?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 23221: Reindent tools/manage-marc-import.tt
[koha.git] / C4 / Search.pm
blob44b12608c0f60b539c2048737e4db4b7abc54ee7
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, $branches, $itemtypes,
316 $query_type, $scan, $opac
317 );
318
319 The all singing, all dancing, multi-server, asynchronous, scanning,
320 searching, record nabbing, facet-building
321
322 See verbose 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, $branches, $itemtypes,
330 $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 @this_facets_array;
520 for my $one_facet (
521 sort {
522 $facets_counter->{$link_value}
523 ->{$b} <=> $facets_counter->{$link_value}
524 ->{$a}
525 } keys %{ $facets_counter->{$link_value} }
526 )
527 {
528 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
529 my $facet_link_value = $one_facet;
530 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
531
532 # fix the length that will display in the label,
533 my $facet_label_value = $one_facet;
534 my $facet_max_length = C4::Context->preference(
535 'FacetLabelTruncationLength')
536 || 20;
537 $facet_label_value =
538 substr( $one_facet, 0, $facet_max_length )
539 . "..."
540 if length($facet_label_value) >
541 $facet_max_length;
542
543 # if it's a branch, label by the name, not the code,
544 if ( $link_value =~ /branch/ ) {
545 if ( defined $branches
546 && ref($branches) eq "HASH"
547 && defined $branches->{$one_facet}
548 && ref( $branches->{$one_facet} ) eq
549 "HASH" )
550 {
551 $facet_label_value =
552 $branches->{$one_facet}
553 ->{'branchname'};
554 }
555 else {
556 $facet_label_value = "*";
557 }
558 }
559
560 # if it's a itemtype, label by the name, not the code,
561 if ( $link_value =~ /itype/ ) {
562 if ( defined $itemtypes
563 && ref($itemtypes) eq "HASH"
564 && defined $itemtypes->{$one_facet}
565 && ref( $itemtypes->{$one_facet} ) eq
566 "HASH" )
567 {
568 $facet_label_value =
569 $itemtypes->{$one_facet}
570 ->{translated_description};
571 }
572 }
573
574 # also, if it's a location code, use the name instead of the code
575 if ( $link_value =~ /location/ ) {
576 # TODO Retrieve all authorised values at once, instead of 1 query per entry
577 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
578 $facet_label_value = $av->count ? $av->next->opac_description : '';
579 }
580
581 # also, if it's a collection code, use the name instead of the code
582 if ( $link_value =~ /ccode/ ) {
583 # TODO Retrieve all authorised values at once, instead of 1 query per entry
584 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
585 $facet_label_value = $av->count ? $av->next->opac_description : '';
586 }
587
588 # but we're down with the whole label being in the link's title.
589 push @this_facets_array,
590 {
591 facet_count =>
592 $facets_counter->{$link_value}
593 ->{$one_facet},
594 facet_label_value => $facet_label_value,
595 facet_title_value => $one_facet,
596 facet_link_value => $facet_link_value,
597 type_link_value => $link_value,
598 }
599 if ($facet_label_value);
600 }
601
602 push @facets_loop,
603 {
604 type_link_value => $link_value,
605 type_id => $link_value . "_id",
606 "type_label_"
607 . $facets_info->{$link_value}->{'label_value'} =>
608 1,
609 facets => \@this_facets_array,
610 }
611 unless (
612 (
613 $facets_info->{$link_value}->{'label_value'} =~
614 /Libraries/
615 )
616 and ( Koha::Libraries->search->count == 1 )
617 );
618 }
619 }
620 }
621 );
622
623 # This sorts the facets into alphabetical order
624 if (@facets_loop) {
625 foreach my $f (@facets_loop) {
626 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
627 }
628 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
629 }
630
631 return ( undef, $results_hashref, \@facets_loop );
632 }
633
634 sub GetFacets {
635
636 my $rs = shift;
637 my $facets;
638
639 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
640
641 if ( $use_zebra_facets ) {
642 $facets = _get_facets_from_zebra( $rs );
643 } else {
644 $facets = _get_facets_from_records( $rs );
645 }
646
647 return $facets;
648 }
649
650 sub _get_facets_from_records {
651
652 my $rs = shift;
653
654 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
655 my $facets_config = getFacets();
656 my $facets = {};
657 my $size = $rs->size();
658 my $jmax = $size > $facets_maxrecs
659 ? $facets_maxrecs
660 : $size;
661
662 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
663
664 my $marc_record = new_record_from_zebra (
665 'biblioserver',
666 $rs->record( $j )->raw()
667 );
668
669 if ( ! defined $marc_record ) {
670 warn "ERROR DECODING RECORD - $@: " .
671 $rs->record( $j )->raw();
672 next;
673 }
674
675 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
676 }
677
678 return $facets;
679 }
680
681 =head2 _get_facets_data_from_record
682
683 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
684
685 Internal function that extracts facets information from a MARC::Record object
686 and populates $facets_counter for using in getRecords.
687
688 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
689 facets for Zebra).
