?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 20996: Fix unit tests
[koha.git] / C4 / Search.pm
blob46fafcda9d32459b8756a141642496df8b321f56
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 $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
658
659 if ( $use_zebra_facets ) {
660 $facets = _get_facets_from_zebra( $rs );
661 } else {
662 $facets = _get_facets_from_records( $rs );
663 }
664
665 return $facets;
666 }
667
668 sub _get_facets_from_records {
669
670 my $rs = shift;
671
672 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
673 my $facets_config = getFacets();
674 my $facets = {};
675 my $size = $rs->size();
676 my $jmax = $size > $facets_maxrecs
677 ? $facets_maxrecs
678 : $size;
679
680 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
681
682 my $marc_record = new_record_from_zebra (
683 'biblioserver',
684 $rs->record( $j )->raw()
685 );
686
687 if ( ! defined $marc_record ) {
688 warn "ERROR DECODING RECORD - $@: " .
689 $rs->record( $j )->raw();
690 next;
691 }
692
693 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
694 }
695
696 return $facets;
697 }
698
699 =head2 _get_facets_data_from_record
700
701 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
702
703 Internal function that extracts facets information from a MARC::Record object
704 and populates $facets_counter for using in getRecords.
705
706 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
707 facets for Zebra).
708
709 =cut
710
711 sub _get_facets_data_from_record {
712
713 my ( $marc_record, $facets, $facets_counter ) = @_;
714
715 for my $facet (@$facets) {
716
717 my @used_datas = ();
718
719 foreach my $tag ( @{ $facet->{ tags } } ) {
720
721 # tag number is the first three digits
722 my $tag_num = substr( $tag, 0, 3 );
723 # subfields are the remainder
724 my $subfield_letters = substr( $tag, 3 );
725
726 my @fields = $marc_record->field( $tag_num );
727 foreach my $field (@fields) {
728 # If $field->indicator(1) eq 'z', it means it is a 'see from'
729 # field introduced because of IncludeSeeFromInSearches, so skip it
730 next if $field->indicator(1) eq 'z';
731
732 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
733
734 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
735 push @used_datas, $data;
736 $facets_counter->{ $facet->{ idx } }->{ $data }++;
737 }
738 }
739 }
740 }
741 }
742
743 =head2 _get_facets_from_zebra
744
745 my $facets = _get_facets_from_zebra( $result_set )
746
747 Retrieves facets for a specified result set. It loops through the facets defined
748 in C4::Koha::getFacets and returns a hash with the following structure:
749
750 { facet_idx => {
751 facet_value => count
752 },
753 ...
754 }
755
756 =cut
757
758 sub _get_facets_from_zebra {
759
760 my $rs = shift;
761
762 # save current elementSetName
763 my $elementSetName = $rs->option( 'elementSetName' );
764
765 my $facets_loop = getFacets();
766 my $facets_data = {};
767 # loop through defined facets and fill the facets hashref
768 foreach my $facet ( @$facets_loop ) {
769
770 my $idx = $facet->{ idx };
771 my $sep = $facet->{ sep };
772 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
773 if ( $facet_values ) {
774 # we've actually got a result
775 $facets_data->{ $idx } = $facet_values;
776 }
777 }
778 # set elementSetName to its previous value to avoid side effects
779 $rs->option( elementSetName => $elementSetName );
780
781 return $facets_data;
782 }
783
784 =head2 _get_facet_from_result_set
785
786 my $facet_values =
787 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
788
789 Internal function that extracts facet information for a specific index ($facet_idx) and
790 returns a hash containing facet values and count:
791
792 {
793 $facet_value => $count ,
794 ...
795 }
796
797 Warning: this function has the side effect of changing the elementSetName for the result
798 set. It is a helper function for the main loop, which takes care of backing it up for
799 restoring.
800
801 =cut
802
803 sub _get_facet_from_result_set {
804
805 my $facet_idx = shift;
806 my $rs = shift;
807 my $sep = shift;
808
809 my $internal_sep = '<*>';
810 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
811
812 return if ( ! defined $facet_idx || ! defined $rs );
813 # zebra's facet element, untokenized index
814 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
815 # configure zebra results for retrieving the desired facet
816 $rs->option( elementSetName => $facet_element );
817 # get the facet record from result set
818 my $facet = $rs->record( 0 )->raw;
819 # if the facet has no restuls...
820 return if !defined $facet;
821 # TODO: benchmark DOM vs. SAX performance
822 my $facet_dom = XML::LibXML->load_xml(
823 string => ($facet)
824 );
825 my @terms = $facet_dom->getElementsByTagName('term');
826 return if ! @terms;
827
828 my $facets = {};
829 foreach my $term ( @terms ) {
830 my $facet_value = $term->textContent;
831 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
832 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
833 }
834
835 return $facets;
836 }
837
838 =head2 _get_facets_info
839
840 my $facets_info = C4::Search::_get_facets_info( $facets )
841
842 Internal function that extracts facets information and properly builds
843 the data structure needed to render facet labels.
844
845 =cut
846
847 sub _get_facets_info {
848
849 my $facets = shift;
850
851 my $facets_info = {};
852
853 for my $facet ( @$facets ) {
854 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
855 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
856 }
857
858 return $facets_info;
859 }
860
861 sub pazGetRecords {
862 my (
863 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
864 $results_per_page, $offset, $expanded_facet, $branches,
865 $query_type, $scan
866 ) = @_;
867 #NOTE: Parameter $branches is not used here !
868
869 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
870 $paz->init();
871 $paz->search($simple_query);
872 sleep 1; # FIXME: WHY?
873
874 # do results
875 my $results_hashref = {};
876 my $stats = XMLin($paz->stat);
877 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
878
879 # for a grouped search result, the number of hits
880 # is the number of groups returned; 'bib_hits' will have
881 # the total number of bibs.
