 use XML::Simple;
29 use C4::XSLT;
30 use C4::Reserves; # GetReserveStatus
31 use C4::Debug;
32 use C4::Charset;
33 use Koha::AuthorisedValues;
34 use Koha::ItemTypes;
35 use Koha::Libraries;
36 use Koha::Patrons;
37 use YAML;
38 use URI::Escape;
39 use Business::ISBN;
40 use MARC::Record;
41 use MARC::Field;
42 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
43
44 BEGIN {
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
46 }
47
48 =head1 NAME
49
50 C4::Search - Functions for searching the Koha catalog.
51
52 =head1 SYNOPSIS
53
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
55
56 =head1 DESCRIPTION
57
58 This module provides searching functions for Koha's bibliographic databases
59
60 =head1 FUNCTIONS
61
62 =cut
63
64 @ISA = qw(Exporter);
65 @EXPORT = qw(
66 &FindDuplicate
67 &SimpleSearch
68 &searchResults
69 &getRecords
70 &buildQuery
71 &GetDistinctValues
72 &enabled_staff_search_views
73 );
74
75 # make all your functions, whether exported or not;
76
77 =head2 FindDuplicate
78
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
80
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
82
83 =cut
84
85 sub FindDuplicate {
86 my ($record) = @_;
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $record, '' );
89 my $sth;
90 my $query;
91 my $search;
92 my $type;
93 my ( $biblionumber, $title );
94
95 # search duplicate on ISBN, easy and fast..
96 # ... normalize first
97 if ( $result->{isbn} ) {
98 $result->{isbn} =~ s/\(.*$//;
99 $result->{isbn} =~ s/\s+$//;
100 $query = "isbn:$result->{isbn}";
101 }
102 else {
103 my $QParser;
104 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
105 my $titleindex;
106 my $authorindex;
107 my $op;
108
109 if ($QParser) {
110 $titleindex = 'title|exact';
111 $authorindex = 'author|exact';
112 $op = '&&';
113 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
114 } else {
115 $titleindex = 'ti,ext';
116 $authorindex = 'au,ext';
117 $op = 'and';
118 }
119
120 $result->{title} =~ s /\\//g;
121 $result->{title} =~ s /\"//g;
122 $result->{title} =~ s /\(//g;
123 $result->{title} =~ s /\)//g;
124
125 # FIXME: instead of removing operators, could just do
126 # quotes around the value
127 $result->{title} =~ s/(and|or|not)//g;
128 $query = "$titleindex:\"$result->{title}\"";
129 if ( $result->{author} ) {
130 $result->{author} =~ s /\\//g;
131 $result->{author} =~ s /\"//g;
132 $result->{author} =~ s /\(//g;
133 $result->{author} =~ s /\)//g;
134
135 # remove valid operators
136 $result->{author} =~ s/(and|or|not)//g;
137 $query .= " $op $authorindex:\"$result->{author}\"";
138 }
139 }
140
141 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
142 my @results;
143 if (!defined $error) {
144 foreach my $possible_duplicate_record (@{$searchresults}) {
145 my $marcrecord = new_record_from_zebra(
146 'biblioserver',
147 $possible_duplicate_record
148 );
149
150 my $result = TransformMarcToKoha( $marcrecord, '' );
151
152 # FIXME :: why 2 $biblionumber ?
153 if ($result) {
154 push @results, $result->{'biblionumber'};
155 push @results, $result->{'title'};
156 }
157 }
158 }
159 return @results;
160 }
161
162 =head2 SimpleSearch
163
164 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
165
166 This function provides a simple search API on the bibliographic catalog
167
168 =over 2
169
170 =item C<input arg:>
171
172 * $query can be a simple keyword or a complete CCL query
173 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
174 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
175 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
176 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
177
178
179 =item C<Return:>
180
181 Returns an array consisting of three elements
182 * $error is undefined unless an error is detected
183 * $results is a reference to an array of records.
184 * $total_hits is the number of hits that would have been returned with no limit
185
186 If an error is returned the two other return elements are undefined. If error itself is undefined
187 the other two elements are always defined
188
189 =item C<usage in the script:>
190
191 =back
192
193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
194
195 if (defined $error) {
196 $template->param(query_error => $error);
197 warn "error: ".$error;
198 output_html_with_http_headers $input, $cookie, $template->output;
199 exit;
200 }
201
202 my $hits = @{$marcresults};
203 my @results;
204
205 for my $r ( @{$marcresults} ) {
206 my $marcrecord = MARC::File::USMARC::decode($r);
207 my $biblio = TransformMarcToKoha($marcrecord,q{});
208
209 #build the iarray of hashs for the template.
210 push @results, {
211 title => $biblio->{'title'},
212 subtitle => $biblio->{'subtitle'},
213 biblionumber => $biblio->{'biblionumber'},
214 author => $biblio->{'author'},
215 publishercode => $biblio->{'publishercode'},
216 publicationyear => $biblio->{'publicationyear'},
217 };
218
219 }
220
221 $template->param(result=>\@results);
222
223 =cut
224
225 sub SimpleSearch {
226 my ( $query, $offset, $max_results, $servers, %options ) = @_;
227
228 return ( 'No query entered', undef, undef ) unless $query;
229 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
230 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
231 my @zoom_queries;
232 my @tmpresults;
233 my @zconns;
234 my $results = [];
235 my $total_hits = 0;
236
237 my $QParser;
238 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
239 if ($QParser) {
240 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
241 }
242
243 # Initialize & Search Zebra
244 for ( my $i = 0 ; $i < @servers ; $i++ ) {
245 eval {
246 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
247 if ($QParser) {
248 $query =~ s/=/:/g unless $options{skip_normalize};
249 $QParser->parse( $query );
250 $query = $QParser->target_syntax($servers[$i]);
251 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
252 } else {
253 $query =~ s/:/=/g unless $options{skip_normalize};
254 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
255 }
256 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
257
258 # error handling
259 my $error =
260 $zconns[$i]->errmsg() . " ("
261 . $zconns[$i]->errcode() . ") "
262 . $zconns[$i]->addinfo() . " "
263 . $zconns[$i]->diagset();
264
265 return ( $error, undef, undef ) if $zconns[$i]->errcode();
266 };
267 if ($@) {
268
269 # caught a ZOOM::Exception
270 my $error =
271 $@->message() . " ("
272 . $@->code() . ") "
273 . $@->addinfo() . " "
274 . $@->diagset();
275 warn $error." for query: $query";
276 return ( $error, undef, undef );
277 }
278 }
279
280 _ZOOM_event_loop(
281 \@zconns,
282 \@tmpresults,
283 sub {
284 my ($i, $size) = @_;
285 my $first_record = defined($offset) ? $offset + 1 : 1;
286 my $hits = $tmpresults[ $i - 1 ]->size();
287 $total_hits += $hits;
288 my $last_record = $hits;
289 if ( defined $max_results && $offset + $max_results < $hits ) {
290 $last_record = $offset + $max_results;
291 }
292
293 for my $j ( $first_record .. $last_record ) {
294 my $record = eval {
295 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
296 ; # 0 indexed
297 };
298 push @{$results}, $record if defined $record;
299 }
300 }
301 );
302
303 foreach my $zoom_query (@zoom_queries) {
304 $zoom_query->destroy();
305 }
306
307 return ( undef, $results, $total_hits );
308 }
309
310 =head2 getRecords
311
312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
313
314 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
315 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
316 $query_type, $scan
317 );
