?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 25452: Restore alternate address email display
[koha.git] / C4 / Search.pm
blob4c07570b4e1ca3c4d074db3af6b7a48f09633edb
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 Koha::RecordProcessor;
38 use YAML;
39 use URI::Escape;
40 use Business::ISBN;
41 use MARC::Record;
42 use MARC::Field;
43 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
44
45 BEGIN {
46 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
47 }
48
49 =head1 NAME
50
51 C4::Search - Functions for searching the Koha catalog.
52
53 =head1 SYNOPSIS
54
55 See opac/opac-search.pl or catalogue/search.pl for example of usage
56
57 =head1 DESCRIPTION
58
59 This module provides searching functions for Koha's bibliographic databases
60
61 =head1 FUNCTIONS
62
63 =cut
64
65 @ISA = qw(Exporter);
66 @EXPORT = qw(
67 &FindDuplicate
68 &SimpleSearch
69 &searchResults
70 &getRecords
71 &buildQuery
72 &GetDistinctValues
73 &enabled_staff_search_views
74 );
75
76 # make all your functions, whether exported or not;
77
78 =head2 FindDuplicate
79
80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
81
82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
83
84 =cut
85
86 sub FindDuplicate {
87 my ($record) = @_;
88 my $dbh = C4::Context->dbh;
89 my $result = TransformMarcToKoha( $record, '' );
90 my $sth;
91 my $query;
92 my $search;
93 my $type;
94 my ( $biblionumber, $title );
95
96 # search duplicate on ISBN, easy and fast..
97 # ... normalize first
98 if ( $result->{isbn} ) {
99 $result->{isbn} =~ s/\(.*$//;
100 $result->{isbn} =~ s/\s+$//;
101 $query = "isbn:$result->{isbn}";
102 }
103 else {
104 my $QParser;
105 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
106 my $titleindex;
107 my $authorindex;
108 my $op;
109
110 if ($QParser) {
111 $titleindex = 'title|exact';
112 $authorindex = 'author|exact';
113 $op = '&&';
114 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
115 } else {
116 $titleindex = 'ti,ext';
117 $authorindex = 'au,ext';
118 $op = 'and';
119 }
120
121 $result->{title} =~ s /\\//g;
122 $result->{title} =~ s /\"//g;
123 $result->{title} =~ s /\(//g;
124 $result->{title} =~ s /\)//g;
125
126 # FIXME: instead of removing operators, could just do
127 # quotes around the value
128 $result->{title} =~ s/(and|or|not)//g;
129 $query = "$titleindex:\"$result->{title}\"";
130 if ( $result->{author} ) {
131 $result->{author} =~ s /\\//g;
132 $result->{author} =~ s /\"//g;
133 $result->{author} =~ s /\(//g;
134 $result->{author} =~ s /\)//g;
135
136 # remove valid operators
137 $result->{author} =~ s/(and|or|not)//g;
138 $query .= " $op $authorindex:\"$result->{author}\"";
139 }
140 }
141
142 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
143 my @results;
144 if (!defined $error) {
145 foreach my $possible_duplicate_record (@{$searchresults}) {
146 my $marcrecord = new_record_from_zebra(
147 'biblioserver',
148 $possible_duplicate_record
149 );
150
151 my $result = TransformMarcToKoha( $marcrecord, '' );
152
153 # FIXME :: why 2 $biblionumber ?
154 if ($result) {
155 push @results, $result->{'biblionumber'};
156 push @results, $result->{'title'};
157 }
158 }
159 }
160 return @results;
161 }
162
163 =head2 SimpleSearch
164
165 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
166
167 This function provides a simple search API on the bibliographic catalog
168
169 =over 2
170
171 =item C<input arg:>
172
173 * $query can be a simple keyword or a complete CCL query
174 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
175 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
176 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
177 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
178
179
180 =item C<Return:>
181
182 Returns an array consisting of three elements
183 * $error is undefined unless an error is detected
184 * $results is a reference to an array of records.
185 * $total_hits is the number of hits that would have been returned with no limit
186
187 If an error is returned the two other return elements are undefined. If error itself is undefined
188 the other two elements are always defined
189
190 =item C<usage in the script:>
191
192 =back
193
194 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
195
196 if (defined $error) {
197 $template->param(query_error => $error);
198 warn "error: ".$error;
199 output_html_with_http_headers $input, $cookie, $template->output;
200 exit;
201 }
202
203 my $hits = @{$marcresults};
204 my @results;
205
206 for my $r ( @{$marcresults} ) {
207 my $marcrecord = MARC::File::USMARC::decode($r);
208 my $biblio = TransformMarcToKoha($marcrecord,q{});
209
210 #build the iarray of hashs for the template.
211 push @results, {
212 title => $biblio->{'title'},
213 subtitle => $biblio->{'subtitle'},
214 biblionumber => $biblio->{'biblionumber'},
215 author => $biblio->{'author'},
216 publishercode => $biblio->{'publishercode'},
217 publicationyear => $biblio->{'publicationyear'},
218 };
219
220 }
221
222 $template->param(result=>\@results);
223
224 =cut
225
226 sub SimpleSearch {
227 my ( $query, $offset, $max_results, $servers, %options ) = @_;
228
229 return ( 'No query entered', undef, undef ) unless $query;
230 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
231 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
232 my @zoom_queries;
233 my @tmpresults;
234 my @zconns;
235 my $results = [];
236 my $total_hits = 0;
237
238 my $QParser;
239 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
240 if ($QParser) {
241 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
242 }
243
244 # Initialize & Search Zebra
245 for ( my $i = 0 ; $i < @servers ; $i++ ) {
246 eval {
247 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
248 if ($QParser) {
249 $query =~ s/=/:/g unless $options{skip_normalize};
250 $QParser->parse( $query );
251 $query = $QParser->target_syntax($servers[$i]);
252 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
253 } else {
254 $query =~ s/:/=/g unless $options{skip_normalize};
255 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
256 }
257 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
258
259 # error handling
260 my $error =
261 $zconns[$i]->errmsg() . " ("
262 . $zconns[$i]->errcode() . ") "
263 . $zconns[$i]->addinfo() . " "
264 . $zconns[$i]->diagset();
265
266 return ( $error, undef, undef ) if $zconns[$i]->errcode();
267 };
268 if ($@) {
269
270 # caught a ZOOM::Exception
271 my $error =
272 $@->message() . " ("
273 . $@->code() . ") "
274 . $@->addinfo() . " "
275 . $@->diagset();
276 warn $error." for query: $query";
277 return ( $error, undef, undef );
278 }
279 }
280
281 _ZOOM_event_loop(
282 \@zconns,
283 \@tmpresults,
284 sub {
285 my ($i, $size) = @_;
286 my $first_record = defined($offset) ? $offset + 1 : 1;
287 my $hits = $tmpresults[ $i - 1 ]->size();
288 $total_hits += $hits;
289 my $last_record = $hits;
290 if ( defined $max_results && $offset + $max_results < $hits ) {
291 $last_record = $offset + $max_results;
292 }
293
294 for my $j ( $first_record .. $last_record ) {
295 my $record = eval {
296 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
297 ; # 0 indexed
298 };
299 push @{$results}, $record if defined $record;
300 }
301 }
302 );
303
304 foreach my $zoom_query (@zoom_queries) {
305 $zoom_query->destroy();
306 }
307
308 return ( undef, $results, $total_hits );
309 }
310
311 =head2 getRecords
312
313 ( undef, $results_hashref, \@facets_loop ) = getRecords (
314
315 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
316 $results_per_page, $offset, $branches, $itemtypes,
317 $query_type, $scan, $opac
318 );