882 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
883 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
884
885 HIT: foreach my $hit (@{ $results->{'hit'} }) {
886 my $recid = $hit->{recid}->[0];
887
888 my $work_title = $hit->{'md-work-title'}->[0];
889 my $work_author;
890 if (exists $hit->{'md-work-author'}) {
891 $work_author = $hit->{'md-work-author'}->[0];
892 }
893 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
894
895 my $result_group = {};
896 $result_group->{'group_label'} = $group_label;
897 $result_group->{'group_merge_key'} = $recid;
898
899 my $count = 1;
900 if (exists $hit->{count}) {
901 $count = $hit->{count}->[0];
902 }
903 $result_group->{'group_count'} = $count;
904
905 for (my $i = 0; $i < $count; $i++) {
906 # FIXME -- may need to worry about diacritics here
907 my $rec = $paz->record($recid, $i);
908 push @{ $result_group->{'RECORDS'} }, $rec;
909 }
910
911 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
912 }
913
914 # pass through facets
915 my $termlist_xml = $paz->termlist('author,subject');
916 my $terms = XMLin($termlist_xml, forcearray => 1);
917 my @facets_loop = ();
918 #die Dumper($results);
919 # foreach my $list (sort keys %{ $terms->{'list'} }) {
920 # my @facets = ();
921 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
922 # push @facets, {
923 # facet_label_value => $facet->{'name'}->[0],
924 # };
925 # }
926 # push @facets_loop, ( {
927 # type_label => $list,
928 # facets => \@facets,
929 # } );
930 # }
931
932 return ( undef, $results_hashref, \@facets_loop );
933 }
934
935 # TRUNCATION
936 sub _detect_truncation {
937 my ( $operand, $index ) = @_;
938 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
939 @regexpr );
940 $operand =~ s/^ //g;
941 my @wordlist = split( /\s/, $operand );
942 foreach my $word (@wordlist) {
943 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
944 push @rightlefttruncated, $word;
945 }
946 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
947 push @lefttruncated, $word;
948 }
949 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
950 push @righttruncated, $word;
951 }
952 elsif ( index( $word, "*" ) < 0 ) {
953 push @nontruncated, $word;
954 }
955 else {
956 push @regexpr, $word;
957 }
958 }
959 return (
960 \@nontruncated, \@righttruncated, \@lefttruncated,
961 \@rightlefttruncated, \@regexpr
962 );
963 }
964
965 # STEMMING
966 sub _build_stemmed_operand {
967 my ($operand,$lang) = @_;
968 require Lingua::Stem::Snowball ;
969 my $stemmed_operand=q{};
970
971 # Stemmer needs language
972 return $operand unless $lang;
973
974 # If operand contains a digit, it is almost certainly an identifier, and should
975 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
976 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
977 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
978 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
979 return $operand if $operand =~ /\d/;
980
981 # FIXME: the locale should be set based on the user's language and/or search choice
982 #warn "$lang";
983 # Make sure we only use the first two letters from the language code
984 $lang = lc(substr($lang, 0, 2));
985 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
986 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
987 if ($lang eq 'nb' || $lang eq 'nn') {
988 $lang = 'no';
989 }
990 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
991 encoding => "UTF-8" );
992
993 my @words = split( / /, $operand );
994 my @stems = $stemmer->stem(\@words);
995 for my $stem (@stems) {
996 $stemmed_operand .= "$stem";
997 $stemmed_operand .= "?"
998 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
999 $stemmed_operand .= " ";
1000 }
1001 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1002 return $stemmed_operand;
1003 }
1004
1005 # FIELD WEIGHTING
1006 sub _build_weighted_query {
1007
1008 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1009 # pretty well but could work much better if we had a smarter query parser
1010 my ( $operand, $stemmed_operand, $index ) = @_;
1011 my $stemming = C4::Context->preference("QueryStemming") || 0;
1012 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1013 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1014 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1015
1016 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1017
1018 # Keyword, or, no index specified
1019 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1020 $weighted_query .=
1021 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1022 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1023 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1024 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1025 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1026 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1027 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1028 if $fuzzy_enabled; # add fuzzy, word list
1029 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1030 if ( $stemming and $stemmed_operand )
1031 ; # add stemming, right truncation
1032 $weighted_query .= " or wrdl,r9=\"$operand\"";
1033
1034 # embedded sorting: 0 a-z; 1 z-a
1035 # $weighted_query .= ") or (sort1,aut=1";
1036 }
1037
1038 # Barcode searches should skip this process
1039 elsif ( $index eq 'bc' ) {
1040 $weighted_query .= "bc=\"$operand\"";
1041 }
1042
1043 # Authority-number searches should skip this process
1044 elsif ( $index eq 'an' ) {
1045 $weighted_query .= "an=\"$operand\"";
1046 }
1047
1048 # If the index is numeric, don't autoquote it.
1049 elsif ( $index =~ /,st-numeric$/ ) {
1050 $weighted_query .= " $index=$operand";
1051 }
1052
1053 # If the index already has more than one qualifier, wrap the operand
1054 # in quotes and pass it back (assumption is that the user knows what they
1055 # are doing and won't appreciate us mucking up their query
1056 elsif ( $index =~ ',' ) {
1057 $weighted_query .= " $index=\"$operand\"";
1058 }
1059
1060 #TODO: build better cases based on specific search indexes
1061 else {
1062 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1063 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1064 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1065 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1066 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1067 if $fuzzy_enabled; # add fuzzy, word list
1068 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1069 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1070 }
1071
1072 $weighted_query .= "))"; # close rank specification
1073 return $weighted_query;
1074 }
1075
1076 =head2 getIndexes
1077
1078 Return an array with available indexes.
1079
1080 =cut
1081
1082 sub getIndexes{
1083 my @indexes = (
1084 # biblio indexes
1085 'ab',
1086 'Abstract',
1087 'acqdate',
1088 'allrecords',
1089 'an',
1090 'Any',
1091 'at',
1092 'au',
1093 'aub',
1094 'aud',
1095 'audience',
1096 'auo',
1097 'aut',
1098 'Author',
1099 'Author-in-order ',
1100 'Author-personal-bibliography',
1101 'Authority-Number',
1102 'authtype',
1103 'bc',
1104 'Bib-level',
1105 'biblionumber',
1106 'bio',
1107 'biography',
1108 'callnum',
1109 'cfn',
1110 'Chronological-subdivision',
1111 'cn-bib-source',
1112 'cn-bib-sort',
1113 'cn-class',
1114 'cn-item',
1115 'cn-prefix',
1116 'cn-suffix',
1117 'cpn',
1118 'Code-institution',
1119 'Conference-name',
1120 'Conference-name-heading',
1121 'Conference-name-see',
1122 'Conference-name-seealso',
1123 'Content-type',
1124 'Control-number',
1125 'copydate',
1126 'Corporate-name',
1127 'Corporate-name-heading',
1128 'Corporate-name-see',
1129 'Corporate-name-seealso',
1130 'Country-publication',
1131 'ctype',
1132 'curriculum',