318
319 The all singing, all dancing, multi-server, asynchronous, scanning,
320 searching, record nabbing, facet-building
321
322 See verbse embedded documentation.
323
324 =cut
325
326 sub getRecords {
327 my (
328 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
329 $results_per_page, $offset, $expanded_facet, $branches,
330 $itemtypes, $query_type, $scan, $opac
331 ) = @_;
332
333 my @servers = @$servers_ref;
334 my @sort_by = @$sort_by_ref;
335 $offset = 0 if $offset < 0;
336
337 # Initialize variables for the ZOOM connection and results object
338 my $zconn;
339 my @zconns;
340 my @results;
341 my $results_hashref = ();
342
343 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
344 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
345
346 # Initialize variables for the faceted results objects
347 my $facets_counter = {};
348 my $facets_info = {};
349 my $facets = getFacets();
350
351 my @facets_loop; # stores the ref to array of hashes for template facets loop
352
353 ### LOOP THROUGH THE SERVERS
354 for ( my $i = 0 ; $i < @servers ; $i++ ) {
355 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
356
357 # perform the search, create the results objects
358 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
359 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
360
361 #$query_to_use = $simple_query if $scan;
362 warn $simple_query if ( $scan and $DEBUG );
363
364 # Check if we've got a query_type defined, if so, use it
365 eval {
366 if ($query_type) {
367 if ($query_type =~ /^ccl/) {
368 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
370 } elsif ($query_type =~ /^cql/) {
371 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
372 } elsif ($query_type =~ /^pqf/) {
373 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
374 } else {
375 warn "Unknown query_type '$query_type'. Results undetermined.";
376 }
377 } elsif ($scan) {
378 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
379 } else {
380 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
381 }
382 };
383 if ($@) {
384 warn "WARNING: query problem with $query_to_use " . $@;
385 }
386
387 # Concatenate the sort_by limits and pass them to the results object
388 # Note: sort will override rank
389 my $sort_by;
390 foreach my $sort (@sort_by) {
391 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
392 $sort_by .= "1=1003 <i ";
393 }
394 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
395 $sort_by .= "1=1003 >i ";
396 }
397 elsif ( $sort eq "popularity_asc" ) {
398 $sort_by .= "1=9003 <i ";
399 }
400 elsif ( $sort eq "popularity_dsc" ) {
401 $sort_by .= "1=9003 >i ";
402 }
403 elsif ( $sort eq "call_number_asc" ) {
404 $sort_by .= "1=8007 <i ";
405 }
406 elsif ( $sort eq "call_number_dsc" ) {
407 $sort_by .= "1=8007 >i ";
408 }
409 elsif ( $sort eq "pubdate_asc" ) {
410 $sort_by .= "1=31 <i ";
411 }
412 elsif ( $sort eq "pubdate_dsc" ) {
413 $sort_by .= "1=31 >i ";
414 }
415 elsif ( $sort eq "acqdate_asc" ) {
416 $sort_by .= "1=32 <i ";
417 }
418 elsif ( $sort eq "acqdate_dsc" ) {
419 $sort_by .= "1=32 >i ";
420 }
421 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
422 $sort_by .= "1=4 <i ";
423 }
424 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
425 $sort_by .= "1=4 >i ";
426 }
427 else {
428 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
429 }
430 }
431 if ( $sort_by && !$scan && $results[$i] ) {
432 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
433 warn "WARNING sort $sort_by failed";
434 }
435 }
436 } # finished looping through servers
437
438 # The big moment: asynchronously retrieve results from all servers
439 _ZOOM_event_loop(
440 \@zconns,
441 \@results,
442 sub {
443 my ( $i, $size ) = @_;
444 my $results_hash;
445
446 # loop through the results
447 $results_hash->{'hits'} = $size;
448 my $times;
449 if ( $offset + $results_per_page <= $size ) {
450 $times = $offset + $results_per_page;
451 }
452 else {
453 $times = $size;
454 }
455
456 for ( my $j = $offset ; $j < $times ; $j++ ) {
457 my $records_hash;
458 my $record;
459
460 ## Check if it's an index scan
461 if ($scan) {
462 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
463
464 # here we create a minimal MARC record and hand it off to the
465 # template just like a normal result ... perhaps not ideal, but
466 # it works for now
467 my $tmprecord = MARC::Record->new();
468 $tmprecord->encoding('UTF-8');
469 my $tmptitle;
470 my $tmpauthor;
471
472 # the minimal record in author/title (depending on MARC flavour)
473 if ( C4::Context->preference("marcflavour") eq
474 "UNIMARC" )
475 {
476 $tmptitle = MARC::Field->new(
477 '200', ' ', ' ',
478 a => $term,
479 f => $occ
480 );
481 $tmprecord->append_fields($tmptitle);
482 }
483 else {
484 $tmptitle =
485 MARC::Field->new( '245', ' ', ' ', a => $term, );
486 $tmpauthor =
487 MARC::Field->new( '100', ' ', ' ', a => $occ, );
488 $tmprecord->append_fields($tmptitle);
489 $tmprecord->append_fields($tmpauthor);
490 }
491 $results_hash->{'RECORDS'}[$j] =
492 $tmprecord->as_usmarc();
493 }
494
495 # not an index scan
496 else {
497 $record = $results[ $i - 1 ]->record($j)->raw();
498 # warn "RECORD $j:".$record;
499 $results_hash->{'RECORDS'}[$j] = $record;
500 }
501
502 }
503 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
504
505 # Fill the facets while we're looping, but only for the
506 # biblioserver and not for a scan
507 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
508 $facets_counter = GetFacets( $results[ $i - 1 ] );
509 $facets_info = _get_facets_info( $facets );
510 }
511
512 # BUILD FACETS
513 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
514 for my $link_value (
515 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
516 keys %$facets_counter
517 )
518 {
519 my $expandable;
520 my $number_of_facets;
521 my @this_facets_array;
522 for my $one_facet (
523 sort {
524 $facets_counter->{$link_value}
525 ->{$b} <=> $facets_counter->{$link_value}
526 ->{$a}
527 } keys %{ $facets_counter->{$link_value} }
528 )
529 {
530 $number_of_facets++;
531 if ( ( $number_of_facets <= 5 )
532 || ( $expanded_facet eq $link_value )
533 || ( $facets_info->{$link_value}->{'expanded'} )
534 )
535 {
536
537 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
538 my $facet_link_value = $one_facet;
539 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
540
541 # fix the length that will display in the label,
542 my $facet_label_value = $one_facet;
543 my $facet_max_length = C4::Context->preference(
544 'FacetLabelTruncationLength')
545 || 20;
546 $facet_label_value =
547 substr( $one_facet, 0, $facet_max_length )
548 . "..."
549 if length($facet_label_value) >
550 $facet_max_length;
551
552 # if it's a branch, label by the name, not the code,
553 if ( $link_value =~ /branch/ ) {
554 if ( defined $branches
555 && ref($branches) eq "HASH"
556 && defined $branches->{$one_facet}
557 && ref( $branches->{$one_facet} ) eq
558 "HASH" )
559 {
560 $facet_label_value =
561 $branches->{$one_facet}
562 ->{'branchname'};
563 }
564 else {
565 $facet_label_value = "*";
566 }
567 }
568
569 # if it's a itemtype, label by the name, not the code,
570 if ( $link_value =~ /itype/ ) {
571 if ( defined $itemtypes
572 && ref($itemtypes) eq "HASH"
573 && defined $itemtypes->{$one_facet}
574 && ref( $itemtypes->{$one_facet} ) eq
575 "HASH" )
576 {
577 $facet_label_value =
578 $itemtypes->{$one_facet}
579 ->{translated_description};
580 }
581 }
582
583 # also, if it's a location code, use the name instead of the code
584 if ( $link_value =~ /location/ ) {
585 # TODO Retrieve all authorised values at once, instead of 1 query per entry
586 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
587 $facet_label_value = $av->count ? $av->next->opac_description : '';
588 }
589
590 # also, if it's a collection code, use the name instead of the code
591 if ( $link_value =~ /ccode/ ) {
592 # TODO Retrieve all authorised values at once, instead of 1 query per entry
593 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
594 $facet_label_value = $av->count ? $av->next->opac_description : '';
595 }
596
597 # but we're down with the whole label being in the link's title.
598 push @this_facets_array,
599 {
600 facet_count =>
601 $facets_counter->{$link_value}
602 ->{$one_facet},
603 facet_label_value => $facet_label_value,
604 facet_title_value => $one_facet,
605 facet_link_value => $facet_link_value,
606 type_link_value => $link_value,
607 }
608 if ($facet_label_value);
609 }
610 }
611
612 # handle expanded option
613 unless ( $facets_info->{$link_value}->{'expanded'} ) {
614 $expandable = 1
615 if ( ( $number_of_facets > 5 )
616 && ( $expanded_facet ne $link_value ) );
617 }
618 push @facets_loop,
619 {
620 type_link_value => $link_value,
621 type_id => $link_value . "_id",
622 "type_label_"
623 . $facets_info->{$link_value}->{'label_value'} =>
624 1,
625 facets => \@this_facets_array,
626 expandable => $expandable,
627 expand => $link_value,
628 }
629 unless (
630 (
631 $facets_info->{$link_value}->{'label_value'} =~
632 /Libraries/
633 )
634 and ( Koha::Libraries->search->count == 1 )
635 );
636 }
637 }
638 }
639 );
640
641 # This sorts the facets into alphabetical order
642 if (@facets_loop) {
643 foreach my $f (@facets_loop) {
644 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
645 }
646 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
647 }
648
649 return ( undef, $results_hashref, \@facets_loop );
650 }
651
652 sub GetFacets {
653
654 my $rs = shift;
655 my $facets;
656
657 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
658 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
659
660 if ( $indexing_mode eq 'dom' &&
661 $use_zebra_facets ) {
662 $facets = _get_facets_from_zebra( $rs );
663 } else {
664 $facets = _get_facets_from_records( $rs );
665 }
666
667 return $facets;
668 }
669
670 sub _get_facets_from_records {
671
672 my $rs = shift;
673
674 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
675 my $facets_config = getFacets();
676 my $facets = {};
677 my $size = $rs->size();
678 my $jmax = $size > $facets_maxrecs
679 ? $facets_maxrecs
680 : $size;
681
682 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
683
684 my $marc_record = new_record_from_zebra (
685 'biblioserver',
686 $rs->record( $j )->raw()
687 );
688
689 if ( ! defined $marc_record ) {
690 warn "ERROR DECODING RECORD - $@: " .