319
320 The all singing, all dancing, multi-server, asynchronous, scanning,
321 searching, record nabbing, facet-building
322
323 See verbose embedded documentation.
324
325 =cut
326
327 sub getRecords {
328 my (
329 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
330 $results_per_page, $offset, $branches, $itemtypes,
331 $query_type, $scan, $opac
332 ) = @_;
333
334 my @servers = @$servers_ref;
335 my @sort_by = @$sort_by_ref;
336 $offset = 0 if $offset < 0;
337
338 # Initialize variables for the ZOOM connection and results object
339 my $zconn;
340 my @zconns;
341 my @results;
342 my $results_hashref = ();
343
344 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
345 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
346
347 # Initialize variables for the faceted results objects
348 my $facets_counter = {};
349 my $facets_info = {};
350 my $facets = getFacets();
351
352 my @facets_loop; # stores the ref to array of hashes for template facets loop
353
354 ### LOOP THROUGH THE SERVERS
355 for ( my $i = 0 ; $i < @servers ; $i++ ) {
356 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
357
358 # perform the search, create the results objects
359 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
360 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
361
362 #$query_to_use = $simple_query if $scan;
363 warn $simple_query if ( $scan and $DEBUG );
364
365 # Check if we've got a query_type defined, if so, use it
366 eval {
367 if ($query_type) {
368 if ($query_type =~ /^ccl/) {
369 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
370 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
371 } elsif ($query_type =~ /^cql/) {
372 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
373 } elsif ($query_type =~ /^pqf/) {
374 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
375 } else {
376 warn "Unknown query_type '$query_type'. Results undetermined.";
377 }
378 } elsif ($scan) {
379 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
380 } else {
381 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
382 }
383 };
384 if ($@) {
385 warn "WARNING: query problem with $query_to_use " . $@;
386 }
387
388 # Concatenate the sort_by limits and pass them to the results object
389 # Note: sort will override rank
390 my $sort_by;
391 foreach my $sort (@sort_by) {
392 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
393 $sort_by .= "1=1003 <i ";
394 }
395 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
396 $sort_by .= "1=1003 >i ";
397 }
398 elsif ( $sort eq "popularity_asc" ) {
399 $sort_by .= "1=9003 <i ";
400 }
401 elsif ( $sort eq "popularity_dsc" ) {
402 $sort_by .= "1=9003 >i ";
403 }
404 elsif ( $sort eq "call_number_asc" ) {
405 $sort_by .= "1=8007 <i ";
406 }
407 elsif ( $sort eq "call_number_dsc" ) {
408 $sort_by .= "1=8007 >i ";
409 }
410 elsif ( $sort eq "pubdate_asc" ) {
411 $sort_by .= "1=31 <i ";
412 }
413 elsif ( $sort eq "pubdate_dsc" ) {
414 $sort_by .= "1=31 >i ";
415 }
416 elsif ( $sort eq "acqdate_asc" ) {
417 $sort_by .= "1=32 <i ";
418 }
419 elsif ( $sort eq "acqdate_dsc" ) {
420 $sort_by .= "1=32 >i ";
421 }
422 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
423 $sort_by .= "1=4 <i ";
424 }
425 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
426 $sort_by .= "1=4 >i ";
427 }
428 else {
429 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
430 }
431 }
432 if ( $sort_by && !$scan && $results[$i] ) {
433 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
434 warn "WARNING sort $sort_by failed";
435 }
436 }
437 } # finished looping through servers
438
439 # The big moment: asynchronously retrieve results from all servers
440 _ZOOM_event_loop(
441 \@zconns,
442 \@results,
443 sub {
444 my ( $i, $size ) = @_;
445 my $results_hash;
446
447 # loop through the results
448 $results_hash->{'hits'} = $size;
449 my $times;
450 if ( $offset + $results_per_page <= $size ) {
451 $times = $offset + $results_per_page;
452 }
453 else {
454 $times = $size;
455 }
456
457 for ( my $j = $offset ; $j < $times ; $j++ ) {
458 my $records_hash;
459 my $record;
460
461 ## Check if it's an index scan
462 if ($scan) {
463 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
464
465 # here we create a minimal MARC record and hand it off to the
466 # template just like a normal result ... perhaps not ideal, but
467 # it works for now
468 my $tmprecord = MARC::Record->new();
469 $tmprecord->encoding('UTF-8');
470 my $tmptitle;
471 my $tmpauthor;
472
473 # the minimal record in author/title (depending on MARC flavour)
474 if ( C4::Context->preference("marcflavour") eq
475 "UNIMARC" )
476 {
477 $tmptitle = MARC::Field->new(
478 '200', ' ', ' ',
479 a => $term,
480 f => $occ
481 );
482 $tmprecord->append_fields($tmptitle);
483 }
484 else {
485 $tmptitle =
486 MARC::Field->new( '245', ' ', ' ', a => $term, );
487 $tmpauthor =
488 MARC::Field->new( '100', ' ', ' ', a => $occ, );
489 $tmprecord->append_fields($tmptitle);
490 $tmprecord->append_fields($tmpauthor);
491 }
492 $results_hash->{'RECORDS'}[$j] =
493 $tmprecord->as_usmarc();
494 }
495
496 # not an index scan
497 else {
498 $record = $results[ $i - 1 ]->record($j)->raw();
499 # warn "RECORD $j:".$record;
500 $results_hash->{'RECORDS'}[$j] = $record;
501 }
502
503 }
504 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
505
506 # Fill the facets while we're looping, but only for the
507 # biblioserver and not for a scan
508 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
509 $facets_counter = GetFacets( $results[ $i - 1 ] );
510 $facets_info = _get_facets_info( $facets );
511 }
512
513 # BUILD FACETS
514 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
515 for my $link_value (
516 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
517 keys %$facets_counter
518 )
519 {
520 my @this_facets_array;
521 for my $one_facet (
522 sort {
523 $facets_counter->{$link_value}
524 ->{$b} <=> $facets_counter->{$link_value}
525 ->{$a}
526 } keys %{ $facets_counter->{$link_value} }
527 )
528 {
529 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
530 my $facet_link_value = $one_facet;
531 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
532
533 # fix the length that will display in the label,
534 my $facet_label_value = $one_facet;
535 my $facet_max_length = C4::Context->preference(
536 'FacetLabelTruncationLength')
537 || 20;
538 $facet_label_value =
539 substr( $one_facet, 0, $facet_max_length )
540 . "..."
541 if length($facet_label_value) >
542 $facet_max_length;
543
544 # if it's a branch, label by the name, not the code,
545 if ( $link_value =~ /branch/ ) {
546 if ( defined $branches
547 && ref($branches) eq "HASH"
548 && defined $branches->{$one_facet}
549 && ref( $branches->{$one_facet} ) eq
550 "HASH" )
551 {
552 $facet_label_value =
553 $branches->{$one_facet}
554 ->{'branchname'};
555 }
556 else {
557 $facet_label_value = "*";
558 }
559 }
560
561 # if it's a itemtype, label by the name, not the code,
562 if ( $link_value =~ /itype/ ) {
563 if ( defined $itemtypes
564 && ref($itemtypes) eq "HASH"
565 && defined $itemtypes->{$one_facet}
566 && ref( $itemtypes->{$one_facet} ) eq
567 "HASH" )
568 {
569 $facet_label_value =
570 $itemtypes->{$one_facet}
571 ->{translated_description};
572 }
573 }
574
575 # also, if it's a location code, use the name instead of the code
576 if ( $link_value =~ /location/ ) {
577 # TODO Retrieve all authorised values at once, instead of 1 query per entry
578 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
579 $facet_label_value = $av->count ? $av->next->opac_description : '';
580 }
581
582 # also, if it's a collection code, use the name instead of the code
583 if ( $link_value =~ /ccode/ ) {
584 # TODO Retrieve all authorised values at once, instead of 1 query per entry
585 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
586 $facet_label_value = $av->count ? $av->next->opac_description : '';
587 }
588
589 # but we're down with the whole label being in the link's title.
590 push @this_facets_array,
591 {
592 facet_count =>
593 $facets_counter->{$link_value}
594 ->{$one_facet},
595 facet_label_value => $facet_label_value,
596 facet_title_value => $one_facet,
597 facet_link_value => $facet_link_value,
598 type_link_value => $link_value,
599 }
600 if ($facet_label_value);
601 }
602
603 push @facets_loop,
604 {
605 type_link_value => $link_value,
606 type_id => $link_value . "_id",
607 "type_label_"
608 . $facets_info->{$link_value}->{'label_value'} =>
609 1,
610 facets => \@this_facets_array,
611 }
612 unless (
613 (
614 $facets_info->{$link_value}->{'label_value'} =~
615 /Libraries/
616 )
617 and ( Koha::Libraries->search->count == 1 )
618 );
619 }
620 }
621 }
622 );
623
624 # This sorts the facets into alphabetical order
625 if (@facets_loop) {
626 foreach my $f (@facets_loop) {
627 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
628 }
629 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
630 }
631
632 return ( undef, $results_hashref, \@facets_loop );
633 }
634
635 sub GetFacets {
636
637 my $rs = shift;
638 my $facets;
639
640 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
641
642 if ( $use_zebra_facets ) {
643 $facets = _get_facets_from_zebra( $rs );
644 } else {
645 $facets = _get_facets_from_records( $rs );
646 }
647
648 return $facets;
649 }
650
651 sub _get_facets_from_records {
652
653 my $rs = shift;
654
655 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
656 my $facets_config = getFacets();
657 my $facets = {};
658 my $size = $rs->size();
659 my $jmax = $size > $facets_maxrecs
660 ? $facets_maxrecs
661 : $size;
662
663 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
664
665 my $marc_record = new_record_from_zebra (
666 'biblioserver',
667 $rs->record( $j )->raw()
668 );
669
670 if ( ! defined $marc_record ) {
671 warn "ERROR DECODING RECORD - $@: " .
672 $rs->record( $j )->raw();
673 next;
674 }
675
676 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
677 }
678
679 return $facets;
680 }
681
682 =head2 _get_facets_data_from_record
683
684 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
685
686 Internal function that extracts facets information from a MARC::Record object
687 and populates $facets_counter for using in getRecords.