690
691 =cut
692
693 sub _get_facets_data_from_record {
694
695 my ( $marc_record, $facets, $facets_counter ) = @_;
696
697 for my $facet (@$facets) {
698
699 my @used_datas = ();
700
701 foreach my $tag ( @{ $facet->{ tags } } ) {
702
703 # tag number is the first three digits
704 my $tag_num = substr( $tag, 0, 3 );
705 # subfields are the remainder
706 my $subfield_letters = substr( $tag, 3 );
707
708 my @fields = $marc_record->field( $tag_num );
709 foreach my $field (@fields) {
710 # If $field->indicator(1) eq 'z', it means it is a 'see from'
711 # field introduced because of IncludeSeeFromInSearches, so skip it
712 next if $field->indicator(1) eq 'z';
713
714 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
715
716 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
717 push @used_datas, $data;
718 $facets_counter->{ $facet->{ idx } }->{ $data }++;
719 }
720 }
721 }
722 }
723 }
724
725 =head2 _get_facets_from_zebra
726
727 my $facets = _get_facets_from_zebra( $result_set )
728
729 Retrieves facets for a specified result set. It loops through the facets defined
730 in C4::Koha::getFacets and returns a hash with the following structure:
731
732 { facet_idx => {
733 facet_value => count
734 },
735 ...
736 }
737
738 =cut
739
740 sub _get_facets_from_zebra {
741
742 my $rs = shift;
743
744 # save current elementSetName
745 my $elementSetName = $rs->option( 'elementSetName' );
746
747 my $facets_loop = getFacets();
748 my $facets_data = {};
749 # loop through defined facets and fill the facets hashref
750 foreach my $facet ( @$facets_loop ) {
751
752 my $idx = $facet->{ idx };
753 my $sep = $facet->{ sep };
754 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
755 if ( $facet_values ) {
756 # we've actually got a result
757 $facets_data->{ $idx } = $facet_values;
758 }
759 }
760 # set elementSetName to its previous value to avoid side effects
761 $rs->option( elementSetName => $elementSetName );
762
763 return $facets_data;
764 }
765
766 =head2 _get_facet_from_result_set
767
768 my $facet_values =
769 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
770
771 Internal function that extracts facet information for a specific index ($facet_idx) and
772 returns a hash containing facet values and count:
773
774 {
775 $facet_value => $count ,
776 ...
777 }
778
779 Warning: this function has the side effect of changing the elementSetName for the result
780 set. It is a helper function for the main loop, which takes care of backing it up for
781 restoring.
782
783 =cut
784
785 sub _get_facet_from_result_set {
786
787 my $facet_idx = shift;
788 my $rs = shift;
789 my $sep = shift;
790
791 my $internal_sep = '<*>';
792 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
793
794 return if ( ! defined $facet_idx || ! defined $rs );
795 # zebra's facet element, untokenized index
796 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
797 # configure zebra results for retrieving the desired facet
798 $rs->option( elementSetName => $facet_element );
799 # get the facet record from result set
800 my $facet = $rs->record( 0 )->raw;
801 # if the facet has no restuls...
802 return if !defined $facet;
803 # TODO: benchmark DOM vs. SAX performance
804 my $facet_dom = XML::LibXML->load_xml(
805 string => ($facet)
806 );
807 my @terms = $facet_dom->getElementsByTagName('term');
808 return if ! @terms;
809
810 my $facets = {};
811 foreach my $term ( @terms ) {
812 my $facet_value = $term->textContent;
813 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
814 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
815 }
816
817 return $facets;
818 }
819
820 =head2 _get_facets_info
821
822 my $facets_info = C4::Search::_get_facets_info( $facets )
823
824 Internal function that extracts facets information and properly builds
825 the data structure needed to render facet labels.
826
827 =cut
828
829 sub _get_facets_info {
830
831 my $facets = shift;
832
833 my $facets_info = {};
834
835 for my $facet ( @$facets ) {
836 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
837 }
838
839 return $facets_info;
840 }
841
842 sub pazGetRecords {
843 my (
844 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
845 $results_per_page, $offset, $branches, $query_type,
846 $scan
847 ) = @_;
848 #NOTE: Parameter $branches is not used here !
849
850 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
851 $paz->init();
852 $paz->search($simple_query);
853 sleep 1; # FIXME: WHY?
854
855 # do results
856 my $results_hashref = {};
857 my $stats = XMLin($paz->stat);
858 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
859
860 # for a grouped search result, the number of hits
861 # is the number of groups returned; 'bib_hits' will have
862 # the total number of bibs.