691 $rs->record( $j )->raw();
692 next;
693 }
694
695 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
696 }
697
698 return $facets;
699 }
700
701 =head2 _get_facets_data_from_record
702
703 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
704
705 Internal function that extracts facets information from a MARC::Record object
706 and populates $facets_counter for using in getRecords.
707
708 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
709 facets for Zebra).
710
711 =cut
712
713 sub _get_facets_data_from_record {
714
715 my ( $marc_record, $facets, $facets_counter ) = @_;
716
717 for my $facet (@$facets) {
718
719 my @used_datas = ();
720
721 foreach my $tag ( @{ $facet->{ tags } } ) {
722
723 # tag number is the first three digits
724 my $tag_num = substr( $tag, 0, 3 );
725 # subfields are the remainder
726 my $subfield_letters = substr( $tag, 3 );
727
728 my @fields = $marc_record->field( $tag_num );
729 foreach my $field (@fields) {
730 # If $field->indicator(1) eq 'z', it means it is a 'see from'
731 # field introduced because of IncludeSeeFromInSearches, so skip it
732 next if $field->indicator(1) eq 'z';
733
734 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
735
736 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
737 push @used_datas, $data;
738 $facets_counter->{ $facet->{ idx } }->{ $data }++;
739 }
740 }
741 }
742 }
743 }
744
745 =head2 _get_facets_from_zebra
746
747 my $facets = _get_facets_from_zebra( $result_set )
748
749 Retrieves facets for a specified result set. It loops through the facets defined
750 in C4::Koha::getFacets and returns a hash with the following structure:
751
752 { facet_idx => {
753 facet_value => count
754 },
755 ...
756 }
757
758 =cut
759
760 sub _get_facets_from_zebra {
761
762 my $rs = shift;
763
764 # save current elementSetName
765 my $elementSetName = $rs->option( 'elementSetName' );
766
767 my $facets_loop = getFacets();
768 my $facets_data = {};
769 # loop through defined facets and fill the facets hashref
770 foreach my $facet ( @$facets_loop ) {
771
772 my $idx = $facet->{ idx };
773 my $sep = $facet->{ sep };
774 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
775 if ( $facet_values ) {
776 # we've actually got a result
777 $facets_data->{ $idx } = $facet_values;
778 }
779 }
780 # set elementSetName to its previous value to avoid side effects
781 $rs->option( elementSetName => $elementSetName );
782
783 return $facets_data;
784 }
785
786 =head2 _get_facet_from_result_set
787
788 my $facet_values =
789 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
790
791 Internal function that extracts facet information for a specific index ($facet_idx) and
792 returns a hash containing facet values and count:
793
794 {
795 $facet_value => $count ,
796 ...
797 }
798
799 Warning: this function has the side effect of changing the elementSetName for the result
800 set. It is a helper function for the main loop, which takes care of backing it up for
801 restoring.
802
803 =cut
804
805 sub _get_facet_from_result_set {
806
807 my $facet_idx = shift;
808 my $rs = shift;
809 my $sep = shift;
810
811 my $internal_sep = '<*>';
812 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
813
814 return if ( ! defined $facet_idx || ! defined $rs );
815 # zebra's facet element, untokenized index
816 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
817 # configure zebra results for retrieving the desired facet
818 $rs->option( elementSetName => $facet_element );
819 # get the facet record from result set
820 my $facet = $rs->record( 0 )->raw;
821 # if the facet has no restuls...
822 return if !defined $facet;
823 # TODO: benchmark DOM vs. SAX performance
824 my $facet_dom = XML::LibXML->load_xml(
825 string => ($facet)
826 );
827 my @terms = $facet_dom->getElementsByTagName('term');
828 return if ! @terms;
829
830 my $facets = {};
831 foreach my $term ( @terms ) {
832 my $facet_value = $term->textContent;
833 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
834 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
835 }
836
837 return $facets;
838 }
839
840 =head2 _get_facets_info
841
842 my $facets_info = C4::Search::_get_facets_info( $facets )
843
844 Internal function that extracts facets information and properly builds
845 the data structure needed to render facet labels.
846
847 =cut
848
849 sub _get_facets_info {
850
851 my $facets = shift;
852
853 my $facets_info = {};
854
855 for my $facet ( @$facets ) {
856 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
857 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
858 }
859
860 return $facets_info;
861 }
862
863 sub pazGetRecords {
864 my (
865 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
866 $results_per_page, $offset, $expanded_facet, $branches,
867 $query_type, $scan
868 ) = @_;
869 #NOTE: Parameter $branches is not used here !
870
871 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
872 $paz->init();
873 $paz->search($simple_query);
874 sleep 1; # FIXME: WHY?
875
876 # do results
877 my $results_hashref = {};
878 my $stats = XMLin($paz->stat);
879 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
880
881 # for a grouped search result, the number of hits
882 # is the number of groups returned; 'bib_hits' will have
883 # the total number of bibs.
884 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
885 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
886
887 HIT: foreach my $hit (@{ $results->{'hit'} }) {
888 my $recid = $hit->{recid}->[0];
889
890 my $work_title = $hit->{'md-work-title'}->[0];
891 my $work_author;
892 if (exists $hit->{'md-work-author'}) {
893 $work_author = $hit->{'md-work-author'}->[0];
894 }
895 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
896
897 my $result_group = {};
898 $result_group->{'group_label'} = $group_label;
899 $result_group->{'group_merge_key'} = $recid;
900
901 my $count = 1;
902 if (exists $hit->{count}) {
903 $count = $hit->{count}->[0];
904 }
905 $result_group->{'group_count'} = $count;
906
907 for (my $i = 0; $i < $count; $i++) {
908 # FIXME -- may need to worry about diacritics here
909 my $rec = $paz->record($recid, $i);
910 push @{ $result_group->{'RECORDS'} }, $rec;
911 }
912
913 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
914 }
915
916 # pass through facets
917 my $termlist_xml = $paz->termlist('author,subject');
918 my $terms = XMLin($termlist_xml, forcearray => 1);
919 my @facets_loop = ();
920 #die Dumper($results);
921 # foreach my $list (sort keys %{ $terms->{'list'} }) {
922 # my @facets = ();
923 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
924 # push @facets, {
925 # facet_label_value => $facet->{'name'}->[0],
926 # };
927 # }
928 # push @facets_loop, ( {
929 # type_label => $list,
930 # facets => \@facets,
931 # } );
932 # }
933
934 return ( undef, $results_hashref, \@facets_loop );
935 }
936
937 # TRUNCATION
938 sub _detect_truncation {
939 my ( $operand, $index ) = @_;
940 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
941 @regexpr );
942 $operand =~ s/^ //g;
943 my @wordlist = split( /\s/, $operand );
944 foreach my $word (@wordlist) {
945 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
946 push @rightlefttruncated, $word;
947 }
948 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
949 push @lefttruncated, $word;
950 }
951 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
952 push @righttruncated, $word;
953 }
954 elsif ( index( $word, "*" ) < 0 ) {
955 push @nontruncated, $word;
956 }
957 else {
958 push @regexpr, $word;
959 }
960 }
961 return (
962 \@nontruncated, \@righttruncated, \@lefttruncated,
963 \@rightlefttruncated, \@regexpr
964 );
965 }
966
967 # STEMMING
968 sub _build_stemmed_operand {
969 my ($operand,$lang) = @_;
970 require Lingua::Stem::Snowball ;
971 my $stemmed_operand=q{};
972
973 # If operand contains a digit, it is almost certainly an identifier, and should
974 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
975 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
976 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
977 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
978 return $operand if $operand =~ /\d/;
979
980 # FIXME: the locale should be set based on the user's language and/or search choice
981 #warn "$lang";
982 # Make sure we only use the first two letters from the language code
983 $lang = lc(substr($lang, 0, 2));
984 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
985 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
986 if ($lang eq 'nb' || $lang eq 'nn') {
987 $lang = 'no';
988 }
989 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
990 encoding => "UTF-8" );
991
992 my @words = split( / /, $operand );
993 my @stems = $stemmer->stem(\@words);
994 for my $stem (@stems) {
995 $stemmed_operand .= "$stem";
996 $stemmed_operand .= "?"