1133 'date-entered-on-file',
1134 'Date-of-acquisition',
1135 'Date-of-publication',
1136 'Date-time-last-modified',
1137 'Dewey-classification',
1138 'Dissertation-information',
1139 'diss',
1140 'dtlm',
1141 'EAN',
1142 'extent',
1143 'fic',
1144 'fiction',
1145 'Form-subdivision',
1146 'format',
1147 'Geographic-subdivision',
1148 'he',
1149 'Heading',
1150 'Heading-use-main-or-added-entry',
1151 'Heading-use-series-added-entry ',
1152 'Heading-use-subject-added-entry',
1153 'Host-item',
1154 'id-other',
1155 'ident',
1156 'Identifier-standard',
1157 'Illustration-code',
1158 'Index-term-genre',
1159 'Index-term-uncontrolled',
1160 'Interest-age-level',
1161 'Interest-grade-level',
1162 'ISBN',
1163 'isbn',
1164 'ISSN',
1165 'issn',
1166 'itemtype',
1167 'kw',
1168 'Koha-Auth-Number',
1169 'l-format',
1170 'language',
1171 'language-original',
1172 'lc-card',
1173 'LC-card-number',
1174 'lcn',
1175 'lex',
1176 'lexile-number',
1177 'llength',
1178 'ln',
1179 'ln-audio',
1180 'ln-subtitle',
1181 'Local-classification',
1182 'Local-number',
1183 'Match-heading',
1184 'Match-heading-see-from',
1185 'Material-type',
1186 'mc-itemtype',
1187 'mc-rtype',
1188 'mus',
1189 'name',
1190 'Music-number',
1191 'Name-geographic',
1192 'Name-geographic-heading',
1193 'Name-geographic-see',
1194 'Name-geographic-seealso',
1195 'nb',
1196 'Note',
1197 'notes',
1198 'ns',
1199 'nt',
1200 'Other-control-number',
1201 'pb',
1202 'Personal-name',
1203 'Personal-name-heading',
1204 'Personal-name-see',
1205 'Personal-name-seealso',
1206 'pl',
1207 'Place-publication',
1208 'pn',
1209 'popularity',
1210 'pubdate',
1211 'Publisher',
1212 'Provider',
1213 'pv',
1214 'Reading-grade-level',
1215 'Record-control-number',
1216 'rcn',
1217 'Record-type',
1218 'rtype',
1219 'se',
1220 'See',
1221 'See-also',
1222 'sn',
1223 'Stock-number',
1224 'su',
1225 'Subject',
1226 'Subject-heading-thesaurus',
1227 'Subject-name-personal',
1228 'Subject-subdivision',
1229 'Summary',
1230 'Suppress',
1231 'su-geo',
1232 'su-na',
1233 'su-to',
1234 'su-ut',
1235 'ut',
1236 'Term-genre-form',
1237 'Term-genre-form-heading',
1238 'Term-genre-form-see',
1239 'Term-genre-form-seealso',
1240 'ti',
1241 'Title',
1242 'Title-cover',
1243 'Title-series',
1244 'Title-uniform',
1245 'Title-uniform-heading',
1246 'Title-uniform-see',
1247 'Title-uniform-seealso',
1248 'totalissues',
1249 'yr',
1250
1251 # items indexes
1252 'acqsource',
1253 'barcode',
1254 'bc',
1255 'branch',
1256 'ccode',
1257 'classification-source',
1258 'cn-sort',
1259 'coded-location-qualifier',
1260 'copynumber',
1261 'damaged',
1262 'datelastborrowed',
1263 'datelastseen',
1264 'holdingbranch',
1265 'homebranch',
1266 'issues',
1267 'item',
1268 'itemnumber',
1269 'itype',
1270 'Local-classification',
1271 'location',
1272 'lost',
1273 'materials-specified',
1274 'mc-ccode',
1275 'mc-itype',
1276 'mc-loc',
1277 'notforloan',
1278 'Number-local-acquisition',
1279 'onloan',
1280 'price',
1281 'renewals',
1282 'replacementprice',
1283 'replacementpricedate',
1284 'reserves',
1285 'restricted',
1286 'stack',
1287 'stocknumber',
1288 'inv',
1289 'uri',
1290 'withdrawn',
1291
1292 # subject related
1293 );
1294
1295 return \@indexes;
1296 }
1297
1298 =head2 _handle_exploding_index
1299
1300 my $query = _handle_exploding_index($index, $term)
1301
1302 Callback routine to generate the search for "exploding" indexes (i.e.
1303 those indexes which are turned into multiple or-connected searches based
1304 on authority data).
1305
1306 =cut
1307
1308 sub _handle_exploding_index {
1309 my ($QParser, $filter, $params, $negate, $server) = @_;
1310 my $index = $filter;
1311 my $term = join(' ', @$params);
1312
1313 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1314
1315 my $marcflavour = C4::Context->preference('marcflavour');
1316
1317 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1318 my $wantedcodes = '';
1319 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1320 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1321 foreach my $auth (@$results) {
1322 my $record = MARC::Record->new_from_usmarc($auth);
1323 my @references = $record->field('5..');
1324 if (@references) {
1325 if ($index eq 'su-br') {
1326 $wantedcodes = 'g';
1327 } elsif ($index eq 'su-na') {
1328 $wantedcodes = 'h';
1329 } elsif ($index eq 'su-rl') {
1330 $wantedcodes = '';
1331 }
1332 foreach my $reference (@references) {
1333 my $codes = $reference->subfield($codesubfield);
1334 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1335 }
1336 }
1337 }
1338 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1339 return $query;
1340 }
1341
1342 =head2 parseQuery
1343
1344 ( $operators, $operands, $indexes, $limits,
1345 $sort_by, $scan, $lang ) =
1346 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1347
1348 Shim function to ease the transition from buildQuery to a new QueryParser.
1349 This function is called at the beginning of buildQuery, and modifies
1350 buildQuery's input. If it can handle the input, it returns a query that
1351 buildQuery will not try to parse.
1352
1353 =cut
1354
1355 sub parseQuery {
1356 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1357
1358 my @operators = $operators ? @$operators : ();
1359 my @indexes = $indexes ? @$indexes : ();
1360 my @operands = $operands ? @$operands : ();
1361 my @limits = $limits ? @$limits : ();
1362 my @sort_by = $sort_by ? @$sort_by : ();
1363
1364 my $query = $operands[0];
1365 my $index;
1366 my $term;
1367 my $query_desc;
1368
1369 my $QParser;
1370 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1371 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1372 undef $QParser if (scalar @limits > 0);
1373
1374 if ($QParser)
1375 {
1376 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1377 $query = '';
1378 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1379 next unless $operands[$ii];
1380 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1381 if ($query);
1382 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1383 $query .= $operands[$ii];
1384 }
1385 elsif ( $indexes[$ii] =~ m/su-/ ) {
1386 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1387 }
1388 else {
1389 $query .=
1390 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1391 }
1392 }
1393 foreach my $limit (@limits) {
1394 }
1395 if ( scalar(@sort_by) > 0 ) {
1396 my $modifier_re =
1397 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1398 $query =~ s/$modifier_re//g;
1399 foreach my $modifier (@sort_by) {
1400 $query .= " #$modifier";
1401 }
1402 }
1403
1404 $query_desc = $query;
1405 $query_desc =~ s/\s+/ /g;
1406 if ( C4::Context->preference("QueryWeightFields") ) {
1407 }
1408 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1409 { 'target_syntax_callback' => \&_handle_exploding_index } );
1410 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1411 { 'target_syntax_callback' => \&_handle_exploding_index } );
1412 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1413 { 'target_syntax_callback' => \&_handle_exploding_index } );
1414 $QParser->parse($query);
1415 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1416 }
1417 else {
1418 require Koha::QueryParser::Driver::PQF;
1419 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1420 s/$modifier_re//g for @operands;
1421 }
1422
1423 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1424 }
1425
1426 =head2 buildQuery
1427
1428 ( $error, $query,
1429 $simple_query, $query_cgi,
1430 $query_desc, $limit,
1431 $limit_cgi, $limit_desc,
1432 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1433
1434 Build queries and limits in CCL, CGI, Human,
1435 handle truncation, stemming, field weighting, fuzziness, etc.