863 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
864 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
865
866 HIT: foreach my $hit (@{ $results->{'hit'} }) {
867 my $recid = $hit->{recid}->[0];
868
869 my $work_title = $hit->{'md-work-title'}->[0];
870 my $work_author;
871 if (exists $hit->{'md-work-author'}) {
872 $work_author = $hit->{'md-work-author'}->[0];
873 }
874 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
875
876 my $result_group = {};
877 $result_group->{'group_label'} = $group_label;
878 $result_group->{'group_merge_key'} = $recid;
879
880 my $count = 1;
881 if (exists $hit->{count}) {
882 $count = $hit->{count}->[0];
883 }
884 $result_group->{'group_count'} = $count;
885
886 for (my $i = 0; $i < $count; $i++) {
887 # FIXME -- may need to worry about diacritics here
888 my $rec = $paz->record($recid, $i);
889 push @{ $result_group->{'RECORDS'} }, $rec;
890 }
891
892 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
893 }
894
895 # pass through facets
896 my $termlist_xml = $paz->termlist('author,subject');
897 my $terms = XMLin($termlist_xml, forcearray => 1);
898 my @facets_loop = ();
899 #die Dumper($results);
900 # foreach my $list (sort keys %{ $terms->{'list'} }) {
901 # my @facets = ();
902 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
903 # push @facets, {
904 # facet_label_value => $facet->{'name'}->[0],
905 # };
906 # }
907 # push @facets_loop, ( {
908 # type_label => $list,
909 # facets => \@facets,
910 # } );
911 # }
912
913 return ( undef, $results_hashref, \@facets_loop );
914 }
915
916 # TRUNCATION
917 sub _detect_truncation {
918 my ( $operand, $index ) = @_;
919 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
920 @regexpr );
921 $operand =~ s/^ //g;
922 my @wordlist = split( /\s/, $operand );
923 foreach my $word (@wordlist) {
924 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
925 push @rightlefttruncated, $word;
926 }
927 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
928 push @lefttruncated, $word;
929 }
930 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
931 push @righttruncated, $word;
932 }
933 elsif ( index( $word, "*" ) < 0 ) {
934 push @nontruncated, $word;
935 }
936 else {
937 push @regexpr, $word;
938 }
939 }
940 return (
941 \@nontruncated, \@righttruncated, \@lefttruncated,
942 \@rightlefttruncated, \@regexpr
943 );
944 }
945
946 # STEMMING
947 sub _build_stemmed_operand {
948 my ($operand,$lang) = @_;
949 require Lingua::Stem::Snowball ;
950 my $stemmed_operand=q{};
951
952 # Stemmer needs language
953 return $operand unless $lang;
954
955 # If operand contains a digit, it is almost certainly an identifier, and should
956 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
957 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
958 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
959 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
960 return $operand if $operand =~ /\d/;
961
962 # FIXME: the locale should be set based on the user's language and/or search choice
963 #warn "$lang";
964 # Make sure we only use the first two letters from the language code
965 $lang = lc(substr($lang, 0, 2));
966 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
967 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
968 if ($lang eq 'nb' || $lang eq 'nn') {
969 $lang = 'no';
970 }
971 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
972 encoding => "UTF-8" );
973
974 my @words = split( / /, $operand );
975 my @stems = $stemmer->stem(\@words);
976 for my $stem (@stems) {
977 $stemmed_operand .= "$stem";
978 $stemmed_operand .= "?"
979 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
980 $stemmed_operand .= " ";
981 }
982 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
983 return $stemmed_operand;
984 }
985
986 # FIELD WEIGHTING
987 sub _build_weighted_query {
988
989 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
990 # pretty well but could work much better if we had a smarter query parser
991 my ( $operand, $stemmed_operand, $index ) = @_;
992 my $stemming = C4::Context->preference("QueryStemming") || 0;
993 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
994 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
995 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
996
997 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
998
999 # Keyword, or, no index specified
1000 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1001 $weighted_query .=
1002 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1003 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1004 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1005 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1006 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1007 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1008 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1009 if $fuzzy_enabled; # add fuzzy, word list
1010 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1011 if ( $stemming and $stemmed_operand )
1012 ; # add stemming, right truncation
1013 $weighted_query .= " or wrdl,r9=\"$operand\"";
1014
1015 # embedded sorting: 0 a-z; 1 z-a
1016 # $weighted_query .= ") or (sort1,aut=1";
1017 }
1018
1019 # Barcode searches should skip this process
1020 elsif ( $index eq 'bc' ) {
1021 $weighted_query .= "bc=\"$operand\"";
1022 }
1023
1024 # Authority-number searches should skip this process
1025 elsif ( $index eq 'an' ) {
1026 $weighted_query .= "an=\"$operand\"";
1027 }
1028
1029 # If the index is numeric, don't autoquote it.
1030 elsif ( $index =~ /,st-numeric$/ ) {
1031 $weighted_query .= " $index=$operand";
1032 }
1033
1034 # If the index already has more than one qualifier, wrap the operand
1035 # in quotes and pass it back (assumption is that the user knows what they
1036 # are doing and won't appreciate us mucking up their query
1037 elsif ( $index =~ ',' ) {
1038 $weighted_query .= " $index=\"$operand\"";
1039 }
1040
1041 #TODO: build better cases based on specific search indexes
1042 else {
1043 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1044 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1045 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1046 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1047 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1048 if $fuzzy_enabled; # add fuzzy, word list
1049 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1050 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1051 }
1052
1053 $weighted_query .= "))"; # close rank specification
1054 return $weighted_query;