997 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
998 $stemmed_operand .= " ";
999 }
1000 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1001 return $stemmed_operand;
1002 }
1003
1004 # FIELD WEIGHTING
1005 sub _build_weighted_query {
1006
1007 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1008 # pretty well but could work much better if we had a smarter query parser
1009 my ( $operand, $stemmed_operand, $index ) = @_;
1010 my $stemming = C4::Context->preference("QueryStemming") || 0;
1011 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1012 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1013 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1014
1015 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1016
1017 # Keyword, or, no index specified
1018 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1019 $weighted_query .=
1020 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1021 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1022 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1023 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1024 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1025 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1026 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1027 if $fuzzy_enabled; # add fuzzy, word list
1028 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1029 if ( $stemming and $stemmed_operand )
1030 ; # add stemming, right truncation
1031 $weighted_query .= " or wrdl,r9=\"$operand\"";
1032
1033 # embedded sorting: 0 a-z; 1 z-a
1034 # $weighted_query .= ") or (sort1,aut=1";
1035 }
1036
1037 # Barcode searches should skip this process
1038 elsif ( $index eq 'bc' ) {
1039 $weighted_query .= "bc=\"$operand\"";
1040 }
1041
1042 # Authority-number searches should skip this process
1043 elsif ( $index eq 'an' ) {
1044 $weighted_query .= "an=\"$operand\"";
1045 }
1046
1047 # If the index is numeric, don't autoquote it.
1048 elsif ( $index =~ /,st-numeric$/ ) {
1049 $weighted_query .= " $index=$operand";
1050 }
1051
1052 # If the index already has more than one qualifier, wrap the operand
1053 # in quotes and pass it back (assumption is that the user knows what they
1054 # are doing and won't appreciate us mucking up their query
1055 elsif ( $index =~ ',' ) {
1056 $weighted_query .= " $index=\"$operand\"";
1057 }
1058
1059 #TODO: build better cases based on specific search indexes
1060 else {
1061 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1062 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1063 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1064 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1065 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1066 if $fuzzy_enabled; # add fuzzy, word list
1067 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1068 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1069 }
1070
1071 $weighted_query .= "))"; # close rank specification
1072 return $weighted_query;
1073 }
1074
1075 =head2 getIndexes
1076
1077 Return an array with available indexes.
1078
1079 =cut
1080
1081 sub getIndexes{
1082 my @indexes = (
1083 # biblio indexes
1084 'ab',
1085 'Abstract',
1086 'acqdate',
1087 'allrecords',
1088 'an',
1089 'Any',
1090 'at',
1091 'au',
1092 'aub',
1093 'aud',
1094 'audience',
1095 'auo',
1096 'aut',
1097 'Author',
1098 'Author-in-order ',
1099 'Author-personal-bibliography',
1100 'Authority-Number',
1101 'authtype',
1102 'bc',
1103 'Bib-level',
1104 'biblionumber',
1105 'bio',
1106 'biography',
1107 'callnum',
1108 'cfn',
1109 'Chronological-subdivision',
1110 'cn-bib-source',
1111 'cn-bib-sort',
1112 'cn-class',
1113 'cn-item',
1114 'cn-prefix',
1115 'cn-suffix',
1116 'cpn',
1117 'Code-institution',
1118 'Conference-name',
1119 'Conference-name-heading',
1120 'Conference-name-see',
1121 'Conference-name-seealso',
1122 'Content-type',
1123 'Control-number',
1124 'copydate',
1125 'Corporate-name',
1126 'Corporate-name-heading',
1127 'Corporate-name-see',
1128 'Corporate-name-seealso',
1129 'Country-publication',
1130 'ctype',
1131 'curriculum',
1132 'date-entered-on-file',
1133 'Date-of-acquisition',
1134 'Date-of-publication',
1135 'Date-time-last-modified',
1136 'Dewey-classification',
1137 'Dissertation-information',
1138 'diss',
1139 'dtlm',
1140 'EAN',
1141 'extent',
1142 'fic',
1143 'fiction',
1144 'Form-subdivision',
1145 'format',
1146 'Geographic-subdivision',
1147 'he',
1148 'Heading',
1149 'Heading-use-main-or-added-entry',
1150 'Heading-use-series-added-entry ',
1151 'Heading-use-subject-added-entry',
1152 'Host-item',
1153 'id-other',
1154 'ident',
1155 'Identifier-standard',
1156 'Illustration-code',
1157 'Index-term-genre',
1158 'Index-term-uncontrolled',
1159 'Interest-age-level',
1160 'Interest-grade-level',
1161 'ISBN',
1162 'isbn',
1163 'ISSN',
1164 'issn',
1165 'itemtype',
1166 'kw',
1167 'Koha-Auth-Number',
1168 'l-format',
1169 'language',
1170 'language-original',
1171 'lc-card',
1172 'LC-card-number',
1173 'lcn',
1174 'lex',
1175 'lexile-number',
1176 'llength',
1177 'ln',
1178 'ln-audio',
1179 'ln-subtitle',
1180 'Local-classification',
1181 'Local-number',
1182 'Match-heading',
1183 'Match-heading-see-from',
1184 'Material-type',
1185 'mc-itemtype',
1186 'mc-rtype',
1187 'mus',
1188 'name',
1189 'Music-number',
1190 'Name-geographic',
1191 'Name-geographic-heading',
1192 'Name-geographic-see',
1193 'Name-geographic-seealso',
1194 'nb',
1195 'Note',
1196 'notes',
1197 'ns',
1198 'nt',
1199 'Other-control-number',
1200 'pb',
1201 'Personal-name',
1202 'Personal-name-heading',
1203 'Personal-name-see',
1204 'Personal-name-seealso',
1205 'pl',
1206 'Place-publication',
1207 'pn',
1208 'popularity',
1209 'pubdate',
1210 'Publisher',
1211 'Provider',
1212 'pv',
1213 'Reading-grade-level',
1214 'Record-control-number',
1215 'rcn',
1216 'Record-type',
1217 'rtype',
1218 'se',
1219 'See',
1220 'See-also',
1221 'sn',
1222 'Stock-number',
1223 'su',
1224 'Subject',
1225 'Subject-heading-thesaurus',
1226 'Subject-name-personal',
1227 'Subject-subdivision',
1228 'Summary',
1229 'Suppress',
1230 'su-geo',
1231 'su-na',
1232 'su-to',
1233 'su-ut',
1234 'ut',
1235 'Term-genre-form',
1236 'Term-genre-form-heading',
1237 'Term-genre-form-see',
1238 'Term-genre-form-seealso',
1239 'ti',
1240 'Title',
1241 'Title-cover',
1242 'Title-series',
1243 'Title-uniform',
1244 'Title-uniform-heading',
1245 'Title-uniform-see',
1246 'Title-uniform-seealso',
1247 'totalissues',
1248 'yr',
1249
1250 # items indexes
1251 'acqsource',
1252 'barcode',
1253 'bc',
1254 'branch',
1255 'ccode',
1256 'classification-source',
1257 'cn-sort',
1258 'coded-location-qualifier',
1259 'copynumber',
1260 'damaged',
1261 'datelastborrowed',
1262 'datelastseen',
1263 'holdingbranch',
1264 'homebranch',
1265 'issues',
1266 'item',
1267 'itemnumber',
1268 'itype',
1269 'Local-classification',
1270 'location',
1271 'lost',
1272 'materials-specified',
1273 'mc-ccode',
1274 'mc-itype',
1275 'mc-loc',
1276 'notforloan',
1277 'Number-local-acquisition',
1278 'onloan',
1279 'price',
1280 'renewals',
1281 'replacementprice',
1282 'replacementpricedate',
1283 'reserves',
1284 'restricted',
1285 'stack',
1286 'stocknumber',
1287 'inv',
1288 'uri',
1289 'withdrawn',
1290
1291 # subject related
1292 );
1293
1294 return \@indexes;
1295 }
1296
1297 =head2 _handle_exploding_index
1298
1299 my $query = _handle_exploding_index($index, $term)
1300
1301 Callback routine to generate the search for "exploding" indexes (i.e.
1302 those indexes which are turned into multiple or-connected searches based
1303 on authority data).
1304
1305 =cut
1306
1307 sub _handle_exploding_index {
1308 my ($QParser, $filter, $params, $negate, $server) = @_;
1309 my $index = $filter;
1310 my $term = join(' ', @$params);
1311
1312 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1313
1314 my $marcflavour = C4::Context->preference('marcflavour');
1315
1316 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1317 my $wantedcodes = '';
1318 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1319 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1320 foreach my $auth (@$results) {
1321 my $record = MARC::Record->new_from_usmarc($auth);
1322 my @references = $record->field('5..');
1323 if (@references) {
1324 if ($index eq 'su-br') {
1325 $wantedcodes = 'g';
1326 } elsif ($index eq 'su-na') {
1327 $wantedcodes = 'h';
1328 } elsif ($index eq 'su-rl') {
1329 $wantedcodes = '';
1330 }
1331 foreach my $reference (@references) {
1332 my $codes = $reference->subfield($codesubfield);
1333 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1334 }
1335 }
1336 }
1337 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1338 return $query;
1339 }
1340
1341 =head2 parseQuery
1342
1343 ( $operators, $operands, $indexes, $limits,
1344 $sort_by, $scan, $lang ) =
1345 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1346
1347 Shim function to ease the transition from buildQuery to a new QueryParser.
1348 This function is called at the beginning of buildQuery, and modifies
1349 buildQuery's input. If it can handle the input, it returns a query that
1350 buildQuery will not try to parse.