1436
1437 See verbose embedded documentation.
1438
1439
1440 =cut
1441
1442 sub buildQuery {
1443 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1444
1445 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1446
1447 my $query_desc;
1448 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1449
1450 # dereference
1451 my @operators = $operators ? @$operators : ();
1452 my @indexes = $indexes ? @$indexes : ();
1453 my @operands = $operands ? @$operands : ();
1454 my @limits = $limits ? @$limits : ();
1455 my @sort_by = $sort_by ? @$sort_by : ();
1456
1457 my $stemming = C4::Context->preference("QueryStemming") || 0;
1458 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1459 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1460 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1461
1462 my $query = $operands[0];
1463 my $simple_query = $operands[0];
1464
1465 # initialize the variables we're passing back
1466 my $query_cgi;
1467 my $query_type;
1468
1469 my $limit;
1470 my $limit_cgi;
1471 my $limit_desc;
1472
1473 my $cclq = 0;
1474 my $cclindexes = getIndexes();
1475 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1476 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1477 my $dx = lc($1);
1478 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1479 }
1480 $query = "ccl=$query" if $cclq;
1481 }
1482
1483 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1484 # DIAGNOSTIC ONLY!!
1485 if ( $query =~ /^ccl=/ ) {
1486 my $q=$';
1487 # This is needed otherwise ccl= and &limit won't work together, and
1488 # this happens when selecting a subject on the opac-detail page
1489 @limits = grep {!/^$/} @limits;
1490 my $original_q = $q; # without available part
1491 unless ( grep { /^available$/ } @limits ) {
1492 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1493 $original_q = $q;
1494 }
1495 if ( @limits ) {
1496 if ( grep { /^available$/ } @limits ) {
1497 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1498 delete $limits['available'];
1499 }
1500 $q .= ' and '.join(' and ', @limits) if @limits;
1501 }
1502 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1503 }
1504 if ( $query =~ /^cql=/ ) {
1505 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1506 }
1507 if ( $query =~ /^pqf=/ ) {
1508 if ($query_desc) {
1509 $query_cgi = "q=".uri_escape_utf8($query_desc);
1510 } else {
1511 $query_desc = $';
1512 $query_cgi = "q=pqf=".uri_escape_utf8($');
1513 }
1514 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1515 }
1516
1517 # pass nested queries directly
1518 # FIXME: need better handling of some of these variables in this case
1519 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1520 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1521 # if ( $query =~ /(\(|\))/ ) {
1522 # return (
1523 # undef, $query, $simple_query, $query_cgi,
1524 # $query, $limit, $limit_cgi, $limit_desc,
1525 # 'ccl'
1526 # );
1527 # }
1528
1529 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1530 # query operands and indexes and add stemming, truncation, field weighting, etc.
1531 # Once we do so, we'll end up with a value in $query, just like if we had an
1532 # incoming $query from the user
1533 else {
1534 $query = ""
1535 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1536 my $previous_operand
1537 ; # a flag used to keep track if there was a previous query
1538 # if there was, we can apply the current operator
1539 # for every operand
1540 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1541
1542 # COMBINE OPERANDS, INDEXES AND OPERATORS
1543 if ( ($operands[$i] // '') ne '' ) {
1544 $operands[$i]=~s/^\s+//;
1545
1546 # A flag to determine whether or not to add the index to the query
1547 my $indexes_set;
1548
1549 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1550 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1551 $weight_fields = 0;
1552 $stemming = 0;
1553 } else {
1554 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1555 }
1556 my $operand = $operands[$i];
1557 my $index = $indexes[$i];
1558
1559 # Add index-specific attributes
1560
1561 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1562 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1563 #which is processed higher up in this sub. Other than that, year searches are typically
1564 #handled as limits which are not processed her either.
1565
1566 # Search ranges: Date of Publication, st-numeric
1567 if ( $index =~ /(yr|st-numeric)/ ) {
1568 #weight_fields/relevance search causes errors with date ranges
1569 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1570 #In the case of YYYY-YYYY, it will return no results
1571 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1572 }
1573
1574 # Date of Acquisition
1575 elsif ( $index =~ /acqdate/ ) {
1576 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1577 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1578 #top of the results just because they have lots of item records matching that date.