1055 }
1056
1057 =head2 getIndexes
1058
1059 Return an array with available indexes.
1060
1061 =cut
1062
1063 sub getIndexes{
1064 my @indexes = (
1065 # biblio indexes
1066 'ab',
1067 'Abstract',
1068 'acqdate',
1069 'allrecords',
1070 'an',
1071 'Any',
1072 'at',
1073 'arl',
1074 'arp',
1075 'au',
1076 'aub',
1077 'aud',
1078 'audience',
1079 'auo',
1080 'aut',
1081 'Author',
1082 'Author-in-order ',
1083 'Author-personal-bibliography',
1084 'Authority-Number',
1085 'authtype',
1086 'bc',
1087 'Bib-level',
1088 'biblionumber',
1089 'bio',
1090 'biography',
1091 'callnum',
1092 'cfn',
1093 'Chronological-subdivision',
1094 'cn-bib-source',
1095 'cn-bib-sort',
1096 'cn-class',
1097 'cn-item',
1098 'cn-prefix',
1099 'cn-suffix',
1100 'cpn',
1101 'Code-institution',
1102 'Conference-name',
1103 'Conference-name-heading',
1104 'Conference-name-see',
1105 'Conference-name-seealso',
1106 'Content-type',
1107 'Control-number',
1108 'copydate',
1109 'Corporate-name',
1110 'Corporate-name-heading',
1111 'Corporate-name-see',
1112 'Corporate-name-seealso',
1113 'Country-publication',
1114 'ctype',
1115 'curriculum',
1116 'date-entered-on-file',
1117 'Date-of-acquisition',
1118 'Date-of-publication',
1119 'Date-time-last-modified',
1120 'Dewey-classification',
1121 'Dissertation-information',
1122 'diss',
1123 'dtlm',
1124 'EAN',
1125 'extent',
1126 'fic',
1127 'fiction',
1128 'Form-subdivision',
1129 'format',
1130 'Geographic-subdivision',
1131 'he',
1132 'Heading',
1133 'Heading-use-main-or-added-entry',
1134 'Heading-use-series-added-entry ',
1135 'Heading-use-subject-added-entry',
1136 'Host-item',
1137 'id-other',
1138 'ident',
1139 'Identifier-standard',
1140 'Illustration-code',
1141 'Index-term-genre',
1142 'Index-term-uncontrolled',
1143 'Interest-age-level',
1144 'Interest-grade-level',
1145 'ISBN',
1146 'isbn',
1147 'ISSN',
1148 'issn',
1149 'itemtype',
1150 'kw',
1151 'Koha-Auth-Number',
1152 'l-format',
1153 'language',
1154 'language-original',
1155 'lc-card',
1156 'LC-card-number',
1157 'lcn',
1158 'lex',
1159 'lexile-number',
1160 'llength',
1161 'ln',
1162 'ln-audio',
1163 'ln-subtitle',
1164 'Local-classification',
1165 'Local-number',
1166 'Match-heading',
1167 'Match-heading-see-from',
1168 'Material-type',
1169 'mc-itemtype',
1170 'mc-rtype',
1171 'mus',
1172 'name',
1173 'Music-number',
1174 'Name-geographic',
1175 'Name-geographic-heading',
1176 'Name-geographic-see',
1177 'Name-geographic-seealso',
1178 'nb',
1179 'Note',
1180 'notes',
1181 'ns',
1182 'nt',
1183 'Other-control-number',
1184 'pb',
1185 'Personal-name',
1186 'Personal-name-heading',
1187 'Personal-name-see',
1188 'Personal-name-seealso',
1189 'pl',
1190 'Place-publication',
1191 'pn',
1192 'popularity',
1193 'pubdate',
1194 'Publisher',
1195 'Provider',
1196 'pv',
1197 'Reading-grade-level',
1198 'Record-control-number',
1199 'rcn',
1200 'Record-type',
1201 'rtype',
1202 'se',
1203 'See',
1204 'See-also',
1205 'sn',
1206 'Stock-number',
1207 'su',
1208 'Subject',
1209 'Subject-heading-thesaurus',
1210 'Subject-name-personal',
1211 'Subject-subdivision',
1212 'Summary',
1213 'Suppress',
1214 'su-geo',
1215 'su-na',
1216 'su-to',
1217 'su-ut',
1218 'ut',
1219 'Term-genre-form',
1220 'Term-genre-form-heading',
1221 'Term-genre-form-see',
1222 'Term-genre-form-seealso',
1223 'ti',
1224 'Title',
1225 'Title-cover',
1226 'Title-series',
1227 'Title-uniform',
1228 'Title-uniform-heading',
1229 'Title-uniform-see',
1230 'Title-uniform-seealso',
1231 'totalissues',
1232 'yr',
1233
1234 # items indexes
1235 'acqsource',
1236 'barcode',
1237 'bc',
1238 'branch',
1239 'ccode',
1240 'classification-source',
1241 'cn-sort',
1242 'coded-location-qualifier',
1243 'copynumber',
1244 'damaged',
1245 'datelastborrowed',
1246 'datelastseen',
1247 'holdingbranch',
1248 'homebranch',
1249 'issues',
1250 'item',
1251 'itemnumber',
1252 'itype',
1253 'Local-classification',
1254 'location',
1255 'lost',
1256 'materials-specified',
1257 'mc-ccode',
1258 'mc-itype',
1259 'mc-loc',
1260 'notforloan',
1261 'Number-local-acquisition',
1262 'onloan',
1263 'price',
1264 'renewals',
1265 'replacementprice',
1266 'replacementpricedate',
1267 'reserves',
1268 'restricted',
1269 'stack',
1270 'stocknumber',
1271 'inv',
1272 'uri',
1273 'withdrawn',
1274
1275 # subject related
1276 );
1277
1278 return \@indexes;
1279 }
1280
1281 =head2 _handle_exploding_index
1282
1283 my $query = _handle_exploding_index($index, $term)
1284
1285 Callback routine to generate the search for "exploding" indexes (i.e.
1286 those indexes which are turned into multiple or-connected searches based
1287 on authority data).
1288
1289 =cut
1290
1291 sub _handle_exploding_index {
1292 my ($QParser, $filter, $params, $negate, $server) = @_;
1293 my $index = $filter;
1294 my $term = join(' ', @$params);
1295
1296 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1297
1298 my $marcflavour = C4::Context->preference('marcflavour');
1299
1300 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1301 my $wantedcodes = '';
1302 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1303 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1304 foreach my $auth (@$results) {
1305 my $record = MARC::Record->new_from_usmarc($auth);
1306 my @references = $record->field('5..');
1307 if (@references) {
1308 if ($index eq 'su-br') {
1309 $wantedcodes = 'g';
1310 } elsif ($index eq 'su-na') {
1311 $wantedcodes = 'h';
1312 } elsif ($index eq 'su-rl') {
1313 $wantedcodes = '';
1314 }
1315 foreach my $reference (@references) {
1316 my $codes = $reference->subfield($codesubfield);
1317 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1318 }
1319 }
1320 }
1321 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1322 return $query;
1323 }
1324
1325 =head2 parseQuery
1326
1327 ( $operators, $operands, $indexes, $limits,
1328 $sort_by, $scan, $lang ) =
1329 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1330
1331 Shim function to ease the transition from buildQuery to a new QueryParser.
1332 This function is called at the beginning of buildQuery, and modifies
1333 buildQuery's input. If it can handle the input, it returns a query that
1334 buildQuery will not try to parse.