1351
1352 =cut
1353
1354 sub parseQuery {
1355 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1356
1357 my @operators = $operators ? @$operators : ();
1358 my @indexes = $indexes ? @$indexes : ();
1359 my @operands = $operands ? @$operands : ();
1360 my @limits = $limits ? @$limits : ();
1361 my @sort_by = $sort_by ? @$sort_by : ();
1362
1363 my $query = $operands[0];
1364 my $index;
1365 my $term;
1366 my $query_desc;
1367
1368 my $QParser;
1369 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1370 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1371 undef $QParser if (scalar @limits > 0);
1372
1373 if ($QParser)
1374 {
1375 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1376 $query = '';
1377 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1378 next unless $operands[$ii];
1379 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1380 if ($query);
1381 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1382 $query .= $operands[$ii];
1383 }
1384 elsif ( $indexes[$ii] =~ m/su-/ ) {
1385 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1386 }
1387 else {
1388 $query .=
1389 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1390 }
1391 }
1392 foreach my $limit (@limits) {
1393 }
1394 if ( scalar(@sort_by) > 0 ) {
1395 my $modifier_re =
1396 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1397 $query =~ s/$modifier_re//g;
1398 foreach my $modifier (@sort_by) {
1399 $query .= " #$modifier";
1400 }
1401 }
1402
1403 $query_desc = $query;
1404 $query_desc =~ s/\s+/ /g;
1405 if ( C4::Context->preference("QueryWeightFields") ) {
1406 }
1407 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1408 { 'target_syntax_callback' => \&_handle_exploding_index } );
1409 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1410 { 'target_syntax_callback' => \&_handle_exploding_index } );
1411 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1412 { 'target_syntax_callback' => \&_handle_exploding_index } );
1413 $QParser->parse($query);
1414 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1415 }
1416 else {
1417 require Koha::QueryParser::Driver::PQF;
1418 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1419 s/$modifier_re//g for @operands;
1420 }
1421
1422 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1423 }
1424
1425 =head2 buildQuery
1426
1427 ( $error, $query,
1428 $simple_query, $query_cgi,
1429 $query_desc, $limit,
1430 $limit_cgi, $limit_desc,
1431 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1432
1433 Build queries and limits in CCL, CGI, Human,
1434 handle truncation, stemming, field weighting, fuzziness, etc.
1435
1436 See verbose embedded documentation.
1437
1438
1439 =cut
1440
1441 sub buildQuery {
1442 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1443
1444 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1445
1446 my $query_desc;
1447 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1448
1449 # dereference
1450 my @operators = $operators ? @$operators : ();
1451 my @indexes = $indexes ? @$indexes : ();
1452 my @operands = $operands ? @$operands : ();
1453 my @limits = $limits ? @$limits : ();
1454 my @sort_by = $sort_by ? @$sort_by : ();
1455
1456 my $stemming = C4::Context->preference("QueryStemming") || 0;
1457 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1458 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1459 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1460
1461 my $query = $operands[0];
1462 my $simple_query = $operands[0];
1463
1464 # initialize the variables we're passing back
1465 my $query_cgi;
1466 my $query_type;
1467
1468 my $limit;
1469 my $limit_cgi;
1470 my $limit_desc;
1471
1472 my $cclq = 0;
1473 my $cclindexes = getIndexes();
1474 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1475 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1476 my $dx = lc($1);
1477 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1478 }
1479 $query = "ccl=$query" if $cclq;
1480 }
1481
1482 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1483 # DIAGNOSTIC ONLY!!
1484 if ( $query =~ /^ccl=/ ) {
1485 my $q=$';
1486 # This is needed otherwise ccl= and &limit won't work together, and
1487 # this happens when selecting a subject on the opac-detail page
1488 @limits = grep {!/^$/} @limits;
1489 my $original_q = $q; # without available part
1490 unless ( grep { /^available$/ } @limits ) {
1491 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1492 $original_q = $q;
1493 }
1494 if ( @limits ) {
1495 if ( grep { /^available$/ } @limits ) {
1496 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1497 delete $limits['available'];
1498 }
1499 $q .= ' and '.join(' and ', @limits) if @limits;
1500 }
1501 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1502 }
1503 if ( $query =~ /^cql=/ ) {
1504 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1505 }
1506 if ( $query =~ /^pqf=/ ) {
1507 if ($query_desc) {
1508 $query_cgi = "q=".uri_escape_utf8($query_desc);
1509 } else {
1510 $query_desc = $';
1511 $query_cgi = "q=pqf=".uri_escape_utf8($');
1512 }
1513 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1514 }
1515
1516 # pass nested queries directly
1517 # FIXME: need better handling of some of these variables in this case
1518 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1519 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1520 # if ( $query =~ /(\(|\))/ ) {
1521 # return (
1522 # undef, $query, $simple_query, $query_cgi,
1523 # $query, $limit, $limit_cgi, $limit_desc,
1524 # 'ccl'
1525 # );
1526 # }
1527
1528 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1529 # query operands and indexes and add stemming, truncation, field weighting, etc.
1530 # Once we do so, we'll end up with a value in $query, just like if we had an
1531 # incoming $query from the user
1532 else {
1533 $query = ""
1534 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1535 my $previous_operand
1536 ; # a flag used to keep track if there was a previous query
1537 # if there was, we can apply the current operator
1538 # for every operand
1539 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1540
1541 # COMBINE OPERANDS, INDEXES AND OPERATORS
1542 if ( ($operands[$i] // '') ne '' ) {
1543 $operands[$i]=~s/^\s+//;
1544
1545 # A flag to determine whether or not to add the index to the query
1546 my $indexes_set;
1547
1548 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1549 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1550 $weight_fields = 0;
1551 $stemming = 0;
1552 } else {
1553 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1554 }
1555 my $operand = $operands[$i];
1556 my $index = $indexes[$i];
1557
1558 # Add index-specific attributes
1559
1560 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1561 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1562 #which is processed higher up in this sub. Other than that, year searches are typically
1563 #handled as limits which are not processed her either.
1564
1565 # Search ranges: Date of Publication, st-numeric
1566 if ( $index =~ /(yr|st-numeric)/ ) {
1567 #weight_fields/relevance search causes errors with date ranges
1568 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1569 #In the case of YYYY-YYYY, it will return no results
1570 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1571 }
1572
1573 # Date of Acquisition
1574 elsif ( $index =~ /acqdate/ ) {
1575 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1576 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1577 #top of the results just because they have lots of item records matching that date.