1579 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1580 #irrelevant here
1581 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1582 }
1583 # ISBN,ISSN,Standard Number, don't need special treatment
1584 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1585 (
1586 $stemming, $auto_truncation,
1587 $weight_fields, $fuzzy_enabled
1588 ) = ( 0, 0, 0, 0 );
1589
1590 if ( $index eq 'nb' ) {
1591 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1592 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1593 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1594 $indexes[$i] = $index = '';
1595 }
1596 }
1597 }
1598
1599 if(not $index){
1600 $index = 'kw';
1601 }
1602
1603 # Set default structure attribute (word list)
1604 my $struct_attr = q{};
1605 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1606 $struct_attr = ",wrdl";
1607 }
1608
1609 # Some helpful index variants
1610 my $index_plus = $index . $struct_attr . ':';
1611 my $index_plus_comma = $index . $struct_attr . ',';
1612
1613 if ($auto_truncation){
1614 unless ( $index =~ /,(st-|phr|ext)/ ) {
1615 #FIXME only valid with LTR scripts
1616 $operand=join(" ",map{
1617 (index($_,"*")>0?"$_":"$_*")
1618 }split (/\s+/,$operand));
1619 warn $operand if $DEBUG;
1620 }
1621 }
1622
1623 # Detect Truncation
1624 my $truncated_operand;
1625 my( $nontruncated, $righttruncated, $lefttruncated,
1626 $rightlefttruncated, $regexpr
1627 ) = _detect_truncation( $operand, $index );
1628 warn
1629 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1630 if $DEBUG;
1631
1632 # Apply Truncation
1633 if (
1634 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1635 scalar(@$rightlefttruncated) > 0 )
1636 {
1637
1638 # Don't field weight or add the index to the query, we do it here
1639 $indexes_set = 1;
1640 undef $weight_fields;
1641 my $previous_truncation_operand;
1642 if (scalar @$nontruncated) {
1643 $truncated_operand .= "$index_plus @$nontruncated ";
1644 $previous_truncation_operand = 1;
1645 }
1646 if (scalar @$righttruncated) {
1647 $truncated_operand .= "and " if $previous_truncation_operand;
1648 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1649 $previous_truncation_operand = 1;
1650 }
1651 if (scalar @$lefttruncated) {
1652 $truncated_operand .= "and " if $previous_truncation_operand;
1653 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1654 $previous_truncation_operand = 1;
1655 }
1656 if (scalar @$rightlefttruncated) {
1657 $truncated_operand .= "and " if $previous_truncation_operand;
1658 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1659 $previous_truncation_operand = 1;
1660 }
1661 }
1662 $operand = $truncated_operand if $truncated_operand;
1663 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1664
1665 # Handle Stemming
1666 my $stemmed_operand;
1667 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1668 if $stemming;
1669
1670 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1671
1672 # Handle Field Weighting
1673 my $weighted_operand;
1674 if ($weight_fields) {
1675 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1676 $operand = $weighted_operand;
1677 $indexes_set = 1;
1678 }
1679
1680 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1681
1682 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1683 query => $query,
1684 query_cgi => $query_cgi,
1685 query_desc => $query_desc,
1686 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1687 parsed_operand => $operand,
1688 original_operand => $operands[$i] // '',
1689 index => $index,
1690 index_plus => $index_plus,
1691 indexes_set => $indexes_set,
1692 previous_operand => $previous_operand,
1693 });
1694
1695 } #/if $operands
1696 } # /for
1697 }
1698 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1699
1700 # add limits
1701 my %group_OR_limits;
1702 my $availability_limit;
1703 foreach my $this_limit (@limits) {
1704 next unless $this_limit;
1705 if ( $this_limit =~ /available/ ) {
1706 #
1707 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1708 ## In English:
1709 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1710 $availability_limit .=
1711 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1712 $limit_cgi .= "&limit=available";
1713 $limit_desc .= "";
1714 }
1715
1716 # group_OR_limits, prefixed by mc-
1717 # OR every member of the group
1718 elsif ( $this_limit =~ /mc/ ) {
1719 my ($k,$v) = split(/:/, $this_limit,2);
1720 if ( $k !~ /mc-i(tem)?type/ ) {
1721 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1722 $this_limit =~ tr/"//d;
1723 $this_limit = $k.":'".$v."'";
1724 }
1725
1726 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1727 $limit_desc .= " or " if $group_OR_limits{$k};
1728 $group_OR_limits{$k} .= "$this_limit";
1729 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1730 $limit_desc .= " $this_limit";
1731 }
1732
1733 # Regular old limits
1734 else {
1735 $limit .= " and " if $limit || $query;
1736 $limit .= "$this_limit";
1737 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1738 if ($this_limit =~ /^branch:(.+)/) {
1739 my $branchcode = $1;
1740 my $library = Koha::Libraries->find( $branchcode );
1741 if (defined $library) {
1742 $limit_desc .= " branch:" . $library->branchname;
1743 } else {
1744 $limit_desc .= " $this_limit";
1745 }
1746 } else {
1747 $limit_desc .= " $this_limit";
1748 }
1749 }
1750 }
1751 foreach my $k (keys (%group_OR_limits)) {
1752 $limit .= " and " if ( $query || $limit );
1753 $limit .= "($group_OR_limits{$k})";
1754 }
1755 if ($availability_limit) {
1756 $limit .= " and " if ( $query || $limit );
1757 $limit .= "($availability_limit)";
1758 }
1759
1760 # Normalize the query and limit strings
1761 # This is flawed , means we can't search anything with : in it
1762 # if user wants to do ccl or cql, start the query with that
1763 # $query =~ s/:/=/g;
1764 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1765 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1766 $query =~ s/(?<=(wrdl)):/=/g;
1767 $query =~ s/(?<=(trn|phr)):/=/g;
1768 $query =~ s/(?<=(st-numeric)):/=/g;
1769 $query =~ s/(?<=(st-year)):/=/g;
1770 $query =~ s/(?<=(st-date-normalized)):/=/g;
1771 $limit =~ s/:/=/g;
1772 for ( $query, $query_desc, $limit, $limit_desc ) {
1773 s/ +/ /g; # remove extra spaces
1774 s/^ //g; # remove any beginning spaces
1775 s/ $//g; # remove any ending spaces
1776 s/==/=/g; # remove double == from query
1777 }
1778 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1779
1780 for ($query_cgi,$simple_query) {
1781 s/"//g;
1782 }
1783 # append the limit to the query
1784 $query .= " " . $limit;
1785
1786 # Warnings if DEBUG
1787 if ($DEBUG) {
1788 warn "QUERY:" . $query;
1789 warn "QUERY CGI:" . $query_cgi;
1790 warn "QUERY DESC:" . $query_desc;
1791 warn "LIMIT:" . $limit;
1792 warn "LIMIT CGI:" . $limit_cgi;
1793 warn "LIMIT DESC:" . $limit_desc;
1794 warn "---------\nLeave buildQuery\n---------";
1795 }
1796
1797 return (
1798 undef, $query, $simple_query, $query_cgi,
1799 $query_desc, $limit, $limit_cgi, $limit_desc,
1800 $query_type
1801 );
1802 }
1803
1804 =head2 _build_initial_query
1805
1806 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1807
1808 Build a section of the initial query containing indexes, operators, and operands.
1809
1810 =cut
1811
1812 sub _build_initial_query {
1813 my ($params) = @_;
1814
1815 my $operator = "";
1816 if ($params->{previous_operand}){
1817 #If there is a previous operand, add a supplied operator or the default 'and'
1818 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1819 }
1820
1821 #NOTE: indexes_set is typically set when doing truncation or field weighting
1822 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1823
1824 #e.g. "kw,wrdl:test"
1825 #e.g. " and kw,wrdl:test"
1826 $params->{query} .= $operator . $operand;
1827
1828 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1829 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1830 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1831
1832 #e.g. " and kw,wrdl: test"
1833 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1834
1835 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1836
1837 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1838 }
1839
1840 =head2 searchResults
1841
1842 my @search_results = searchResults($search_context, $searchdesc, $hits,
1843 $results_per_page, $offset, $scan,
1844 @marcresults);
1845
1846 Format results in a form suitable for passing to the template
1847
1848 =cut
1849
1850 # IMO this subroutine is pretty messy still -- it's responsible for
1851 # building the HTML output for the template
1852 sub searchResults {
1853 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1854 my $dbh = C4::Context->dbh;
1855 my @newresults;
1856
1857 require C4::Items;
1858
1859 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1860 my ($is_opac, $hidelostitems);
1861 if ($search_context eq 'opac') {
1862 $hidelostitems = C4::Context->preference('hidelostitems');
1863 $is_opac = 1;
1864 }
1865
1866 #Build branchnames hash
1867 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1868
1869 # FIXME - We build an authorised values hash here, using the default framework
1870 # though it is possible to have different authvals for different fws.