1335
1336 =cut
1337
1338 sub parseQuery {
1339 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1340
1341 my @operators = $operators ? @$operators : ();
1342 my @indexes = $indexes ? @$indexes : ();
1343 my @operands = $operands ? @$operands : ();
1344 my @limits = $limits ? @$limits : ();
1345 my @sort_by = $sort_by ? @$sort_by : ();
1346
1347 my $query = $operands[0];
1348 my $index;
1349 my $term;
1350 my $query_desc;
1351
1352 my $QParser;
1353 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1354 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1355 undef $QParser if (scalar @limits > 0);
1356
1357 if ($QParser)
1358 {
1359 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1360 $query = '';
1361 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1362 next unless $operands[$ii];
1363 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1364 if ($query);
1365 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1366 $query .= $operands[$ii];
1367 }
1368 elsif ( $indexes[$ii] =~ m/su-/ ) {
1369 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1370 }
1371 else {
1372 $query .=
1373 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1374 }
1375 }
1376 foreach my $limit (@limits) {
1377 }
1378 if ( scalar(@sort_by) > 0 ) {
1379 my $modifier_re =
1380 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1381 $query =~ s/$modifier_re//g;
1382 foreach my $modifier (@sort_by) {
1383 $query .= " #$modifier";
1384 }
1385 }
1386
1387 $query_desc = $query;
1388 $query_desc =~ s/\s+/ /g;
1389 if ( C4::Context->preference("QueryWeightFields") ) {
1390 }
1391 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1392 { 'target_syntax_callback' => \&_handle_exploding_index } );
1393 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1394 { 'target_syntax_callback' => \&_handle_exploding_index } );
1395 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1396 { 'target_syntax_callback' => \&_handle_exploding_index } );
1397 $QParser->parse($query);
1398 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1399 }
1400 else {
1401 require Koha::QueryParser::Driver::PQF;
1402 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1403 s/$modifier_re//g for @operands;
1404 }
1405
1406 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1407 }
1408
1409 =head2 buildQuery
1410
1411 ( $error, $query,
1412 $simple_query, $query_cgi,
1413 $query_desc, $limit,
1414 $limit_cgi, $limit_desc,
1415 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1416
1417 Build queries and limits in CCL, CGI, Human,
1418 handle truncation, stemming, field weighting, fuzziness, etc.
1419
1420 See verbose embedded documentation.
1421
1422
1423 =cut
1424
1425 sub buildQuery {
1426 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1427
1428 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1429
1430 my $query_desc;
1431 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1432
1433 # dereference
1434 my @operators = $operators ? @$operators : ();
1435 my @indexes = $indexes ? @$indexes : ();
1436 my @operands = $operands ? @$operands : ();
1437 my @limits = $limits ? @$limits : ();
1438 my @sort_by = $sort_by ? @$sort_by : ();
1439
1440 my $stemming = C4::Context->preference("QueryStemming") || 0;
1441 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1442 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1443 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1444
1445 my $query = $operands[0];
1446 my $simple_query = $operands[0];
1447
1448 # initialize the variables we're passing back
1449 my $query_cgi;
1450 my $query_type;
1451
1452 my $limit;
1453 my $limit_cgi;
1454 my $limit_desc;
1455
1456 my $cclq = 0;
1457 my $cclindexes = getIndexes();
1458 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1459 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1460 my $dx = lc($1);
1461 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1462 }
1463 $query = "ccl=$query" if $cclq;
1464 }
1465
1466 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1467 # DIAGNOSTIC ONLY!!
1468 if ( $query =~ /^ccl=/ ) {
1469 my $q=$';
1470 # This is needed otherwise ccl= and &limit won't work together, and
1471 # this happens when selecting a subject on the opac-detail page
1472 @limits = grep {!/^$/} @limits;
1473 my $original_q = $q; # without available part
1474 unless ( grep { /^available$/ } @limits ) {
1475 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1476 $original_q = $q;
1477 }
1478 if ( @limits ) {
1479 if ( grep { /^available$/ } @limits ) {
1480 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1481 delete $limits['available'];
1482 }
1483 $q .= ' and '.join(' and ', @limits) if @limits;
1484 }
1485 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1486 }
1487 if ( $query =~ /^cql=/ ) {
1488 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1489 }
1490 if ( $query =~ /^pqf=/ ) {
1491 if ($query_desc) {
1492 $query_cgi = "q=".uri_escape_utf8($query_desc);
1493 } else {
1494 $query_desc = $';
1495 $query_cgi = "q=pqf=".uri_escape_utf8($');
1496 }
1497 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1498 }
1499
1500 # pass nested queries directly
1501 # FIXME: need better handling of some of these variables in this case
1502 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1503 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1504 # if ( $query =~ /(\(|\))/ ) {
1505 # return (
1506 # undef, $query, $simple_query, $query_cgi,
1507 # $query, $limit, $limit_cgi, $limit_desc,
1508 # 'ccl'
1509 # );
1510 # }
1511
1512 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1513 # query operands and indexes and add stemming, truncation, field weighting, etc.