1578 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1579 #irrelevant here
1580 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1581 }
1582 # ISBN,ISSN,Standard Number, don't need special treatment
1583 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1584 (
1585 $stemming, $auto_truncation,
1586 $weight_fields, $fuzzy_enabled
1587 ) = ( 0, 0, 0, 0 );
1588
1589 if ( $index eq 'nb' ) {
1590 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1591 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1592 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1593 $indexes[$i] = $index = '';
1594 }
1595 }
1596 }
1597
1598 if(not $index){
1599 $index = 'kw';
1600 }
1601
1602 # Set default structure attribute (word list)
1603 my $struct_attr = q{};
1604 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1605 $struct_attr = ",wrdl";
1606 }
1607
1608 # Some helpful index variants
1609 my $index_plus = $index . $struct_attr . ':';
1610 my $index_plus_comma = $index . $struct_attr . ',';
1611
1612 if ($auto_truncation){
1613 unless ( $index =~ /,(st-|phr|ext)/ ) {
1614 #FIXME only valid with LTR scripts
1615 $operand=join(" ",map{
1616 (index($_,"*")>0?"$_":"$_*")
1617 }split (/\s+/,$operand));
1618 warn $operand if $DEBUG;
1619 }
1620 }
1621
1622 # Detect Truncation
1623 my $truncated_operand;
1624 my( $nontruncated, $righttruncated, $lefttruncated,
1625 $rightlefttruncated, $regexpr
1626 ) = _detect_truncation( $operand, $index );
1627 warn
1628 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1629 if $DEBUG;
1630
1631 # Apply Truncation
1632 if (
1633 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1634 scalar(@$rightlefttruncated) > 0 )
1635 {
1636
1637 # Don't field weight or add the index to the query, we do it here
1638 $indexes_set = 1;
1639 undef $weight_fields;
1640 my $previous_truncation_operand;
1641 if (scalar @$nontruncated) {
1642 $truncated_operand .= "$index_plus @$nontruncated ";
1643 $previous_truncation_operand = 1;
1644 }
1645 if (scalar @$righttruncated) {
1646 $truncated_operand .= "and " if $previous_truncation_operand;
1647 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1648 $previous_truncation_operand = 1;
1649 }
1650 if (scalar @$lefttruncated) {
1651 $truncated_operand .= "and " if $previous_truncation_operand;
1652 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1653 $previous_truncation_operand = 1;
1654 }
1655 if (scalar @$rightlefttruncated) {
1656 $truncated_operand .= "and " if $previous_truncation_operand;
1657 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1658 $previous_truncation_operand = 1;
1659 }
1660 }
1661 $operand = $truncated_operand if $truncated_operand;
1662 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1663
1664 # Handle Stemming
1665 my $stemmed_operand;
1666 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1667 if $stemming;
1668
1669 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1670
1671 # Handle Field Weighting
1672 my $weighted_operand;
1673 if ($weight_fields) {
1674 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1675 $operand = $weighted_operand;
1676 $indexes_set = 1;
1677 }
1678
1679 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1680
1681 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1682 query => $query,
1683 query_cgi => $query_cgi,
1684 query_desc => $query_desc,
1685 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1686 parsed_operand => $operand,
1687 original_operand => $operands[$i] // '',
1688 index => $index,
1689 index_plus => $index_plus,
1690 indexes_set => $indexes_set,
1691 previous_operand => $previous_operand,
1692 });
1693
1694 } #/if $operands
1695 } # /for
1696 }
1697 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1698
1699 # add limits
1700 my %group_OR_limits;
1701 my $availability_limit;
1702 foreach my $this_limit (@limits) {
1703 next unless $this_limit;
1704 if ( $this_limit =~ /available/ ) {
1705 #
1706 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1707 ## In English:
1708 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1709 $availability_limit .=
1710 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1711 $limit_cgi .= "&limit=available";
1712 $limit_desc .= "";
1713 }
1714
1715 # group_OR_limits, prefixed by mc-
1716 # OR every member of the group
1717 elsif ( $this_limit =~ /mc/ ) {
1718 my ($k,$v) = split(/:/, $this_limit,2);
1719 if ( $k !~ /mc-i(tem)?type/ ) {
1720 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1721 $this_limit =~ tr/"//d;
1722 $this_limit = $k.":'".$v."'";
1723 }
1724
1725 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1726 $limit_desc .= " or " if $group_OR_limits{$k};
1727 $group_OR_limits{$k} .= "$this_limit";
1728 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1729 $limit_desc .= " $this_limit";
1730 }
1731
1732 # Regular old limits
1733 else {
1734 $limit .= " and " if $limit || $query;
1735 $limit .= "$this_limit";
1736 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1737 if ($this_limit =~ /^branch:(.+)/) {
1738 my $branchcode = $1;
1739 my $library = Koha::Libraries->find( $branchcode );
1740 if (defined $library) {
1741 $limit_desc .= " branch:" . $library->branchname;
1742 } else {
1743 $limit_desc .= " $this_limit";
1744 }
1745 } else {
1746 $limit_desc .= " $this_limit";
1747 }
1748 }
1749 }
1750 foreach my $k (keys (%group_OR_limits)) {
1751 $limit .= " and " if ( $query || $limit );
1752 $limit .= "($group_OR_limits{$k})";
1753 }
1754 if ($availability_limit) {
1755 $limit .= " and " if ( $query || $limit );
1756 $limit .= "($availability_limit)";
1757 }
1758
1759 # Normalize the query and limit strings
1760 # This is flawed , means we can't search anything with : in it
1761 # if user wants to do ccl or cql, start the query with that
1762 # $query =~ s/:/=/g;
1763 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1764 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1765 $query =~ s/(?<=(wrdl)):/=/g;
1766 $query =~ s/(?<=(trn|phr)):/=/g;
1767 $query =~ s/(?<=(st-numeric)):/=/g;
1768 $query =~ s/(?<=(st-year)):/=/g;
1769 $query =~ s/(?<=(st-date-normalized)):/=/g;
1770 $limit =~ s/:/=/g;
1771 for ( $query, $query_desc, $limit, $limit_desc ) {
1772 s/ +/ /g; # remove extra spaces
1773 s/^ //g; # remove any beginning spaces
1774 s/ $//g; # remove any ending spaces
1775 s/==/=/g; # remove double == from query
1776 }
1777 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1778
1779 for ($query_cgi,$simple_query) {
1780 s/"//g;
1781 }
1782 # append the limit to the query
1783 $query .= " " . $limit;
1784
1785 # Warnings if DEBUG
1786 if ($DEBUG) {
1787 warn "QUERY:" . $query;
1788 warn "QUERY CGI:" . $query_cgi;
1789 warn "QUERY DESC:" . $query_desc;
1790 warn "LIMIT:" . $limit;
1791 warn "LIMIT CGI:" . $limit_cgi;
1792 warn "LIMIT DESC:" . $limit_desc;
1793 warn "---------\nLeave buildQuery\n---------";
1794 }
1795
1796 return (
1797 undef, $query, $simple_query, $query_cgi,
1798 $query_desc, $limit, $limit_cgi, $limit_desc,
1799 $query_type
1800 );
1801 }
1802
1803 =head2 _build_initial_query
1804
1805 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1806
1807 Build a section of the initial query containing indexes, operators, and operands.
1808
1809 =cut
1810
1811 sub _build_initial_query {
1812 my ($params) = @_;
1813
1814 my $operator = "";
1815 if ($params->{previous_operand}){
1816 #If there is a previous operand, add a supplied operator or the default 'and'
1817 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1818 }
1819
1820 #NOTE: indexes_set is typically set when doing truncation or field weighting
1821 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1822
1823 #e.g. "kw,wrdl:test"
1824 #e.g. " and kw,wrdl:test"
1825 $params->{query} .= $operator . $operand;
1826
1827 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1828 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1829 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1830
1831 #e.g. " and kw,wrdl: test"
1832 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1833
1834 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1835
1836 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1837 }
1838
1839 =head2 searchResults
1840
1841 my @search_results = searchResults($search_context, $searchdesc, $hits,
1842 $results_per_page, $offset, $scan,
1843 @marcresults);
1844
1845 Format results in a form suitable for passing to the template
1846
1847 =cut
1848
1849 # IMO this subroutine is pretty messy still -- it's responsible for
1850 # building the HTML output for the template
1851 sub searchResults {
1852 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1853 my $dbh = C4::Context->dbh;
1854 my @newresults;
1855
1856 require C4::Items;
1857
1858 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1859 my ($is_opac, $hidelostitems);
1860 if ($search_context eq 'opac') {
1861 $hidelostitems = C4::Context->preference('hidelostitems');
1862 $is_opac = 1;
1863 }
1864
1865 #Build branchnames hash
1866 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1867
1868 # FIXME - We build an authorised values hash here, using the default framework
1869 # though it is possible to have different authvals for different fws.