1871
1872 my $shelflocations =
1873 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1874
1875 # get notforloan authorised value list (see $shelflocations FIXME)
1876 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1877 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1878
1879 #Get itemtype hash
1880 my $itemtypes = Koha::ItemTypes->search_with_localization;
1881 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1882
1883 #search item field code
1884 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1885
1886 ## find column names of items related to MARC
1887 my %subfieldstosearch;
1888 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1889 for my $column ( @columns ) {
1890 my ( $tagfield, $tagsubfield ) =
1891 &GetMarcFromKohaField( "items." . $column, "" );
1892 if ( defined $tagsubfield ) {
1893 $subfieldstosearch{$column} = $tagsubfield;
1894 }
1895 }
1896
1897 # handle which records to actually retrieve
1898 my $times;
1899 if ( $hits && $offset + $results_per_page <= $hits ) {
1900 $times = $offset + $results_per_page;
1901 }
1902 else {
1903 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1904 }
1905
1906 my $marcflavour = C4::Context->preference("marcflavour");
1907 # We get the biblionumber position in MARC
1908 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1909
1910 # set stuff for XSLT processing here once, not later again for every record we retrieved
1911 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1912 my $xslsyspref = $interface . "XSLTResultsDisplay";
1913 my $xslfile = C4::Context->preference($xslsyspref);
1914 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1915 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1916
1917 # loop through all of the records we've retrieved
1918 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1919
1920 my $marcrecord;
1921 if ($scan) {
1922 # For Scan searches we built USMARC data
1923 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1924 } else {
1925 # Normal search, render from Zebra's output
1926 $marcrecord = new_record_from_zebra(
1927 'biblioserver',
1928 $marcresults->[$i]
1929 );
1930
1931 if ( ! defined $marcrecord ) {
1932 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1933 next;
1934 }
1935 }
1936
1937 my $fw = $scan
1938 ? undef
1939 : $bibliotag < 10
1940 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1941 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1942
1943 SetUTF8Flag($marcrecord);
1944 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1945 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1946 $oldbiblio->{result_number} = $i + 1;
1947
1948 # add imageurl to itemtype if there is one
1949 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1950
1951 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1952 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1953 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1954 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1955 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1956
1957 # edition information, if any
1958 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1959 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1960 # Build summary if there is one (the summary is defined in the itemtypes table)
1961 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1962 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1963 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1964 my @fields = $marcrecord->fields();
1965
1966 my $newsummary;
1967 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1968 my $tags = {};
1969 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1970 $tag =~ /(.{3})(.)/;
1971 if($marcrecord->field($1)){
1972 my @abc = $marcrecord->field($1)->subfield($2);
1973 $tags->{$tag} = $#abc + 1 ;
1974 }
1975 }
1976
1977 # We catch how many times to repeat this line
1978 my $max = 0;
1979 foreach my $tag (keys(%$tags)){
1980 $max = $tags->{$tag} if($tags->{$tag} > $max);
1981 }
1982
1983 # we replace, and repeat each line
1984 for (my $i = 0 ; $i < $max ; $i++){
1985 my $newline = $line;
1986
1987 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1988 $tag =~ /(.{3})(.)/;
1989
1990 if($marcrecord->field($1)){
1991 my @repl = $marcrecord->field($1)->subfield($2);
1992 my $subfieldvalue = $repl[$i];
1993 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1994 }
1995 }
1996 $newsummary .= "$newline\n";
1997 }
1998 }
1999
2000 $newsummary =~ s/\[(.*?)]//g;
2001 $newsummary =~ s/\n/<br\/>/g;
2002 $oldbiblio->{summary} = $newsummary;
2003 }
2004
2005 # Pull out the items fields
2006 my @fields = $marcrecord->field($itemtag);
2007 my $marcflavor = C4::Context->preference("marcflavour");
2008 # adding linked items that belong to host records
2009 my $analyticsfield = '773';
2010 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2011 $analyticsfield = '773';
2012 } elsif ($marcflavor eq 'UNIMARC') {
2013 $analyticsfield = '461';
2014 }
2015 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2016 my $hostbiblionumber = $hostfield->subfield("0");
2017 my $linkeditemnumber = $hostfield->subfield("9");
2018 if( $hostbiblionumber ) {
2019 my $hostbiblio = GetMarcBiblio({
2020 biblionumber => $hostbiblionumber,
2021 embed_items => 1 });
2022 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2023 if( $hostbiblio ) {
2024 my @hostitems = $hostbiblio->field($itemfield);
2025 foreach my $hostitem (@hostitems){
2026 if ($hostitem->subfield("9") eq $linkeditemnumber){
2027 my $linkeditem =$hostitem;
2028 # append linked items if they exist
2029 push @fields, $linkeditem if $linkeditem;
2030 }
2031 }
2032 }
2033 }
2034 }
2035
2036 # Setting item statuses for display
2037 my @available_items_loop;
2038 my @onloan_items_loop;
2039 my @other_items_loop;
2040
2041 my $available_items;
2042 my $onloan_items;
2043 my $other_items;
2044
2045 my $ordered_count = 0;
2046 my $available_count = 0;
2047 my $onloan_count = 0;
2048 my $longoverdue_count = 0;
2049 my $other_count = 0;
2050 my $withdrawn_count = 0;
2051 my $itemlost_count = 0;
2052 my $hideatopac_count = 0;
2053 my $itembinding_count = 0;
2054 my $itemdamaged_count = 0;
2055 my $item_in_transit_count = 0;
2056 my $can_place_holds = 0;
2057 my $item_onhold_count = 0;
2058 my $notforloan_count = 0;
2059 my $items_count = scalar(@fields);
2060 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2061 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2062 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2063
2064 # loop through every item
2065 foreach my $field (@fields) {
2066 my $item;
2067
2068 # populate the items hash
2069 foreach my $code ( keys %subfieldstosearch ) {
2070 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2071 }
2072 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2073
2074 # OPAC hidden items
2075 if ($is_opac) {
2076 # hidden because lost
2077 if ($hidelostitems && $item->{itemlost}) {
2078 $hideatopac_count++;
2079 next;
2080 }
2081 # hidden based on OpacHiddenItems syspref
2082 my @hi = C4::Items::GetHiddenItemnumbers($item);
2083 if (scalar @hi) {
2084 push @hiddenitems, @hi;
2085 $hideatopac_count++;
2086 next;
2087 }
2088 }
2089
2090 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2091 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2092
2093 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2094 if ($item->{$hbranch}) {
2095 $item->{'branchname'} = $branches{$item->{$hbranch}};
2096 }
2097 elsif ($item->{$otherbranch}) { # Last resort
2098 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2099 }
2100
2101 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2102 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2103 my $userenv = C4::Context->userenv;
2104 if ( $item->{onloan}
2105 && $userenv
2106 && $userenv->{number}
2107 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2108 {
2109 $onloan_count++;
2110 my $key = $prefix . $item->{onloan} . $item->{barcode};
2111 $onloan_items->{$key}->{due_date} = $item->{onloan};
2112 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2113 $onloan_items->{$key}->{branchname} = $item->{branchname};
2114 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2115 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2116 $onloan_items->{$key}->{description} = $item->{description};
2117 $onloan_items->{$key}->{imageurl} =
2118 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2119
2120 # if something's checked out and lost, mark it as 'long overdue'
2121 if ( $item->{itemlost} ) {
2122 $onloan_items->{$key}->{longoverdue}++;
2123 $longoverdue_count++;
2124 }
2125 else { # can place holds as long as item isn't lost
2126 $can_place_holds = 1;
2127 }
2128 }
2129
2130 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2131 else {
2132
2133 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2134
2135 # item is on order
2136 if ( $item->{notforloan} < 0 ) {
2137 $ordered_count++;
2138 } elsif ( $item->{notforloan} > 0 ) {
2139 $notforloan_count++;
2140 }
2141
2142 # is item in transit?