1514 # Once we do so, we'll end up with a value in $query, just like if we had an
1515 # incoming $query from the user
1516 else {
1517 $query = ""
1518 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1519 my $previous_operand
1520 ; # a flag used to keep track if there was a previous query
1521 # if there was, we can apply the current operator
1522 # for every operand
1523 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1524
1525 # COMBINE OPERANDS, INDEXES AND OPERATORS
1526 if ( ($operands[$i] // '') ne '' ) {
1527 $operands[$i]=~s/^\s+//;
1528
1529 # A flag to determine whether or not to add the index to the query
1530 my $indexes_set;
1531
1532 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1533 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1534 $weight_fields = 0;
1535 $stemming = 0;
1536 } else {
1537 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1538 }
1539 my $operand = $operands[$i];
1540 my $index = $indexes[$i];
1541
1542 # Add index-specific attributes
1543
1544 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1545 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1546 #which is processed higher up in this sub. Other than that, year searches are typically
1547 #handled as limits which are not processed her either.
1548
1549 # Search ranges: Date of Publication, st-numeric
1550 if ( $index =~ /(yr|st-numeric)/ ) {
1551 #weight_fields/relevance search causes errors with date ranges
1552 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1553 #In the case of YYYY-YYYY, it will return no results
1554 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1555 }
1556
1557 # Date of Acquisition
1558 elsif ( $index =~ /acqdate/ ) {
1559 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1560 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1561 #top of the results just because they have lots of item records matching that date.
1562 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1563 #irrelevant here
1564 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1565 }
1566 # ISBN,ISSN,Standard Number, don't need special treatment
1567 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1568 (
1569 $stemming, $auto_truncation,
1570 $weight_fields, $fuzzy_enabled
1571 ) = ( 0, 0, 0, 0 );
1572
1573 if ( $index eq 'nb' ) {
1574 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1575 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1576 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1577 $indexes[$i] = $index = '';
1578 }
1579 }
1580 }
1581
1582 if(not $index){
1583 $index = 'kw';
1584 }
1585
1586 # Set default structure attribute (word list)
1587 my $struct_attr = q{};
1588 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1589 $struct_attr = ",wrdl";
1590 }
1591
1592 # Some helpful index variants
1593 my $index_plus = $index . $struct_attr . ':';
1594 my $index_plus_comma = $index . $struct_attr . ',';
1595
1596 if ($auto_truncation){
1597 unless ( $index =~ /,(st-|phr|ext)/ ) {
1598 #FIXME only valid with LTR scripts
1599 $operand=join(" ",map{
1600 (index($_,"*")>0?"$_":"$_*")
1601 }split (/\s+/,$operand));
1602 warn $operand if $DEBUG;
1603 }
1604 }
1605
1606 # Detect Truncation
1607 my $truncated_operand;
1608 my( $nontruncated, $righttruncated, $lefttruncated,
1609 $rightlefttruncated, $regexpr
1610 ) = _detect_truncation( $operand, $index );
1611 warn
1612 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1613 if $DEBUG;
1614
1615 # Apply Truncation
1616 if (
1617 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1618 scalar(@$rightlefttruncated) > 0 )
1619 {
1620
1621 # Don't field weight or add the index to the query, we do it here
1622 $indexes_set = 1;
1623 undef $weight_fields;
1624 my $previous_truncation_operand;
1625 if (scalar @$nontruncated) {
1626 $truncated_operand .= "$index_plus @$nontruncated ";
1627 $previous_truncation_operand = 1;
1628 }
1629 if (scalar @$righttruncated) {
1630 $truncated_operand .= "and " if $previous_truncation_operand;
1631 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1632 $previous_truncation_operand = 1;
1633 }
1634 if (scalar @$lefttruncated) {
1635 $truncated_operand .= "and " if $previous_truncation_operand;
1636 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1637 $previous_truncation_operand = 1;
1638 }
1639 if (scalar @$rightlefttruncated) {
1640 $truncated_operand .= "and " if $previous_truncation_operand;
1641 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1642 $previous_truncation_operand = 1;
1643 }
1644 }
1645 $operand = $truncated_operand if $truncated_operand;
1646 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1647
1648 # Handle Stemming
1649 my $stemmed_operand;
1650 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1651 if $stemming;
1652
1653 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1654
1655 # Handle Field Weighting
1656 my $weighted_operand;
1657 if ($weight_fields) {
1658 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1659 $operand = $weighted_operand;
1660 $indexes_set = 1;
1661 }
1662
1663 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1664
1665 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1666 query => $query,
1667 query_cgi => $query_cgi,
1668 query_desc => $query_desc,
1669 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1670 parsed_operand => $operand,
1671 original_operand => $operands[$i] // '',
1672 index => $index,
1673 index_plus => $index_plus,
1674 indexes_set => $indexes_set,
1675 previous_operand => $previous_operand,
1676 });
1677
1678 } #/if $operands