1870
1871 my $shelflocations =
1872 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1873
1874 # get notforloan authorised value list (see $shelflocations FIXME)
1875 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1876 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1877
1878 #Get itemtype hash
1879 my $itemtypes = Koha::ItemTypes->search_with_localization;
1880 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1881
1882 #search item field code
1883 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1884
1885 ## find column names of items related to MARC
1886 my %subfieldstosearch;
1887 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1888 for my $column ( @columns ) {
1889 my ( $tagfield, $tagsubfield ) =
1890 &GetMarcFromKohaField( "items." . $column, "" );
1891 if ( defined $tagsubfield ) {
1892 $subfieldstosearch{$column} = $tagsubfield;
1893 }
1894 }
1895
1896 # handle which records to actually retrieve
1897 my $times;
1898 if ( $hits && $offset + $results_per_page <= $hits ) {
1899 $times = $offset + $results_per_page;
1900 }
1901 else {
1902 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1903 }
1904
1905 my $marcflavour = C4::Context->preference("marcflavour");
1906 # We get the biblionumber position in MARC
1907 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1908
1909 # set stuff for XSLT processing here once, not later again for every record we retrieved
1910 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1911 my $xslsyspref = $interface . "XSLTResultsDisplay";
1912 my $xslfile = C4::Context->preference($xslsyspref);
1913 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1914 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1915
1916 # loop through all of the records we've retrieved
1917 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1918
1919 my $marcrecord;
1920 if ($scan) {
1921 # For Scan searches we built USMARC data
1922 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1923 } else {
1924 # Normal search, render from Zebra's output
1925 $marcrecord = new_record_from_zebra(
1926 'biblioserver',
1927 $marcresults->[$i]
1928 );
1929
1930 if ( ! defined $marcrecord ) {
1931 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1932 next;
1933 }
1934 }
1935
1936 my $fw = $scan
1937 ? undef
1938 : $bibliotag < 10
1939 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1940 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1941
1942 SetUTF8Flag($marcrecord);
1943 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1944 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1945 $oldbiblio->{result_number} = $i + 1;
1946
1947 # add imageurl to itemtype if there is one
1948 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1949
1950 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1951 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1952 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1953 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1954 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1955
1956 # edition information, if any
1957 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1958 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1959 # Build summary if there is one (the summary is defined in the itemtypes table)
1960 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1961 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1962 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1963 my @fields = $marcrecord->fields();
1964
1965 my $newsummary;
1966 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1967 my $tags = {};
1968 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1969 $tag =~ /(.{3})(.)/;
1970 if($marcrecord->field($1)){
1971 my @abc = $marcrecord->field($1)->subfield($2);
1972 $tags->{$tag} = $#abc + 1 ;
1973 }
1974 }
1975
1976 # We catch how many times to repeat this line
1977 my $max = 0;
1978 foreach my $tag (keys(%$tags)){
1979 $max = $tags->{$tag} if($tags->{$tag} > $max);
1980 }
1981
1982 # we replace, and repeat each line
1983 for (my $i = 0 ; $i < $max ; $i++){
1984 my $newline = $line;
1985
1986 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1987 $tag =~ /(.{3})(.)/;
1988
1989 if($marcrecord->field($1)){
1990 my @repl = $marcrecord->field($1)->subfield($2);
1991 my $subfieldvalue = $repl[$i];
1992 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1993 }
1994 }
1995 $newsummary .= "$newline\n";
1996 }
1997 }
1998
1999 $newsummary =~ s/\[(.*?)]//g;
2000 $newsummary =~ s/\n/<br\/>/g;
2001 $oldbiblio->{summary} = $newsummary;
2002 }
2003
2004 # Pull out the items fields
2005 my @fields = $marcrecord->field($itemtag);
2006 my $marcflavor = C4::Context->preference("marcflavour");
2007 # adding linked items that belong to host records
2008 my $analyticsfield = '773';
2009 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2010 $analyticsfield = '773';
2011 } elsif ($marcflavor eq 'UNIMARC') {
2012 $analyticsfield = '461';
2013 }
2014 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2015 my $hostbiblionumber = $hostfield->subfield("0");
2016 my $linkeditemnumber = $hostfield->subfield("9");
2017 if( $hostbiblionumber ) {
2018 my $hostbiblio = GetMarcBiblio({
2019 biblionumber => $hostbiblionumber,
2020 embed_items => 1 });
2021 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2022 if( $hostbiblio ) {
2023 my @hostitems = $hostbiblio->field($itemfield);
2024 foreach my $hostitem (@hostitems){
2025 if ($hostitem->subfield("9") eq $linkeditemnumber){
2026 my $linkeditem =$hostitem;
2027 # append linked items if they exist
2028 push @fields, $linkeditem if $linkeditem;
2029 }
2030 }
2031 }
2032 }
2033 }
2034
2035 # Setting item statuses for display
2036 my @available_items_loop;
2037 my @onloan_items_loop;
2038 my @other_items_loop;
2039
2040 my $available_items;
2041 my $onloan_items;
2042 my $other_items;
2043
2044 my $ordered_count = 0;
2045 my $available_count = 0;
2046 my $onloan_count = 0;
2047 my $longoverdue_count = 0;
2048 my $other_count = 0;
2049 my $withdrawn_count = 0;
2050 my $itemlost_count = 0;
2051 my $hideatopac_count = 0;
2052 my $itembinding_count = 0;
2053 my $itemdamaged_count = 0;
2054 my $item_in_transit_count = 0;
2055 my $can_place_holds = 0;
2056 my $item_onhold_count = 0;
2057 my $notforloan_count = 0;
2058 my $items_count = scalar(@fields);
2059 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2060 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2061 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2062
2063 # loop through every item
2064 foreach my $field (@fields) {
2065 my $item;
2066
2067 # populate the items hash
2068 foreach my $code ( keys %subfieldstosearch ) {
2069 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2070 }
2071 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2072
2073 # OPAC hidden items
2074 if ($is_opac) {
2075 # hidden because lost
2076 if ($hidelostitems && $item->{itemlost}) {
2077 $hideatopac_count++;
2078 next;
2079 }
2080 # hidden based on OpacHiddenItems syspref
2081 my @hi = C4::Items::GetHiddenItemnumbers($item);
2082 if (scalar @hi) {
2083 push @hiddenitems, @hi;
2084 $hideatopac_count++;
2085 next;
2086 }
2087 }
2088
2089 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2090 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2091
2092 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2093 if ($item->{$hbranch}) {
2094 $item->{'branchname'} = $branches{$item->{$hbranch}};
2095 }
2096 elsif ($item->{$otherbranch}) { # Last resort
2097 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2098 }
2099
2100 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2101 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2102 my $userenv = C4::Context->userenv;
2103 if ( $item->{onloan}
2104 && $userenv
2105 && $userenv->{number}
2106 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2107 {
2108 $onloan_count++;
2109 my $key = $prefix . $item->{onloan} . $item->{barcode};
2110 $onloan_items->{$key}->{due_date} = $item->{onloan};
2111 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2112 $onloan_items->{$key}->{branchname} = $item->{branchname};
2113 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2114 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2115 $onloan_items->{$key}->{description} = $item->{description};
2116 $onloan_items->{$key}->{imageurl} =
2117 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2118
2119 # if something's checked out and lost, mark it as 'long overdue'
2120 if ( $item->{itemlost} ) {
2121 $onloan_items->{$key}->{longoverdue}++;
2122 $longoverdue_count++;
2123 }
2124 else { # can place holds as long as item isn't lost
2125 $can_place_holds = 1;
2126 }
2127 }
2128
2129 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2130 else {
2131
2132 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2133
2134 # item is on order
2135 if ( $item->{notforloan} < 0 ) {
2136 $ordered_count++;
2137 } elsif ( $item->{notforloan} > 0 ) {
2138 $notforloan_count++;
2139 }
2140
2141 # is item in transit?
2142 my $transfertwhen = '';
2143 my ($transfertfrom, $transfertto);
2144
2145 # is item on the reserve shelf?
2146 my $reservestatus = '';
2147
2148 unless ($item->{withdrawn}
2149 || $item->{itemlost}
2150 || $item->{damaged}
2151 || $item->{notforloan}
2152 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2153 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2154
2155 # A couple heuristics to limit how many times
2156 # we query the database for item transfer information, sacrificing
2157 # accuracy in some cases for speed;
2158 #
2159 # 1. don't query if item has one of the other statuses
2160 # 2. don't check transit status if the bib has
2161 # more than 20 items
2162 #
2163 # FIXME: to avoid having the query the database like this, and to make
2164 # the in transit status count as unavailable for search limiting,
2165 # should map transit status to record indexed in Zebra.
2166 #
2167 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2168 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2169 }
2170
2171 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2172 if ( $item->{withdrawn}
2173 || $item->{itemlost}
2174 || $item->{damaged}
2175 || $item->{notforloan}
2176 || $reservestatus eq 'Waiting'
2177 || ($transfertwhen ne ''))
2178 {
2179 $withdrawn_count++ if $item->{withdrawn};
2180 $itemlost_count++ if $item->{itemlost};
2181 $itemdamaged_count++ if $item->{damaged};
2182 $item_in_transit_count++ if $transfertwhen ne '';
2183 $item_onhold_count++ if $reservestatus eq 'Waiting';
2184 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2185
2186 # can place a hold on a item if
2187 # not lost nor withdrawn
2188 # not damaged unless AllowHoldsOnDamagedItems is true
2189 # item is either for loan or on order (notforloan < 0)
2190 $can_place_holds = 1
2191 if (
2192 !$item->{itemlost}
2193 && !$item->{withdrawn}
2194 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2195 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2196 );
2197
2198 $other_count++;
2199
2200 my $key = $prefix . $item->{status};
2201 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2202 $other_items->{$key}->{$_} = $item->{$_};
2203 }
2204 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2205 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2206 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2207 $other_items->{$key}->{count}++ if $item->{$hbranch};
2208 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2209 $other_items->{$key}->{description} = $item->{description};
2210 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2211 }
2212 # item is available
2213 else {
2214 $can_place_holds = 1;
2215 $available_count++;
2216 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2217 foreach (qw(branchname itemcallnumber description)) {
2218 $available_items->{$prefix}->{$_} = $item->{$_};
2219 }
2220 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2221 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2222 }
2223 }
2224 } # notforloan, item level and biblioitem level
2225
2226 # if all items are hidden, do not show the record
2227 if ($items_count > 0 && $hideatopac_count == $items_count) {
2228 next;
2229 }
2230
2231 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2232 for my $key ( sort keys %$onloan_items ) {
2233 (++$onloanitemscount > $maxitems) and last;
2234 push @onloan_items_loop, $onloan_items->{$key};
2235 }
2236 for my $key ( sort keys %$other_items ) {
2237 (++$otheritemscount > $maxitems) and last;
2238 push @other_items_loop, $other_items->{$key};
2239 }
2240 for my $key ( sort keys %$available_items ) {
2241 (++$availableitemscount > $maxitems) and last;
2242 push @available_items_loop, $available_items->{$key}
2243 }
2244
2245 # XSLT processing of some stuff
2246 # we fetched the sysprefs already before the loop through all retrieved record!