2143 my $transfertwhen = '';
2144 my ($transfertfrom, $transfertto);
2145
2146 # is item on the reserve shelf?
2147 my $reservestatus = '';
2148
2149 unless ($item->{withdrawn}
2150 || $item->{itemlost}
2151 || $item->{damaged}
2152 || $item->{notforloan}
2153 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2154 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2155
2156 # A couple heuristics to limit how many times
2157 # we query the database for item transfer information, sacrificing
2158 # accuracy in some cases for speed;
2159 #
2160 # 1. don't query if item has one of the other statuses
2161 # 2. don't check transit status if the bib has
2162 # more than 20 items
2163 #
2164 # FIXME: to avoid having the query the database like this, and to make
2165 # the in transit status count as unavailable for search limiting,
2166 # should map transit status to record indexed in Zebra.
2167 #
2168 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2169 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2170 }
2171
2172 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2173 if ( $item->{withdrawn}
2174 || $item->{itemlost}
2175 || $item->{damaged}
2176 || $item->{notforloan}
2177 || $reservestatus eq 'Waiting'
2178 || ($transfertwhen ne ''))
2179 {
2180 $withdrawn_count++ if $item->{withdrawn};
2181 $itemlost_count++ if $item->{itemlost};
2182 $itemdamaged_count++ if $item->{damaged};
2183 $item_in_transit_count++ if $transfertwhen ne '';
2184 $item_onhold_count++ if $reservestatus eq 'Waiting';
2185 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2186
2187 # can place a hold on a item if
2188 # not lost nor withdrawn
2189 # not damaged unless AllowHoldsOnDamagedItems is true
2190 # item is either for loan or on order (notforloan < 0)
2191 $can_place_holds = 1
2192 if (
2193 !$item->{itemlost}
2194 && !$item->{withdrawn}
2195 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2196 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2197 );
2198
2199 $other_count++;
2200
2201 my $key = $prefix . $item->{status};
2202 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2203 $other_items->{$key}->{$_} = $item->{$_};
2204 }
2205 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2206 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2207 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2208 $other_items->{$key}->{count}++ if $item->{$hbranch};
2209 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2210 $other_items->{$key}->{description} = $item->{description};
2211 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2212 }
2213 # item is available
2214 else {
2215 $can_place_holds = 1;
2216 $available_count++;
2217 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2218 foreach (qw(branchname itemcallnumber description)) {
2219 $available_items->{$prefix}->{$_} = $item->{$_};
2220 }
2221 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2222 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2223 }
2224 }
2225 } # notforloan, item level and biblioitem level
2226
2227 # if all items are hidden, do not show the record
2228 if ($items_count > 0 && $hideatopac_count == $items_count) {
2229 next;
2230 }
2231
2232 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2233 for my $key ( sort keys %$onloan_items ) {
2234 (++$onloanitemscount > $maxitems) and last;
2235 push @onloan_items_loop, $onloan_items->{$key};
2236 }
2237 for my $key ( sort keys %$other_items ) {
2238 (++$otheritemscount > $maxitems) and last;
2239 push @other_items_loop, $other_items->{$key};
2240 }
2241 for my $key ( sort keys %$available_items ) {
2242 (++$availableitemscount > $maxitems) and last;
2243 push @available_items_loop, $available_items->{$key}
2244 }
2245
2246 # XSLT processing of some stuff
2247 # we fetched the sysprefs already before the loop through all retrieved record!
2248 if (!$scan && $xslfile) {
2249 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2250 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2251 }
2252
2253 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2254 if (!C4::Context->preference("item-level_itypes")) {
2255 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2256 $can_place_holds = 0;
2257 }
2258 }
2259 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2260 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2261 $oldbiblio->{items_count} = $items_count;
2262 $oldbiblio->{available_items_loop} = \@available_items_loop;
2263 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2264 $oldbiblio->{other_items_loop} = \@other_items_loop;
2265 $oldbiblio->{availablecount} = $available_count;
2266 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2267 $oldbiblio->{onloancount} = $onloan_count;
2268 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2269 $oldbiblio->{othercount} = $other_count;
2270 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2271 $oldbiblio->{withdrawncount} = $withdrawn_count;
2272 $oldbiblio->{itemlostcount} = $itemlost_count;
2273 $oldbiblio->{damagedcount} = $itemdamaged_count;
2274 $oldbiblio->{intransitcount} = $item_in_transit_count;
2275 $oldbiblio->{onholdcount} = $item_onhold_count;
2276 $oldbiblio->{orderedcount} = $ordered_count;
2277 $oldbiblio->{notforloancount} = $notforloan_count;
2278
2279 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2280 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2281 my $subfields = substr $fieldspec, 3;
2282 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2283 my @alternateholdingsinfo = ();
2284 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2285 my $alternateholdingscount = 0;
2286
2287 for my $field (@holdingsfields) {
2288 my %holding = ( holding => '' );
2289 my $havesubfield = 0;
2290 for my $subfield ($field->subfields()) {
2291 if ((index $subfields, $$subfield[0]) >= 0) {
2292 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2293 $holding{'holding'} .= $$subfield[1];
2294 $havesubfield++;
2295 }
2296 }
2297 if ($havesubfield) {
2298 push(@alternateholdingsinfo, \%holding);
2299 $alternateholdingscount++;
2300 }
2301 }
2302
2303 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2304 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2305 }
2306
2307 push( @newresults, $oldbiblio );
2308 }
2309
2310 return @newresults;
2311 }
2312
2313 =head2 enabled_staff_search_views
2314
2315 %hash = enabled_staff_search_views()
2316
2317 This function returns a hash that contains three flags obtained from the system
2318 preferences, used to determine whether a particular staff search results view
2319 is enabled.