688
689 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
690 facets for Zebra).
691
692 =cut
693
694 sub _get_facets_data_from_record {
695
696 my ( $marc_record, $facets, $facets_counter ) = @_;
697
698 for my $facet (@$facets) {
699
700 my @used_datas = ();
701
702 foreach my $tag ( @{ $facet->{ tags } } ) {
703
704 # tag number is the first three digits
705 my $tag_num = substr( $tag, 0, 3 );
706 # subfields are the remainder
707 my $subfield_letters = substr( $tag, 3 );
708
709 my @fields = $marc_record->field( $tag_num );
710 foreach my $field (@fields) {
711 # If $field->indicator(1) eq 'z', it means it is a 'see from'
712 # field introduced because of IncludeSeeFromInSearches, so skip it
713 next if $field->indicator(1) eq 'z';
714
715 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
716
717 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
718 push @used_datas, $data;
719 $facets_counter->{ $facet->{ idx } }->{ $data }++;
720 }
721 }
722 }
723 }
724 }
725
726 =head2 _get_facets_from_zebra
727
728 my $facets = _get_facets_from_zebra( $result_set )
729
730 Retrieves facets for a specified result set. It loops through the facets defined
731 in C4::Koha::getFacets and returns a hash with the following structure:
732
733 { facet_idx => {
734 facet_value => count
735 },
736 ...
737 }
738
739 =cut
740
741 sub _get_facets_from_zebra {
742
743 my $rs = shift;
744
745 # save current elementSetName
746 my $elementSetName = $rs->option( 'elementSetName' );
747
748 my $facets_loop = getFacets();
749 my $facets_data = {};
750 # loop through defined facets and fill the facets hashref
751 foreach my $facet ( @$facets_loop ) {
752
753 my $idx = $facet->{ idx };
754 my $sep = $facet->{ sep };
755 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
756 if ( $facet_values ) {
757 # we've actually got a result
758 $facets_data->{ $idx } = $facet_values;
759 }
760 }
761 # set elementSetName to its previous value to avoid side effects
762 $rs->option( elementSetName => $elementSetName );
763
764 return $facets_data;
765 }
766
767 =head2 _get_facet_from_result_set
768
769 my $facet_values =
770 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
771
772 Internal function that extracts facet information for a specific index ($facet_idx) and
773 returns a hash containing facet values and count:
774
775 {
776 $facet_value => $count ,
777 ...
778 }
779
780 Warning: this function has the side effect of changing the elementSetName for the result
781 set. It is a helper function for the main loop, which takes care of backing it up for
782 restoring.
783
784 =cut
785
786 sub _get_facet_from_result_set {
787
788 my $facet_idx = shift;
789 my $rs = shift;
790 my $sep = shift;
791
792 my $internal_sep = '<*>';
793 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
794
795 return if ( ! defined $facet_idx || ! defined $rs );
796 # zebra's facet element, untokenized index
797 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
798 # configure zebra results for retrieving the desired facet
799 $rs->option( elementSetName => $facet_element );
800 # get the facet record from result set
801 my $facet = $rs->record( 0 )->raw;
802 # if the facet has no restuls...
803 return if !defined $facet;
804 # TODO: benchmark DOM vs. SAX performance
805 my $facet_dom = XML::LibXML->load_xml(
806 string => ($facet)
807 );
808 my @terms = $facet_dom->getElementsByTagName('term');
809 return if ! @terms;
810
811 my $facets = {};
812 foreach my $term ( @terms ) {
813 my $facet_value = $term->textContent;
814 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
815 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
816 }
817
818 return $facets;
819 }
820
821 =head2 _get_facets_info
822
823 my $facets_info = C4::Search::_get_facets_info( $facets )
824
825 Internal function that extracts facets information and properly builds
826 the data structure needed to render facet labels.
827
828 =cut
829
830 sub _get_facets_info {
831
832 my $facets = shift;
833
834 my $facets_info = {};
835
836 for my $facet ( @$facets ) {
837 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
838 }
839
840 return $facets_info;
841 }
842
843 sub pazGetRecords {
844 my (
845 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
846 $results_per_page, $offset, $branches, $query_type,
847 $scan
848 ) = @_;
849 #NOTE: Parameter $branches is not used here !
850
851 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
852 $paz->init();
853 $paz->search($simple_query);
854 sleep 1; # FIXME: WHY?
855
856 # do results
857 my $results_hashref = {};
858 my $stats = XMLin($paz->stat);
859 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
860
861 # for a grouped search result, the number of hits
862 # is the number of groups returned; 'bib_hits' will have
863 # the total number of bibs.
864 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
865 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
866
867 HIT: foreach my $hit (@{ $results->{'hit'} }) {
868 my $recid = $hit->{recid}->[0];
869
870 my $work_title = $hit->{'md-work-title'}->[0];
871 my $work_author;
872 if (exists $hit->{'md-work-author'}) {
873 $work_author = $hit->{'md-work-author'}->[0];
874 }
875 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
876
877 my $result_group = {};
878 $result_group->{'group_label'} = $group_label;
879 $result_group->{'group_merge_key'} = $recid;
880
881 my $count = 1;
882 if (exists $hit->{count}) {
883 $count = $hit->{count}->[0];
884 }
885 $result_group->{'group_count'} = $count;
886
887 for (my $i = 0; $i < $count; $i++) {
888 # FIXME -- may need to worry about diacritics here
889 my $rec = $paz->record($recid, $i);
890 push @{ $result_group->{'RECORDS'} }, $rec;
891 }
892
893 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
894 }
895
896 # pass through facets
897 my $termlist_xml = $paz->termlist('author,subject');
898 my $terms = XMLin($termlist_xml, forcearray => 1);
899 my @facets_loop = ();
900 #die Dumper($results);
901 # foreach my $list (sort keys %{ $terms->{'list'} }) {
902 # my @facets = ();
903 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
904 # push @facets, {
905 # facet_label_value => $facet->{'name'}->[0],
906 # };
907 # }
908 # push @facets_loop, ( {
909 # type_label => $list,
910 # facets => \@facets,
911 # } );
912 # }
913
914 return ( undef, $results_hashref, \@facets_loop );
915 }
916
917 # TRUNCATION
918 sub _detect_truncation {
919 my ( $operand, $index ) = @_;
920 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
921 @regexpr );
922 $operand =~ s/^ //g;
923 my @wordlist = split( /\s/, $operand );
924 foreach my $word (@wordlist) {
925 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
926 push @rightlefttruncated, $word;
927 }
928 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
929 push @lefttruncated, $word;
930 }
931 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
932 push @righttruncated, $word;
933 }
934 elsif ( index( $word, "*" ) < 0 ) {
935 push @nontruncated, $word;
936 }
937 else {
938 push @regexpr, $word;
939 }
940 }
941 return (
942 \@nontruncated, \@righttruncated, \@lefttruncated,
943 \@rightlefttruncated, \@regexpr
944 );
945 }
946
947 # STEMMING
948 sub _build_stemmed_operand {
949 my ($operand,$lang) = @_;
950 require Lingua::Stem::Snowball ;
951 my $stemmed_operand=q{};
952
953 # Stemmer needs language
954 return $operand unless $lang;
955
956 # If operand contains a digit, it is almost certainly an identifier, and should
957 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
958 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
959 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
960 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
961 return $operand if $operand =~ /\d/;
962
963 # FIXME: the locale should be set based on the user's language and/or search choice
964 #warn "$lang";
965 # Make sure we only use the first two letters from the language code
966 $lang = lc(substr($lang, 0, 2));
967 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
968 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
969 if ($lang eq 'nb' || $lang eq 'nn') {
970 $lang = 'no';
971 }
972 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
973 encoding => "UTF-8" );
974
975 my @words = split( / /, $operand );
976 my @stems = $stemmer->stem(\@words);
977 for my $stem (@stems) {
978 $stemmed_operand .= "$stem";
979 $stemmed_operand .= "?"
980 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
981 $stemmed_operand .= " ";
982 }
983 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
984 return $stemmed_operand;
985 }
986
987 # FIELD WEIGHTING
988 sub _build_weighted_query {
989
990 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
991 # pretty well but could work much better if we had a smarter query parser
992 my ( $operand, $stemmed_operand, $index ) = @_;
993 my $stemming = C4::Context->preference("QueryStemming") || 0;
994 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
995 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
996 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
997
998 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
999
1000 # Keyword, or, no index specified
1001 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1002 $weighted_query .=
1003 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1004 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1005 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1006 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1007 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1008 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1009 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1010 if $fuzzy_enabled; # add fuzzy, word list
1011 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1012 if ( $stemming and $stemmed_operand )
1013 ; # add stemming, right truncation
1014 $weighted_query .= " or wrdl,r9=\"$operand\"";
1015
1016 # embedded sorting: 0 a-z; 1 z-a
1017 # $weighted_query .= ") or (sort1,aut=1";
1018 }
1019
1020 # Barcode searches should skip this process
1021 elsif ( $index eq 'bc' ) {
1022 $weighted_query .= "bc=\"$operand\"";
1023 }
1024
1025 # Authority-number searches should skip this process
1026 elsif ( $index eq 'an' ) {
1027 $weighted_query .= "an=\"$operand\"";
1028 }
1029
1030 # If the index is numeric, don't autoquote it.