1679 } # /for
1680 }
1681 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1682
1683 # add limits
1684 my %group_OR_limits;
1685 my $availability_limit;
1686 foreach my $this_limit (@limits) {
1687 next unless $this_limit;
1688 if ( $this_limit =~ /available/ ) {
1689 #
1690 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1691 ## In English:
1692 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1693 $availability_limit .=
1694 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1695 $limit_cgi .= "&limit=available";
1696 $limit_desc .= "";
1697 }
1698
1699 # group_OR_limits, prefixed by mc-
1700 # OR every member of the group
1701 elsif ( $this_limit =~ /mc/ ) {
1702 my ($k,$v) = split(/:/, $this_limit,2);
1703 if ( $k !~ /mc-i(tem)?type/ ) {
1704 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1705 $this_limit =~ tr/"//d;
1706 $this_limit = $k.":'".$v."'";
1707 }
1708
1709 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1710 $limit_desc .= " or " if $group_OR_limits{$k};
1711 $group_OR_limits{$k} .= "$this_limit";
1712 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1713 $limit_desc .= " $this_limit";
1714 }
1715
1716 # Regular old limits
1717 else {
1718 $limit .= " and " if $limit || $query;
1719 $limit .= "$this_limit";
1720 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1721 if ($this_limit =~ /^branch:(.+)/) {
1722 my $branchcode = $1;
1723 my $library = Koha::Libraries->find( $branchcode );
1724 if (defined $library) {
1725 $limit_desc .= " branch:" . $library->branchname;
1726 } else {
1727 $limit_desc .= " $this_limit";
1728 }
1729 } else {
1730 $limit_desc .= " $this_limit";
1731 }
1732 }
1733 }
1734 foreach my $k (keys (%group_OR_limits)) {
1735 $limit .= " and " if ( $query || $limit );
1736 $limit .= "($group_OR_limits{$k})";
1737 }
1738 if ($availability_limit) {
1739 $limit .= " and " if ( $query || $limit );
1740 $limit .= "($availability_limit)";
1741 }
1742
1743 # Normalize the query and limit strings
1744 # This is flawed , means we can't search anything with : in it
1745 # if user wants to do ccl or cql, start the query with that
1746 # $query =~ s/:/=/g;
1747 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1748 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1749 $query =~ s/(?<=(wrdl)):/=/g;
1750 $query =~ s/(?<=(trn|phr)):/=/g;
1751 $query =~ s/(?<=(st-numeric)):/=/g;
1752 $query =~ s/(?<=(st-year)):/=/g;
1753 $query =~ s/(?<=(st-date-normalized)):/=/g;
1754 $limit =~ s/:/=/g;
1755 for ( $query, $query_desc, $limit, $limit_desc ) {
1756 s/ +/ /g; # remove extra spaces
1757 s/^ //g; # remove any beginning spaces
1758 s/ $//g; # remove any ending spaces
1759 s/==/=/g; # remove double == from query
1760 }
1761 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1762
1763 for ($query_cgi,$simple_query) {
1764 s/"//g;
1765 }
1766 # append the limit to the query
1767 $query .= " " . $limit;
1768
1769 # Warnings if DEBUG
1770 if ($DEBUG) {
1771 warn "QUERY:" . $query;
1772 warn "QUERY CGI:" . $query_cgi;
1773 warn "QUERY DESC:" . $query_desc;
1774 warn "LIMIT:" . $limit;
1775 warn "LIMIT CGI:" . $limit_cgi;
1776 warn "LIMIT DESC:" . $limit_desc;
1777 warn "---------\nLeave buildQuery\n---------";
1778 }
1779
1780 return (
1781 undef, $query, $simple_query, $query_cgi,
1782 $query_desc, $limit, $limit_cgi, $limit_desc,
1783 $query_type
1784 );
1785 }
1786
1787 =head2 _build_initial_query
1788
1789 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1790
1791 Build a section of the initial query containing indexes, operators, and operands.
1792
1793 =cut
1794
1795 sub _build_initial_query {
1796 my ($params) = @_;
1797
1798 my $operator = "";
1799 if ($params->{previous_operand}){
1800 #If there is a previous operand, add a supplied operator or the default 'and'
1801 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1802 }
1803
1804 #NOTE: indexes_set is typically set when doing truncation or field weighting
1805 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1806
1807 #e.g. "kw,wrdl:test"
1808 #e.g. " and kw,wrdl:test"
1809 $params->{query} .= $operator . $operand;
1810
1811 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1812 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1813 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1814
1815 #e.g. " and kw,wrdl: test"
1816 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1817
1818 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1819
1820 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1821 }
1822
1823 =head2 searchResults
1824
1825 my @search_results = searchResults($search_context, $searchdesc, $hits,
1826 $results_per_page, $offset, $scan,
1827 @marcresults);
1828
1829 Format results in a form suitable for passing to the template
1830
1831 =cut
1832
1833 # IMO this subroutine is pretty messy still -- it's responsible for
1834 # building the HTML output for the template
1835 sub searchResults {
1836 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1837 my $dbh = C4::Context->dbh;
1838 my @newresults;
1839
1840 require C4::Items;
1841
1842 $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet';
1843 my ($is_opac, $hidelostitems);
1844 if ($search_context->{'interface'} eq 'opac') {
1845 $hidelostitems = C4::Context->preference('hidelostitems');
1846 $is_opac = 1;
1847 }
1848
1849 #Build branchnames hash
1850 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1851
1852 # FIXME - We build an authorised values hash here, using the default framework
1853 # though it is possible to have different authvals for different fws.