2247 if (!$scan && $xslfile) {
2248 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2249 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2250 }
2251
2252 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2253 if (!C4::Context->preference("item-level_itypes")) {
2254 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2255 $can_place_holds = 0;
2256 }
2257 }
2258 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2259 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2260 $oldbiblio->{items_count} = $items_count;
2261 $oldbiblio->{available_items_loop} = \@available_items_loop;
2262 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2263 $oldbiblio->{other_items_loop} = \@other_items_loop;
2264 $oldbiblio->{availablecount} = $available_count;
2265 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2266 $oldbiblio->{onloancount} = $onloan_count;
2267 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2268 $oldbiblio->{othercount} = $other_count;
2269 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2270 $oldbiblio->{withdrawncount} = $withdrawn_count;
2271 $oldbiblio->{itemlostcount} = $itemlost_count;
2272 $oldbiblio->{damagedcount} = $itemdamaged_count;
2273 $oldbiblio->{intransitcount} = $item_in_transit_count;
2274 $oldbiblio->{onholdcount} = $item_onhold_count;
2275 $oldbiblio->{orderedcount} = $ordered_count;
2276 $oldbiblio->{notforloancount} = $notforloan_count;
2277
2278 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2279 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2280 my $subfields = substr $fieldspec, 3;
2281 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2282 my @alternateholdingsinfo = ();
2283 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2284 my $alternateholdingscount = 0;
2285
2286 for my $field (@holdingsfields) {
2287 my %holding = ( holding => '' );
2288 my $havesubfield = 0;
2289 for my $subfield ($field->subfields()) {
2290 if ((index $subfields, $$subfield[0]) >= 0) {
2291 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2292 $holding{'holding'} .= $$subfield[1];
2293 $havesubfield++;
2294 }
2295 }
2296 if ($havesubfield) {
2297 push(@alternateholdingsinfo, \%holding);
2298 $alternateholdingscount++;
2299 }
2300 }
2301
2302 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2303 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2304 }
2305
2306 push( @newresults, $oldbiblio );
2307 }
2308
2309 return @newresults;
2310 }
2311
2312 =head2 enabled_staff_search_views
2313
2314 %hash = enabled_staff_search_views()
2315
2316 This function returns a hash that contains three flags obtained from the system
2317 preferences, used to determine whether a particular staff search results view
2318 is enabled.
2319
2320 =over 2
2321
2322 =item C<Output arg:>
2323
2324 * $hash{can_view_MARC} is true only if the MARC view is enabled
2325 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2326 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2327
2328 =item C<usage in the script:>
2329
2330 =back
2331
2332 $template->param ( C4::Search::enabled_staff_search_views );
2333
2334 =cut
2335
2336 sub enabled_staff_search_views
2337 {
2338 return (
2339 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2340 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2341 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2342 );
2343 }
2344
2345 =head2 z3950_search_args
2346
2347 $arrayref = z3950_search_args($matchpoints)
2348
2349 This function returns an array reference that contains the search parameters to be
2350 passed to the Z39.50 search script (z3950_search.pl). The array elements
2351 are hash refs whose keys are name and value, and whose values are the
2352 name of a search parameter, the value of that search parameter and the URL encoded
2353 value of that parameter.
2354
2355 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2356
2357 The search parameter values are obtained from the bibliographic record whose
2358 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2359
2360 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2361 a general purpose search argument. In this case, the returned array contains only
2362 entry: the key is 'title' and the value is derived from $matchpoints.
2363
2364 If a search parameter value is undefined or empty, it is not included in the returned
2365 array.
2366
2367 The returned array reference may be passed directly to the template parameters.
2368
2369 =over 2
2370
2371 =item C<Output arg:>
2372
2373 * $array containing hash refs as described above
2374
2375 =item C<usage in the script:>
2376
2377 =back
2378
2379 $data = Biblio::GetBiblioData($bibno);
2380 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2381
2382 *OR*
2383
2384 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2385
2386 =cut
2387
2388 sub z3950_search_args {
2389 my $bibrec = shift;
2390
2391 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2392 my $isbn = Business::ISBN->new( $isbn_string );
2393
2394 if (defined $isbn && $isbn->is_valid)
2395 {
2396 if ( ref($bibrec) ) {
2397 $bibrec->{isbn} = $isbn_string;
2398 $bibrec->{title} = undef;
2399 } else {
2400 $bibrec = { isbn => $isbn_string };
2401 }
2402 }
2403 else {
2404 $bibrec = { title => $bibrec } if !ref $bibrec;
2405 }
2406 my $array = [];
2407 for my $field (qw/ lccn isbn issn title author dewey subject /)
2408 {
2409 push @$array, { name => $field, value => $bibrec->{$field} }
2410 if defined $bibrec->{$field};
2411 }
2412 return $array;
2413 }
2414
2415 =head2 GetDistinctValues($field);
2416
2417 C<$field> is a reference to the fields array
2418
2419 =cut
2420
2421 sub GetDistinctValues {
2422 my ($fieldname,$string)=@_;
2423 # returns a reference to a hash of references to branches...
2424 if ($fieldname=~/\./){
2425 my ($table,$column)=split /\./, $fieldname;
2426 my $dbh = C4::Context->dbh;
2427 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2428 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 ");
2429 $sth->execute;
2430 my $elements=$sth->fetchall_arrayref({});
2431 return $elements;
2432 }
2433 else {
2434 $string||= qq("");
2435 my @servers=qw<biblioserver authorityserver>;
2436 my (@zconns,@results);
2437 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2438 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2439 $results[$i] =
2440 $zconns[$i]->scan(
2441 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2442 );
2443 }
2444 # The big moment: asynchronously retrieve results from all servers
2445 my @elements;
2446 _ZOOM_event_loop(
2447 \@zconns,
2448 \@results,
2449 sub {
2450 my ( $i, $size ) = @_;
2451 for ( my $j = 0 ; $j < $size ; $j++ ) {
2452 my %hashscan;
2453 @hashscan{qw(value cnt)} =
2454 $results[ $i - 1 ]->display_term($j);
2455 push @elements, \%hashscan;
2456 }
2457 }
2458 );
2459 return \@elements;
2460 }
2461 }
2462
2463 =head2 _ZOOM_event_loop
2464
2465 _ZOOM_event_loop(\@zconns, \@results, sub {
2466 my ( $i, $size ) = @_;
2467 ....
2468 } );
2469
2470 Processes a ZOOM event loop and passes control to a closure for
2471 processing the results, and destroying the resultsets.
2472
2473 =cut
2474
2475 sub _ZOOM_event_loop {
2476 my ($zconns, $results, $callback) = @_;
2477 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2478 my $ev = $zconns->[ $i - 1 ]->last_event();
2479 if ( $ev == ZOOM::Event::ZEND ) {
2480 next unless $results->[ $i - 1 ];
2481 my $size = $results->[ $i - 1 ]->size();
2482 if ( $size > 0 ) {
2483 $callback->($i, $size);
2484 }
2485 }
2486 }
2487
2488 foreach my $result (@$results) {
2489 $result->destroy();
2490 }
2491 }
2492
2493 =head2 new_record_from_zebra
2494
2495 Given raw data from a searchengine result set, return a MARC::Record object
2496
2497 This helper function is needed to take into account all the involved
2498 system preferences and configuration variables to properly create the
2499 MARC::Record object.
2500
2501 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2502 data. If we are using DOM, then it has to be MARCXML.
2503
2504 If we are using elasticsearch, it'll already be a MARC::Record and this
2505 function needs a new name.
2506
2507 =cut
2508
2509 sub new_record_from_zebra {
2510
2511 my $server = shift;
2512 my $raw_data = shift;
2513 # Set the default indexing modes
2514 my $search_engine = C4::Context->preference("SearchEngine");
2515 if ($search_engine eq 'Elasticsearch') {
2516 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2517 }
2518 my $index_mode = ( $server eq 'biblioserver' )
2519 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2520 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2521
2522 my $marc_record = eval {
2523 if ( $index_mode eq 'dom' ) {
2524 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2525 } else {
2526 MARC::Record->new_from_usmarc( $raw_data );
2527 }
2528 };
2529
2530 if ($@) {
2531 return;
2532 } else {
2533 return $marc_record;
2534 }
2535
2536 }
2537
2538 END { } # module clean-up code here (global destructor)
2539
2540 1;
2541 __END__
2542
2543 =head1 AUTHOR
2544
2545 Koha Development Team <http://koha-community.org/>
2546
2547 =cut
Koha ILS