1031 elsif ( $index =~ /,st-numeric$/ ) {
1032 $weighted_query .= " $index=$operand";
1033 }
1034
1035 # If the index already has more than one qualifier, wrap the operand
1036 # in quotes and pass it back (assumption is that the user knows what they
1037 # are doing and won't appreciate us mucking up their query
1038 elsif ( $index =~ ',' ) {
1039 $weighted_query .= " $index=\"$operand\"";
1040 }
1041
1042 #TODO: build better cases based on specific search indexes
1043 else {
1044 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1045 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1046 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1047 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1048 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1049 if $fuzzy_enabled; # add fuzzy, word list
1050 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1051 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1052 }
1053
1054 $weighted_query .= "))"; # close rank specification
1055 return $weighted_query;
1056 }
1057
1058 =head2 getIndexes
1059
1060 Return an array with available indexes.
1061
1062 =cut
1063
1064 sub getIndexes{
1065 my @indexes = (
1066 # biblio indexes
1067 'ab',
1068 'Abstract',
1069 'acqdate',
1070 'allrecords',
1071 'an',
1072 'Any',
1073 'at',
1074 'arl',
1075 'arp',
1076 'au',
1077 'aub',
1078 'aud',
1079 'audience',
1080 'auo',
1081 'aut',
1082 'Author',
1083 'Author-in-order ',
1084 'Author-personal-bibliography',
1085 'Authority-Number',
1086 'authtype',
1087 'bc',
1088 'Bib-level',
1089 'biblionumber',
1090 'bio',
1091 'biography',
1092 'callnum',
1093 'cfn',
1094 'Chronological-subdivision',
1095 'cn-bib-source',
1096 'cn-bib-sort',
1097 'cn-class',
1098 'cn-item',
1099 'cn-prefix',
1100 'cn-suffix',
1101 'cpn',
1102 'Code-institution',
1103 'Conference-name',
1104 'Conference-name-heading',
1105 'Conference-name-see',
1106 'Conference-name-seealso',
1107 'Content-type',
1108 'Control-number',
1109 'copydate',
1110 'Corporate-name',
1111 'Corporate-name-heading',
1112 'Corporate-name-see',
1113 'Corporate-name-seealso',
1114 'Country-publication',
1115 'ctype',
1116 'curriculum',
1117 'date-entered-on-file',
1118 'Date-of-acquisition',
1119 'Date-of-publication',
1120 'Date-time-last-modified',
1121 'Dewey-classification',
1122 'Dissertation-information',
1123 'diss',
1124 'dtlm',
1125 'EAN',
1126 'extent',
1127 'fic',
1128 'fiction',
1129 'Form-subdivision',
1130 'format',
1131 'Geographic-subdivision',
1132 'he',
1133 'Heading',
1134 'Heading-use-main-or-added-entry',
1135 'Heading-use-series-added-entry ',
1136 'Heading-use-subject-added-entry',
1137 'Host-item',
1138 'id-other',
1139 'ident',
1140 'Identifier-standard',
1141 'Illustration-code',
1142 'Index-term-genre',
1143 'Index-term-uncontrolled',
1144 'Interest-age-level',
1145 'Interest-grade-level',
1146 'ISBN',
1147 'isbn',
1148 'ISSN',
1149 'issn',
1150 'itemtype',
1151 'kw',
1152 'Koha-Auth-Number',
1153 'l-format',
1154 'language',
1155 'language-original',
1156 'lc-card',
1157 'LC-card-number',
1158 'lcn',
1159 'lex',
1160 'lexile-number',
1161 'llength',
1162 'ln',
1163 'ln-audio',
1164 'ln-subtitle',
1165 'Local-classification',
1166 'Local-number',
1167 'Match-heading',
1168 'Match-heading-see-from',
1169 'Material-type',
1170 'mc-itemtype',
1171 'mc-rtype',
1172 'mus',
1173 'name',
1174 'Music-number',
1175 'Name-geographic',
1176 'Name-geographic-heading',
1177 'Name-geographic-see',
1178 'Name-geographic-seealso',
1179 'nb',
1180 'Note',
1181 'notes',
1182 'ns',
1183 'nt',
1184 'Other-control-number',
1185 'pb',
1186 'Personal-name',
1187 'Personal-name-heading',
1188 'Personal-name-see',
1189 'Personal-name-seealso',
1190 'pl',
1191 'Place-publication',
1192 'pn',
1193 'popularity',
1194 'pubdate',
1195 'Publisher',
1196 'Provider',
1197 'pv',
1198 'Reading-grade-level',
1199 'Record-control-number',
1200 'rcn',
1201 'Record-type',
1202 'rtype',
1203 'se',
1204 'See',
1205 'See-also',
1206 'sn',
1207 'Stock-number',
1208 'su',
1209 'Subject',
1210 'Subject-heading-thesaurus',
1211 'Subject-name-personal',
1212 'Subject-subdivision',
1213 'Summary',
1214 'Suppress',
1215 'su-geo',
1216 'su-na',
1217 'su-to',
1218 'su-ut',
1219 'ut',
1220 'Term-genre-form',
1221 'Term-genre-form-heading',
1222 'Term-genre-form-see',
1223 'Term-genre-form-seealso',
1224 'ti',
1225 'Title',
1226 'Title-cover',
1227 'Title-series',
1228 'Title-uniform',
1229 'Title-uniform-heading',
1230 'Title-uniform-see',
1231 'Title-uniform-seealso',
1232 'totalissues',
1233 'yr',
1234
1235 # items indexes
1236 'acqsource',
1237 'barcode',
1238 'bc',
1239 'branch',
1240 'ccode',
1241 'classification-source',
1242 'cn-sort',
1243 'coded-location-qualifier',
1244 'copynumber',
1245 'damaged',
1246 'datelastborrowed',
1247 'datelastseen',
1248 'holdingbranch',
1249 'homebranch',
1250 'issues',
1251 'item',
1252 'itemnumber',
1253 'itype',
1254 'Local-classification',
1255 'location',
1256 'lost',
1257 'materials-specified',
1258 'mc-ccode',
1259 'mc-itype',
1260 'mc-loc',
1261 'notforloan',
1262 'Number-local-acquisition',
1263 'onloan',
1264 'price',
1265 'renewals',
1266 'replacementprice',
1267 'replacementpricedate',
1268 'reserves',
1269 'restricted',
1270 'stack',
1271 'stocknumber',
1272 'inv',
1273 'uri',
1274 'withdrawn',
1275
1276 # subject related
1277 );
1278
1279 return \@indexes;
1280 }
1281
1282 =head2 _handle_exploding_index
1283
1284 my $query = _handle_exploding_index($index, $term)
1285
1286 Callback routine to generate the search for "exploding" indexes (i.e.
1287 those indexes which are turned into multiple or-connected searches based
1288 on authority data).
1289
1290 =cut
1291
1292 sub _handle_exploding_index {
1293 my ($QParser, $filter, $params, $negate, $server) = @_;
1294 my $index = $filter;
1295 my $term = join(' ', @$params);
1296
1297 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1298
1299 my $marcflavour = C4::Context->preference('marcflavour');
1300
1301 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1302 my $wantedcodes = '';
1303 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1304 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1305 foreach my $auth (@$results) {
1306 my $record = MARC::Record->new_from_usmarc($auth);
1307 my @references = $record->field('5..');
1308 if (@references) {
1309 if ($index eq 'su-br') {
1310 $wantedcodes = 'g';
1311 } elsif ($index eq 'su-na') {
1312 $wantedcodes = 'h';
1313 } elsif ($index eq 'su-rl') {
1314 $wantedcodes = '';
1315 }
1316 foreach my $reference (@references) {
1317 my $codes = $reference->subfield($codesubfield);
1318 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1319 }
1320 }
1321 }
1322 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1323 return $query;
1324 }
1325
1326 =head2 parseQuery
1327
1328 ( $operators, $operands, $indexes, $limits,
1329 $sort_by, $scan, $lang ) =
1330 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1331
1332 Shim function to ease the transition from buildQuery to a new QueryParser.
1333 This function is called at the beginning of buildQuery, and modifies
1334 buildQuery's input. If it can handle the input, it returns a query that
1335 buildQuery will not try to parse.