2320
2321 =over 2
2322
2323 =item C<Output arg:>
2324
2325 * $hash{can_view_MARC} is true only if the MARC view is enabled
2326 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2327 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2328
2329 =item C<usage in the script:>
2330
2331 =back
2332
2333 $template->param ( C4::Search::enabled_staff_search_views );
2334
2335 =cut
2336
2337 sub enabled_staff_search_views
2338 {
2339 return (
2340 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2341 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2342 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2343 );
2344 }
2345
2346 =head2 z3950_search_args
2347
2348 $arrayref = z3950_search_args($matchpoints)
2349
2350 This function returns an array reference that contains the search parameters to be
2351 passed to the Z39.50 search script (z3950_search.pl). The array elements
2352 are hash refs whose keys are name and value, and whose values are the
2353 name of a search parameter, the value of that search parameter and the URL encoded
2354 value of that parameter.
2355
2356 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2357
2358 The search parameter values are obtained from the bibliographic record whose
2359 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2360
2361 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2362 a general purpose search argument. In this case, the returned array contains only
2363 entry: the key is 'title' and the value is derived from $matchpoints.
2364
2365 If a search parameter value is undefined or empty, it is not included in the returned
2366 array.
2367
2368 The returned array reference may be passed directly to the template parameters.
2369
2370 =over 2
2371
2372 =item C<Output arg:>
2373
2374 * $array containing hash refs as described above
2375
2376 =item C<usage in the script:>
2377
2378 =back
2379
2380 $data = Biblio::GetBiblioData($bibno);
2381 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2382
2383 *OR*
2384
2385 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2386
2387 =cut
2388
2389 sub z3950_search_args {
2390 my $bibrec = shift;
2391
2392 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2393 my $isbn = Business::ISBN->new( $isbn_string );
2394
2395 if (defined $isbn && $isbn->is_valid)
2396 {
2397 if ( ref($bibrec) ) {
2398 $bibrec->{isbn} = $isbn_string;
2399 $bibrec->{title} = undef;
2400 } else {
2401 $bibrec = { isbn => $isbn_string };
2402 }
2403 }
2404 else {
2405 $bibrec = { title => $bibrec } if !ref $bibrec;
2406 }
2407 my $array = [];
2408 for my $field (qw/ lccn isbn issn title author dewey subject /)
2409 {
2410 push @$array, { name => $field, value => $bibrec->{$field} }
2411 if defined $bibrec->{$field};
2412 }
2413 return $array;
2414 }
2415
2416 =head2 GetDistinctValues($field);
2417
2418 C<$field> is a reference to the fields array
2419
2420 =cut
2421
2422 sub GetDistinctValues {
2423 my ($fieldname,$string)=@_;
2424 # returns a reference to a hash of references to branches...
2425 if ($fieldname=~/\./){
2426 my ($table,$column)=split /\./, $fieldname;
2427 my $dbh = C4::Context->dbh;
2428 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2429 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 ");
2430 $sth->execute;
2431 my $elements=$sth->fetchall_arrayref({});
2432 return $elements;
2433 }
2434 else {
2435 $string||= qq("");
2436 my @servers=qw<biblioserver authorityserver>;
2437 my (@zconns,@results);
2438 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2439 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2440 $results[$i] =
2441 $zconns[$i]->scan(
2442 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2443 );
2444 }
2445 # The big moment: asynchronously retrieve results from all servers
2446 my @elements;
2447 _ZOOM_event_loop(
2448 \@zconns,
2449 \@results,
2450 sub {
2451 my ( $i, $size ) = @_;
2452 for ( my $j = 0 ; $j < $size ; $j++ ) {
2453 my %hashscan;
2454 @hashscan{qw(value cnt)} =
2455 $results[ $i - 1 ]->display_term($j);
2456 push @elements, \%hashscan;
2457 }
2458 }
2459 );
2460 return \@elements;
2461 }
2462 }
2463
2464 =head2 _ZOOM_event_loop
2465
2466 _ZOOM_event_loop(\@zconns, \@results, sub {
2467 my ( $i, $size ) = @_;
2468 ....
2469 } );
2470
2471 Processes a ZOOM event loop and passes control to a closure for
2472 processing the results, and destroying the resultsets.
2473
2474 =cut
2475
2476 sub _ZOOM_event_loop {
2477 my ($zconns, $results, $callback) = @_;
2478 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2479 my $ev = $zconns->[ $i - 1 ]->last_event();
2480 if ( $ev == ZOOM::Event::ZEND ) {
2481 next unless $results->[ $i - 1 ];
2482 my $size = $results->[ $i - 1 ]->size();
2483 if ( $size > 0 ) {
2484 $callback->($i, $size);
2485 }
2486 }
2487 }
2488
2489 foreach my $result (@$results) {
2490 $result->destroy();
2491 }
2492 }
2493
2494 =head2 new_record_from_zebra
2495
2496 Given raw data from a searchengine result set, return a MARC::Record object
2497
2498 This helper function is needed to take into account all the involved
2499 system preferences and configuration variables to properly create the
2500 MARC::Record object.
2501
2502 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2503 data. If we are using DOM, then it has to be MARCXML.
2504
2505 If we are using elasticsearch, it'll already be a MARC::Record and this
2506 function needs a new name.
2507
2508 =cut
2509
2510 sub new_record_from_zebra {
2511
2512 my $server = shift;
2513 my $raw_data = shift;
2514 # Set the default indexing modes
2515 my $search_engine = C4::Context->preference("SearchEngine");
2516 if ($search_engine eq 'Elasticsearch') {
2517 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2518 }
2519 my $index_mode = ( $server eq 'biblioserver' )
2520 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2521 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2522
2523 my $marc_record = eval {
2524 if ( $index_mode eq 'dom' ) {
2525 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2526 } else {
2527 MARC::Record->new_from_usmarc( $raw_data );
2528 }
2529 };
2530
2531 if ($@) {
2532 return;
2533 } else {
2534 return $marc_record;
2535 }
2536
2537 }
2538
2539 END { } # module clean-up code here (global destructor)
2540
2541 1;
2542 __END__
2543
2544 =head1 AUTHOR
2545
2546 Koha Development Team <http://koha-community.org/>
2547
2548 =cut
Koha ILS