1336
1337 =cut
1338
1339 sub parseQuery {
1340 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1341
1342 my @operators = $operators ? @$operators : ();
1343 my @indexes = $indexes ? @$indexes : ();
1344 my @operands = $operands ? @$operands : ();
1345 my @limits = $limits ? @$limits : ();
1346 my @sort_by = $sort_by ? @$sort_by : ();
1347
1348 my $query = $operands[0];
1349 my $index;
1350 my $term;
1351 my $query_desc;
1352
1353 my $QParser;
1354 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1355 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1356 undef $QParser if (scalar @limits > 0);
1357
1358 if ($QParser)
1359 {
1360 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1361 $query = '';
1362 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1363 next unless $operands[$ii];
1364 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1365 if ($query);
1366 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1367 $query .= $operands[$ii];
1368 }
1369 elsif ( $indexes[$ii] =~ m/su-/ ) {
1370 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1371 }
1372 else {
1373 $query .=
1374 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1375 }
1376 }
1377 foreach my $limit (@limits) {
1378 }
1379 if ( scalar(@sort_by) > 0 ) {
1380 my $modifier_re =
1381 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1382 $query =~ s/$modifier_re//g;
1383 foreach my $modifier (@sort_by) {
1384 $query .= " #$modifier";
1385 }
1386 }
1387
1388 $query_desc = $query;
1389 $query_desc =~ s/\s+/ /g;
1390 if ( C4::Context->preference("QueryWeightFields") ) {
1391 }
1392 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1393 { 'target_syntax_callback' => \&_handle_exploding_index } );
1394 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1395 { 'target_syntax_callback' => \&_handle_exploding_index } );
1396 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1397 { 'target_syntax_callback' => \&_handle_exploding_index } );
1398 $QParser->parse($query);
1399 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1400 }
1401 else {
1402 require Koha::QueryParser::Driver::PQF;
1403 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1404 s/$modifier_re//g for @operands;
1405 }
1406
1407 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1408 }
1409
1410 =head2 buildQuery
1411
1412 ( $error, $query,
1413 $simple_query, $query_cgi,
1414 $query_desc, $limit,
1415 $limit_cgi, $limit_desc,
1416 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1417
1418 Build queries and limits in CCL, CGI, Human,
1419 handle truncation, stemming, field weighting, fuzziness, etc.
1420
1421 See verbose embedded documentation.
1422
1423
1424 =cut
1425
1426 sub buildQuery {
1427 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1428
1429 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1430
1431 my $query_desc;
1432 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1433
1434 # dereference
1435 my @operators = $operators ? @$operators : ();
1436 my @indexes = $indexes ? @$indexes : ();
1437 my @operands = $operands ? @$operands : ();
1438 my @limits = $limits ? @$limits : ();
1439 my @sort_by = $sort_by ? @$sort_by : ();
1440
1441 my $stemming = C4::Context->preference("QueryStemming") || 0;
1442 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1443 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1444 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1445
1446 my $query = $operands[0];
1447 my $simple_query = $operands[0];
1448
1449 # initialize the variables we're passing back
1450 my $query_cgi;
1451 my $query_type;
1452
1453 my $limit;
1454 my $limit_cgi;
1455 my $limit_desc;
1456
1457 my $cclq = 0;
1458 my $cclindexes = getIndexes();
1459 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1460 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1461 my $dx = lc($1);
1462 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1463 }
1464 $query = "ccl=$query" if $cclq;
1465 }
1466
1467 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1468 # DIAGNOSTIC ONLY!!
1469 if ( $query =~ /^ccl=/ ) {
1470 my $q=$';
1471 # This is needed otherwise ccl= and &limit won't work together, and
1472 # this happens when selecting a subject on the opac-detail page
1473 @limits = grep {!/^$/} @limits;
1474 my $original_q = $q; # without available part
1475 unless ( grep { /^available$/ } @limits ) {
1476 $q =~ s| and \( \(allrecords,AlwaysMatches=''\) and \(not-onloan-count,st-numeric >= 1\) and \(lost,st-numeric=0\) \)||;
1477 $original_q = $q;
1478 }
1479 if ( @limits ) {
1480 if ( grep { /^available$/ } @limits ) {
1481 $q .= q| and ( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )|;
1482 delete $limits['available'];
1483 }
1484 $q .= ' and '.join(' and ', @limits) if @limits;
1485 }
1486 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1487 }
1488 if ( $query =~ /^cql=/ ) {
1489 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1490 }
1491 if ( $query =~ /^pqf=/ ) {
1492 if ($query_desc) {
1493 $query_cgi = "q=".uri_escape_utf8($query_desc);
1494 } else {
1495 $query_desc = $';
1496 $query_cgi = "q=pqf=".uri_escape_utf8($');
1497 }
1498 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1499 }
1500
1501 # pass nested queries directly
1502 # FIXME: need better handling of some of these variables in this case
1503 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1504 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1505 # if ( $query =~ /(\(|\))/ ) {
1506 # return (
1507 # undef, $query, $simple_query, $query_cgi,
1508 # $query, $limit, $limit_cgi, $limit_desc,
1509 # 'ccl'
1510 # );
1511 # }
1512
1513 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1514 # query operands and indexes and add stemming, truncation, field weighting, etc.
1515 # Once we do so, we'll end up with a value in $query, just like if we had an
1516 # incoming $query from the user
1517 else {
1518 $query = ""
1519 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1520 my $previous_operand
1521 ; # a flag used to keep track if there was a previous query
1522 # if there was, we can apply the current operator
1523 # for every operand
1524 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1525
1526 # COMBINE OPERANDS, INDEXES AND OPERATORS
1527 if ( ($operands[$i] // '') ne '' ) {
1528 $operands[$i]=~s/^\s+//;
1529
1530 # A flag to determine whether or not to add the index to the query
1531 my $indexes_set;
1532
1533 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1534 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1535 $weight_fields = 0;
1536 $stemming = 0;
1537 } else {
1538 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1539 }
1540 my $operand = $operands[$i];
1541 my $index = $indexes[$i];
1542
1543 # Add index-specific attributes
1544
1545 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1546 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1547 #which is processed higher up in this sub. Other than that, year searches are typically
1548 #handled as limits which are not processed her either.
1549
1550 # Search ranges: Date of Publication, st-numeric
1551 if ( $index =~ /(yr|st-numeric)/ ) {
1552 #weight_fields/relevance search causes errors with date ranges
1553 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1554 #In the case of YYYY-YYYY, it will return no results
1555 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1556 }
1557
1558 # Date of Acquisition
1559 elsif ( $index =~ /acqdate/ ) {
1560 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1561 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1562 #top of the results just because they have lots of item records matching that date.
1563 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1564 #irrelevant here
1565 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1566 }
1567 # ISBN,ISSN,Standard Number, don't need special treatment
1568 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1569 (
1570 $stemming, $auto_truncation,
1571 $weight_fields, $fuzzy_enabled
1572 ) = ( 0, 0, 0, 0 );
1573
1574 if ( $index eq 'nb' ) {
1575 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1576 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1577 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1578 $indexes[$i] = $index = '';
1579 }
1580 }
1581 }
1582
1583 if(not $index){
1584 $index = 'kw';
1585 }
1586
1587 # Set default structure attribute (word list)
1588 my $struct_attr = q{};
1589 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1590 $struct_attr = ",wrdl";
1591 }
1592
1593 # Some helpful index variants
1594 my $index_plus = $index . $struct_attr . ':';
1595 my $index_plus_comma = $index . $struct_attr . ',';
1596
1597 if ($auto_truncation){
1598 unless ( $index =~ /,(st-|phr|ext)/ ) {
1599 #FIXME only valid with LTR scripts
1600 $operand=join(" ",map{
1601 (index($_,"*")>0?"$_":"$_*")
1602 }split (/\s+/,$operand));
1603 warn $operand if $DEBUG;
1604 }
1605 }
1606
1607 # Detect Truncation
1608 my $truncated_operand;
1609 my( $nontruncated, $righttruncated, $lefttruncated,
1610 $rightlefttruncated, $regexpr
1611 ) = _detect_truncation( $operand, $index );
1612 warn
1613 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1614 if $DEBUG;
1615
1616 # Apply Truncation
1617 if (
1618 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1619 scalar(@$rightlefttruncated) > 0 )
1620 {
1621
1622 # Don't field weight or add the index to the query, we do it here
1623 $indexes_set = 1;
1624 undef $weight_fields;
1625 my $previous_truncation_operand;
1626 if (scalar @$nontruncated) {
1627 $truncated_operand .= "$index_plus @$nontruncated ";
1628 $previous_truncation_operand = 1;
1629 }
1630 if (scalar @$righttruncated) {
1631 $truncated_operand .= "and " if $previous_truncation_operand;
1632 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1633 $previous_truncation_operand = 1;
1634 }
1635 if (scalar @$lefttruncated) {
1636 $truncated_operand .= "and " if $previous_truncation_operand;
1637 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1638 $previous_truncation_operand = 1;
1639 }
1640 if (scalar @$rightlefttruncated) {
1641 $truncated_operand .= "and " if $previous_truncation_operand;
1642 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1643 $previous_truncation_operand = 1;
1644 }
1645 }
1646 $operand = $truncated_operand if $truncated_operand;
1647 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1648
1649 # Handle Stemming
1650 my $stemmed_operand;
1651 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1652 if $stemming;
1653
1654 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1655
1656 # Handle Field Weighting
1657 my $weighted_operand;
1658 if ($weight_fields) {
1659 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1660 $operand = $weighted_operand;
1661 $indexes_set = 1;
1662 }
1663
1664 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1665
1666 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1667 query => $query,
1668 query_cgi => $query_cgi,
1669 query_desc => $query_desc,
1670 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1671 parsed_operand => $operand,
1672 original_operand => $operands[$i] // '',
1673 index => $index,
1674 index_plus => $index_plus,
1675 indexes_set => $indexes_set,
1676 previous_operand => $previous_operand,
1677 });
1678
1679 } #/if $operands
1680 } # /for
1681 }
1682 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1683
1684 # add limits
1685 my %group_OR_limits;
1686 my $availability_limit;
1687 foreach my $this_limit (@limits) {
1688 next unless $this_limit;
1689 if ( $this_limit =~ /available/ ) {
1690 #
1691 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1692 ## In English:
1693 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1694 $availability_limit .=
1695 "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )";
1696 $limit_cgi .= "&limit=available";
1697 $limit_desc .= "";
1698 }
1699
1700 # group_OR_limits, prefixed by mc-
1701 # OR every member of the group
1702 elsif ( $this_limit =~ /mc/ ) {
1703 my ($k,$v) = split(/:/, $this_limit,2);
1704 if ( $k !~ /mc-i(tem)?type/ ) {
1705 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1706 $this_limit =~ tr/"//d;
1707 $this_limit = $k.":'".$v."'";
1708 }
1709
1710 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1711 $limit_desc .= " or " if $group_OR_limits{$k};
1712 $group_OR_limits{$k} .= "$this_limit";
1713 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1714 $limit_desc .= " $this_limit";
1715 }
1716
1717 # Regular old limits
1718 else {
1719 $limit .= " and " if $limit || $query;
1720 $limit .= "$this_limit";
1721 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1722 if ($this_limit =~ /^branch:(.+)/) {
1723 my $branchcode = $1;
1724 my $library = Koha::Libraries->find( $branchcode );
1725 if (defined $library) {
1726 $limit_desc .= " branch:" . $library->branchname;
1727 } else {
1728 $limit_desc .= " $this_limit";
1729 }
1730 } else {
1731 $limit_desc .= " $this_limit";
1732 }
1733 }
1734 }
1735 foreach my $k (keys (%group_OR_limits)) {
1736 $limit .= " and " if ( $query || $limit );
1737 $limit .= "($group_OR_limits{$k})";
1738 }
1739 if ($availability_limit) {
1740 $limit .= " and " if ( $query || $limit );
1741 $limit .= "($availability_limit)";
1742 }
1743
1744 # Normalize the query and limit strings
1745 # This is flawed , means we can't search anything with : in it
1746 # if user wants to do ccl or cql, start the query with that
1747 # $query =~ s/:/=/g;
1748 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1749 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1750 $query =~ s/(?<=(wrdl)):/=/g;
1751 $query =~ s/(?<=(trn|phr)):/=/g;
1752 $query =~ s/(?<=(st-numeric)):/=/g;
1753 $query =~ s/(?<=(st-year)):/=/g;
1754 $query =~ s/(?<=(st-date-normalized)):/=/g;
1755 $limit =~ s/:/=/g;
1756 for ( $query, $query_desc, $limit, $limit_desc ) {
1757 s/ +/ /g; # remove extra spaces
1758 s/^ //g; # remove any beginning spaces
1759 s/ $//g; # remove any ending spaces
1760 s/==/=/g; # remove double == from query
1761 }
1762 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1763
1764 for ($query_cgi,$simple_query) {
1765 s/"//g;
1766 }
1767 # append the limit to the query
1768 $query .= " " . $limit;
1769
1770 # Warnings if DEBUG
1771 if ($DEBUG) {
1772 warn "QUERY:" . $query;
1773 warn "QUERY CGI:" . $query_cgi;
1774 warn "QUERY DESC:" . $query_desc;
1775 warn "LIMIT:" . $limit;
1776 warn "LIMIT CGI:" . $limit_cgi;
1777 warn "LIMIT DESC:" . $limit_desc;
1778 warn "---------\nLeave buildQuery\n---------";
1779 }
1780
1781 return (
1782 undef, $query, $simple_query, $query_cgi,
1783 $query_desc, $limit, $limit_cgi, $limit_desc,
1784 $query_type
1785 );
1786 }
1787
1788 =head2 _build_initial_query
1789
1790 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1791
1792 Build a section of the initial query containing indexes, operators, and operands.
1793
1794 =cut
1795
1796 sub _build_initial_query {
1797 my ($params) = @_;
1798
1799 my $operator = "";
1800 if ($params->{previous_operand}){
1801 #If there is a previous operand, add a supplied operator or the default 'and'
1802 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1803 }
1804
1805 #NOTE: indexes_set is typically set when doing truncation or field weighting
1806 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1807
1808 #e.g. "kw,wrdl:test"
1809 #e.g. " and kw,wrdl:test"
1810 $params->{query} .= $operator . $operand;
1811
1812 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1813 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1814 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1815
1816 #e.g. " and kw,wrdl: test"
1817 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1818
1819 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1820
1821 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1822 }
1823
1824 =head2 searchResults
1825
1826 my @search_results = searchResults($search_context, $searchdesc, $hits,
1827 $results_per_page, $offset, $scan,
1828 @marcresults);
1829
1830 Format results in a form suitable for passing to the template
1831
1832 =cut
1833
1834 # IMO this subroutine is pretty messy still -- it's responsible for
1835 # building the HTML output for the template
1836 sub searchResults {
1837 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1838 my $dbh = C4::Context->dbh;
1839 my @newresults;
1840
1841 require C4::Items;
1842
1843 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1844 my ($is_opac, $hidelostitems);
1845 if ($search_context->{'interface'} eq 'opac') {
1846 $hidelostitems = C4::Context->preference('hidelostitems');
1847 $is_opac = 1;
1848 }
1849
1850 my $record_processor = Koha::RecordProcessor->new({
1851 filters => 'ViewPolicy'
1852 });
1853
1854 #Build branchnames hash
1855 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1856
1857 # FIXME - We build an authorised values hash here, using the default framework
1858 # though it is possible to have different authvals for different fws.
1859
1860 my $shelflocations =
1861 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1862
1863 # get notforloan authorised value list (see $shelflocations FIXME)
1864 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1865 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1866
1867 #Get itemtype hash
1868 my $itemtypes = Koha::ItemTypes->search_with_localization;
1869 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1870
1871 #search item field code
1872 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1873
1874 ## find column names of items related to MARC
1875 my %subfieldstosearch;
1876 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1877 for my $column ( @columns ) {
1878 my ( $tagfield, $tagsubfield ) =
1879 &GetMarcFromKohaField( "items." . $column, "" );
1880 if ( defined $tagsubfield ) {
1881 $subfieldstosearch{$column} = $tagsubfield;
1882 }
1883 }
1884
1885 # handle which records to actually retrieve
1886 my $times;
1887 if ( $hits && $offset + $results_per_page <= $hits ) {
1888 $times = $offset + $results_per_page;
1889 }
1890 else {
1891 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1892 }
1893
1894 my $marcflavour = C4::Context->preference("marcflavour");
1895 # We get the biblionumber position in MARC
1896 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1897
1898 # set stuff for XSLT processing here once, not later again for every record we retrieved
1899 my $xslfile;
1900 my $xslsyspref;
1901 if( $is_opac ){
1902 $xslsyspref = "OPACXSLTResultsDisplay";
1903 $xslfile = C4::Context->preference( $xslsyspref );
1904 } else {
1905 $xslsyspref = "XSLTResultsDisplay";
1906 $xslfile = C4::Context->preference( $xslsyspref ) || "default";
1907 }
1908 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1909 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1910
1911 my $userenv = C4::Context->userenv;
1912 my $logged_in_user
1913 = ( defined $userenv and $userenv->{number} )
1914 ? Koha::Patrons->find( $userenv->{number} )
1915 : undef;
1916 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1917
1918 # loop through all of the records we've retrieved
1919 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1920
1921 my $marcrecord;
1922 if ($scan) {
1923 # For Scan searches we built USMARC data
1924 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1925 } else {
1926 # Normal search, render from Zebra's output
1927 $marcrecord = new_record_from_zebra(
1928 'biblioserver',
1929 $marcresults->[$i]
1930 );
1931
1932 if ( ! defined $marcrecord ) {
1933 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1934 next;
1935 }
1936 }
1937
1938 my $fw = $scan
1939 ? undef
1940 : $bibliotag < 10
1941 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1942 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1943
1944 SetUTF8Flag($marcrecord);
1945 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1946 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1947 $oldbiblio->{result_number} = $i + 1;
1948
1949 # add imageurl to itemtype if there is one
1950 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1951
1952 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1953 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1954 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1955 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1956 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1957
1958 # edition information, if any
1959 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1960 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1961 # Build summary if there is one (the summary is defined in the itemtypes table)
1962 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1963 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1964 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1965 my @fields = $marcrecord->fields();
1966
1967 my $newsummary;
1968 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1969 my $tags = {};
1970 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1971 $tag =~ /(.{3})(.)/;
1972 if($marcrecord->field($1)){
1973 my @abc = $marcrecord->field($1)->subfield($2);
1974 $tags->{$tag} = $#abc + 1 ;
1975 }
1976 }
1977
1978 # We catch how many times to repeat this line
1979 my $max = 0;
1980 foreach my $tag (keys(%$tags)){
1981 $max = $tags->{$tag} if($tags->{$tag} > $max);
1982 }
1983
1984 # we replace, and repeat each line
1985 for (my $i = 0 ; $i < $max ; $i++){
1986 my $newline = $line;
1987
1988 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1989 $tag =~ /(.{3})(.)/;
1990
1991 if($marcrecord->field($1)){
1992 my @repl = $marcrecord->field($1)->subfield($2);
1993 my $subfieldvalue = $repl[$i];
1994 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1995 }
1996 }
1997 $newsummary .= "$newline\n";
1998 }
1999 }
2000
2001 $newsummary =~ s/\[(.*?)]//g;
2002 $newsummary =~ s/\n/<br\/>/g;
2003 $oldbiblio->{summary} = $newsummary;
2004 }
2005
2006 # Pull out the items fields
2007 my @fields = $marcrecord->field($itemtag);
2008 my $marcflavor = C4::Context->preference("marcflavour");
2009
2010 # adding linked items that belong to host records
2011 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
2012 my $analyticsfield = '773';
2013 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2014 $analyticsfield = '773';
2015 } elsif ($marcflavor eq 'UNIMARC') {
2016 $analyticsfield = '461';
2017 }
2018 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2019 my $hostbiblionumber = $hostfield->subfield("0");
2020 my $linkeditemnumber = $hostfield->subfield("9");
2021 if( $hostbiblionumber ) {
2022 my $hostbiblio = GetMarcBiblio({
2023 biblionumber => $hostbiblionumber,
2024 embed_items => 1 });
2025 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2026 if( $hostbiblio ) {
2027 my @hostitems = $hostbiblio->field($itemfield);
2028 foreach my $hostitem (@hostitems){
2029 if ($hostitem->subfield("9") eq $linkeditemnumber){
2030 my $linkeditem =$hostitem;
2031 # append linked items if they exist
2032 push @fields, $linkeditem if $linkeditem;
2033 }
2034 }
2035 }
2036 }
2037 }
2038 }
2039
2040 # Setting item statuses for display
2041 my @available_items_loop;
2042 my @onloan_items_loop;
2043 my @other_items_loop;
2044
2045 my $available_items;
2046 my $onloan_items;
2047 my $other_items;
2048
2049 my $ordered_count = 0;
2050 my $available_count = 0;
2051 my $onloan_count = 0;
2052 my $longoverdue_count = 0;
2053 my $other_count = 0;
2054 my $withdrawn_count = 0;
2055 my $itemlost_count = 0;
2056 my $hideatopac_count = 0;
2057 my $itembinding_count = 0;
2058 my $itemdamaged_count = 0;
2059 my $item_in_transit_count = 0;
2060 my $can_place_holds = 0;
2061 my $item_onhold_count = 0;
2062 my $notforloan_count = 0;
2063 my $items_count = scalar(@fields);
2064 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2065 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2066 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2067
2068 # loop through every item
2069 foreach my $field (@fields) {
2070 my $item;
2071
2072 # populate the items hash
2073 foreach my $code ( keys %subfieldstosearch ) {
2074 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2075 }
2076 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2077
2078 # OPAC hidden items
2079 if ($is_opac) {
2080 # hidden because lost
2081 if ($hidelostitems && $item->{itemlost}) {
2082 $hideatopac_count++;
2083 next;
2084 }
2085 # hidden based on OpacHiddenItems syspref
2086 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
2087 if (scalar @hi) {
2088 push @hiddenitems, @hi;
2089 $hideatopac_count++;
2090 next;
2091 }
2092 }
2093
2094 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2095 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2096
2097 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2098 if ($item->{$hbranch}) {
2099 $item->{'branchname'} = $branches{$item->{$hbranch}};
2100 }
2101 elsif ($item->{$otherbranch}) { # Last resort
2102 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2103 }
2104
2105 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2106 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2107 if ( $item->{onloan}
2108 and $logged_in_user
2109 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
2110 {
2111 $onloan_count++;
2112 my $key = $prefix . $item->{onloan} . $item->{barcode};
2113 $onloan_items->{$key}->{due_date} = $item->{onloan};
2114 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2115 $onloan_items->{$key}->{branchname} = $item->{branchname};
2116 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2117 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2118 $onloan_items->{$key}->{description} = $item->{description};
2119 $onloan_items->{$key}->{imageurl} =
2120 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2121
2122 # if something's checked out and lost, mark it as 'long overdue'
2123 if ( $item->{itemlost} ) {
2124 $onloan_items->{$key}->{longoverdue}++;
2125 $longoverdue_count++;
2126 }
2127 else { # can place holds as long as item isn't lost
2128 $can_place_holds = 1;
2129 }
2130 }
2131
2132 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2133 else {
2134
2135 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2136
2137 # item is on order
2138 if ( $item->{notforloan} < 0 ) {
2139 $ordered_count++;
2140 } elsif ( $item->{notforloan} > 0 ) {
2141 $notforloan_count++;
2142 }
2143
2144 # is item in transit?
2145 my $transfertwhen = '';
2146 my ($transfertfrom, $transfertto);
2147
2148 # is item on the reserve shelf?
2149 my $reservestatus = '';
2150
2151 unless ($item->{withdrawn}
2152 || $item->{itemlost}
2153 || $item->{damaged}
2154 || $item->{notforloan}
2155 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2156 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2157
2158 # A couple heuristics to limit how many times
2159 # we query the database for item transfer information, sacrificing
2160 # accuracy in some cases for speed;
2161 #
2162 # 1. don't query if item has one of the other statuses
2163 # 2. don't check transit status if the bib has
2164 # more than 20 items
2165 #
2166 # FIXME: to avoid having the query the database like this, and to make
2167 # the in transit status count as unavailable for search limiting,
2168 # should map transit status to record indexed in Zebra.
2169 #
2170 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2171 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2172 }
2173
2174 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2175 if ( $item->{withdrawn}
2176 || $item->{itemlost}
2177 || $item->{damaged}
2178 || $item->{notforloan}
2179 || $reservestatus eq 'Waiting'
2180 || ($transfertwhen ne ''))
2181 {
2182 $withdrawn_count++ if $item->{withdrawn};
2183 $itemlost_count++ if $item->{itemlost};
2184 $itemdamaged_count++ if $item->{damaged};
2185 $item_in_transit_count++ if $transfertwhen ne '';
2186 $item_onhold_count++ if $reservestatus eq 'Waiting';
2187 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2188
2189 # can place a hold on a item if
2190 # not lost nor withdrawn
2191 # not damaged unless AllowHoldsOnDamagedItems is true
2192 # item is either for loan or on order (notforloan < 0)
2193 $can_place_holds = 1
2194 if (
2195 !$item->{itemlost}
2196 && !$item->{withdrawn}
2197 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2198 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2199 );
2200
2201 $other_count++;
2202
2203 my $key = $prefix . $item->{status};
2204 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2205 $other_items->{$key}->{$_} = $item->{$_};
2206 }
2207 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2208 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2209 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2210 $other_items->{$key}->{count}++ if $item->{$hbranch};
2211 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2212 $other_items->{$key}->{description} = $item->{description};
2213 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2214 }
2215 # item is available
2216 else {
2217 $can_place_holds = 1;
2218 $available_count++;
2219 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2220 foreach (qw(branchname itemcallnumber description)) {
2221 $available_items->{$prefix}->{$_} = $item->{$_};
2222 }
2223 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2224 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2225 }
2226 }
2227 } # notforloan, item level and biblioitem level
2228
2229 # if all items are hidden, do not show the record
2230 if ($items_count > 0 && $hideatopac_count == $items_count) {
2231 next;
2232 }
2233
2234 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2235 for my $key ( sort keys %$onloan_items ) {
2236 (++$onloanitemscount > $maxitems) and last;
2237 push @onloan_items_loop, $onloan_items->{$key};
2238 }
2239 for my $key ( sort keys %$other_items ) {
2240 (++$otheritemscount > $maxitems) and last;
2241 push @other_items_loop, $other_items->{$key};
2242 }
2243 for my $key ( sort keys %$available_items ) {
2244 (++$availableitemscount > $maxitems) and last;
2245 push @available_items_loop, $available_items->{$key}
2246 }
2247
2248 # XSLT processing of some stuff
2249 # we fetched the sysprefs already before the loop through all retrieved record!
2250 if (!$scan && $xslfile) {
2251 $record_processor->options({
2252 frameworkcode => $fw,
2253 interface => $search_context->{'interface'}
2254 });
2255
2256 $record_processor->process($marcrecord);
2257 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2258 }
2259
2260 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2261 if (!C4::Context->preference("item-level_itypes")) {
2262 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2263 $can_place_holds = 0;
2264 }
2265 }
2266 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2267 $oldbiblio->{items_count} = $items_count;
2268 $oldbiblio->{available_items_loop} = \@available_items_loop;
2269 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2270 $oldbiblio->{other_items_loop} = \@other_items_loop;
2271 $oldbiblio->{availablecount} = $available_count;
2272 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2273 $oldbiblio->{onloancount} = $onloan_count;
2274 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2275 $oldbiblio->{othercount} = $other_count;
2276 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2277 $oldbiblio->{withdrawncount} = $withdrawn_count;
2278 $oldbiblio->{itemlostcount} = $itemlost_count;
2279 $oldbiblio->{damagedcount} = $itemdamaged_count;
2280 $oldbiblio->{intransitcount} = $item_in_transit_count;
2281 $oldbiblio->{onholdcount} = $item_onhold_count;
2282 $oldbiblio->{orderedcount} = $ordered_count;
2283 $oldbiblio->{notforloancount} = $notforloan_count;
2284
2285 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2286 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2287 my $subfields = substr $fieldspec, 3;
2288 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2289 my @alternateholdingsinfo = ();
2290 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2291 my $alternateholdingscount = 0;
2292
2293 for my $field (@holdingsfields) {
2294 my %holding = ( holding => '' );
2295 my $havesubfield = 0;
2296 for my $subfield ($field->subfields()) {
2297 if ((index $subfields, $$subfield[0]) >= 0) {
2298 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2299 $holding{'holding'} .= $$subfield[1];
2300 $havesubfield++;
2301 }
2302 }
2303 if ($havesubfield) {
2304 push(@alternateholdingsinfo, \%holding);
2305 $alternateholdingscount++;
2306 }
2307 }
2308
2309 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2310 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2311 }
2312
2313 push( @newresults, $oldbiblio );
2314 }
2315
2316 return @newresults;
2317 }
2318
2319 =head2 enabled_staff_search_views
2320
2321 %hash = enabled_staff_search_views()
2322
2323 This function returns a hash that contains three flags obtained from the system
2324 preferences, used to determine whether a particular staff search results view
2325 is enabled.
2326
2327 =over 2
2328
2329 =item C<Output arg:>
2330
2331 * $hash{can_view_MARC} is true only if the MARC view is enabled
2332 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2333 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2334
2335 =item C<usage in the script:>
2336
2337 =back
2338
2339 $template->param ( C4::Search::enabled_staff_search_views );
2340
2341 =cut
2342
2343 sub enabled_staff_search_views
2344 {
2345 return (
2346 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2347 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2348 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2349 );
2350 }
2351
2352 =head2 z3950_search_args
2353
2354 $arrayref = z3950_search_args($matchpoints)
2355
2356 This function returns an array reference that contains the search parameters to be
2357 passed to the Z39.50 search script (z3950_search.pl). The array elements
2358 are hash refs whose keys are name and value, and whose values are the
2359 name of a search parameter, the value of that search parameter and the URL encoded
2360 value of that parameter.
2361
2362 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2363
2364 The search parameter values are obtained from the bibliographic record whose
2365 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2366
2367 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2368 a general purpose search argument. In this case, the returned array contains only
2369 entry: the key is 'title' and the value is derived from $matchpoints.
2370
2371 If a search parameter value is undefined or empty, it is not included in the returned
2372 array.
2373
2374 The returned array reference may be passed directly to the template parameters.
2375
2376 =over 2
2377
2378 =item C<Output arg:>
2379
2380 * $array containing hash refs as described above
2381
2382 =item C<usage in the script:>
2383
2384 =back
2385
2386 $data = Biblio::GetBiblioData($bibno);
2387 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2388
2389 *OR*
2390
2391 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2392
2393 =cut
2394
2395 sub z3950_search_args {
2396 my $bibrec = shift;
2397
2398 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2399 my $isbn = Business::ISBN->new( $isbn_string );
2400
2401 if (defined $isbn && $isbn->is_valid)
2402 {
2403 if ( ref($bibrec) ) {
2404 $bibrec->{isbn} = $isbn_string;
2405 $bibrec->{title} = undef;
2406 } else {
2407 $bibrec = { isbn => $isbn_string };
2408 }
2409 }
2410 else {
2411 $bibrec = { title => $bibrec } if !ref $bibrec;
2412 }
2413 my $array = [];
2414 for my $field (qw/ lccn isbn issn title author dewey subject /)
2415 {
2416 push @$array, { name => $field, value => $bibrec->{$field} }
2417 if defined $bibrec->{$field};
2418 }
2419 return $array;
2420 }
2421
2422 =head2 GetDistinctValues($field);
2423
2424 C<$field> is a reference to the fields array
2425
2426 =cut
2427
2428 sub GetDistinctValues {
2429 my ($fieldname,$string)=@_;
2430 # returns a reference to a hash of references to branches...
2431 if ($fieldname=~/\./){
2432 my ($table,$column)=split /\./, $fieldname;
2433 my $dbh = C4::Context->dbh;
2434 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2435 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 ");
2436 $sth->execute;
2437 my $elements=$sth->fetchall_arrayref({});
2438 return $elements;
2439 }
2440 else {
2441 $string||= qq("");
2442 my @servers=qw<biblioserver authorityserver>;
2443 my (@zconns,@results);
2444 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2445 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2446 $results[$i] =
2447 $zconns[$i]->scan(
2448 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2449 );
2450 }
2451 # The big moment: asynchronously retrieve results from all servers
2452 my @elements;
2453 _ZOOM_event_loop(
2454 \@zconns,
2455 \@results,
2456 sub {
2457 my ( $i, $size ) = @_;
2458 for ( my $j = 0 ; $j < $size ; $j++ ) {
2459 my %hashscan;
2460 @hashscan{qw(value cnt)} =
2461 $results[ $i - 1 ]->display_term($j);
2462 push @elements, \%hashscan;
2463 }
2464 }
2465 );
2466 return \@elements;
2467 }
2468 }
2469
2470 =head2 _ZOOM_event_loop
2471
2472 _ZOOM_event_loop(\@zconns, \@results, sub {
2473 my ( $i, $size ) = @_;
2474 ....
2475 } );
2476
2477 Processes a ZOOM event loop and passes control to a closure for
2478 processing the results, and destroying the resultsets.
2479
2480 =cut
2481
2482 sub _ZOOM_event_loop {
2483 my ($zconns, $results, $callback) = @_;
2484 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2485 my $ev = $zconns->[ $i - 1 ]->last_event();
2486 if ( $ev == ZOOM::Event::ZEND ) {
2487 next unless $results->[ $i - 1 ];
2488 my $size = $results->[ $i - 1 ]->size();
2489 if ( $size > 0 ) {
2490 $callback->($i, $size);
2491 }
2492 }
2493 }
2494
2495 foreach my $result (@$results) {
2496 $result->destroy();
2497 }
2498 }
2499
2500 =head2 new_record_from_zebra
2501
2502 Given raw data from a searchengine result set, return a MARC::Record object
2503
2504 This helper function is needed to take into account all the involved
2505 system preferences and configuration variables to properly create the
2506 MARC::Record object.
2507
2508 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2509 data. If we are using DOM, then it has to be MARCXML.
2510
2511 If we are using elasticsearch, it'll already be a MARC::Record and this
2512 function needs a new name.
2513
2514 =cut
2515
2516 sub new_record_from_zebra {
2517
2518 my $server = shift;
2519 my $raw_data = shift;
2520 # Set the default indexing modes
2521 my $search_engine = C4::Context->preference("SearchEngine");
2522 if ($search_engine eq 'Elasticsearch') {
2523 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2524 }
2525 my $index_mode = ( $server eq 'biblioserver' )
2526 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2527 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2528
2529 my $marc_record = eval {
2530 if ( $index_mode eq 'dom' ) {
2531 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2532 } else {
2533 MARC::Record->new_from_usmarc( $raw_data );
2534 }
2535 };
2536
2537 if ($@) {
2538 return;
2539 } else {
2540 return $marc_record;
2541 }
2542
2543 }
2544
2545 END { } # module clean-up code here (global destructor)
2546
2547 1;
2548 __END__
2549
2550 =head1 AUTHOR
2551
2552 Koha Development Team <http://koha-community.org/>
2553
2554 =cut
Koha ILS