1854
1855 my $shelflocations =
1856 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1857
1858 # get notforloan authorised value list (see $shelflocations FIXME)
1859 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1860 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1861
1862 #Get itemtype hash
1863 my $itemtypes = Koha::ItemTypes->search_with_localization;
1864 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1865
1866 #search item field code
1867 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1868
1869 ## find column names of items related to MARC
1870 my %subfieldstosearch;
1871 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1872 for my $column ( @columns ) {
1873 my ( $tagfield, $tagsubfield ) =
1874 &GetMarcFromKohaField( "items." . $column, "" );
1875 if ( defined $tagsubfield ) {
1876 $subfieldstosearch{$column} = $tagsubfield;
1877 }
1878 }
1879
1880 # handle which records to actually retrieve
1881 my $times;
1882 if ( $hits && $offset + $results_per_page <= $hits ) {
1883 $times = $offset + $results_per_page;
1884 }
1885 else {
1886 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1887 }
1888
1889 my $marcflavour = C4::Context->preference("marcflavour");
1890 # We get the biblionumber position in MARC
1891 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1892
1893 # set stuff for XSLT processing here once, not later again for every record we retrieved
1894 my $xslfile;
1895 my $xslsyspref;
1896 if( $is_opac ){
1897 $xslsyspref = "OPACXSLTResultsDisplay";
1898 $xslfile = C4::Context->preference( $xslsyspref );
1899 } else {
1900 $xslsyspref = "XSLTResultsDisplay";
1901 $xslfile = C4::Context->preference( $xslsyspref ) || "default";
1902 }
1903 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1904 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1905
1906 my $userenv = C4::Context->userenv;
1907 my $logged_in_user
1908 = ( defined $userenv and $userenv->{number} )
1909 ? Koha::Patrons->find( $userenv->{number} )
1910 : undef;
1911 my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0;
1912
1913 # loop through all of the records we've retrieved
1914 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1915
1916 my $marcrecord;
1917 if ($scan) {
1918 # For Scan searches we built USMARC data
1919 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1920 } else {
1921 # Normal search, render from Zebra's output
1922 $marcrecord = new_record_from_zebra(
1923 'biblioserver',
1924 $marcresults->[$i]
1925 );
1926
1927 if ( ! defined $marcrecord ) {
1928 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1929 next;
1930 }
1931 }
1932
1933 my $fw = $scan
1934 ? undef
1935 : $bibliotag < 10
1936 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1937 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1938
1939 SetUTF8Flag($marcrecord);
1940 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1941 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1942 $oldbiblio->{result_number} = $i + 1;
1943
1944 # add imageurl to itemtype if there is one
1945 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1946
1947 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1948 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1949 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1950 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1951 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1952
1953 # edition information, if any
1954 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1955 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1956 # Build summary if there is one (the summary is defined in the itemtypes table)
1957 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1958 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1959 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1960 my @fields = $marcrecord->fields();
1961
1962 my $newsummary;
1963 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1964 my $tags = {};
1965 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1966 $tag =~ /(.{3})(.)/;
1967 if($marcrecord->field($1)){
1968 my @abc = $marcrecord->field($1)->subfield($2);
1969 $tags->{$tag} = $#abc + 1 ;
1970 }
1971 }
1972
1973 # We catch how many times to repeat this line
1974 my $max = 0;
1975 foreach my $tag (keys(%$tags)){
1976 $max = $tags->{$tag} if($tags->{$tag} > $max);
1977 }
1978
1979 # we replace, and repeat each line
1980 for (my $i = 0 ; $i < $max ; $i++){
1981 my $newline = $line;
1982
1983 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1984 $tag =~ /(.{3})(.)/;
1985
1986 if($marcrecord->field($1)){
1987 my @repl = $marcrecord->field($1)->subfield($2);
1988 my $subfieldvalue = $repl[$i];
1989 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1990 }
1991 }
1992 $newsummary .= "$newline\n";
1993 }
1994 }
1995
1996 $newsummary =~ s/\[(.*?)]//g;
1997 $newsummary =~ s/\n/<br\/>/g;
1998 $oldbiblio->{summary} = $newsummary;
1999 }
2000
2001 # Pull out the items fields
2002 my @fields = $marcrecord->field($itemtag);
2003 my $marcflavor = C4::Context->preference("marcflavour");
2004
2005 # adding linked items that belong to host records
2006 if ( C4::Context->preference('EasyAnalyticalRecords') ) {
2007 my $analyticsfield = '773';
2008 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2009 $analyticsfield = '773';
2010 } elsif ($marcflavor eq 'UNIMARC') {
2011 $analyticsfield = '461';
2012 }
2013 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2014 my $hostbiblionumber = $hostfield->subfield("0");
2015 my $linkeditemnumber = $hostfield->subfield("9");
2016 if( $hostbiblionumber ) {
2017 my $hostbiblio = GetMarcBiblio({
2018 biblionumber => $hostbiblionumber,
2019 embed_items => 1 });
2020 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2021 if( $hostbiblio ) {
2022 my @hostitems = $hostbiblio->field($itemfield);
2023 foreach my $hostitem (@hostitems){
2024 if ($hostitem->subfield("9") eq $linkeditemnumber){
2025 my $linkeditem =$hostitem;
2026 # append linked items if they exist
2027 push @fields, $linkeditem if $linkeditem;
2028 }
2029 }
2030 }
2031 }
2032 }
2033 }
2034
2035 # Setting item statuses for display
2036 my @available_items_loop;
2037 my @onloan_items_loop;
2038 my @other_items_loop;
2039
2040 my $available_items;
2041 my $onloan_items;
2042 my $other_items;
2043
2044 my $ordered_count = 0;
2045 my $available_count = 0;
2046 my $onloan_count = 0;
2047 my $longoverdue_count = 0;
2048 my $other_count = 0;
2049 my $withdrawn_count = 0;
2050 my $itemlost_count = 0;
2051 my $hideatopac_count = 0;
2052 my $itembinding_count = 0;
2053 my $itemdamaged_count = 0;
2054 my $item_in_transit_count = 0;
2055 my $can_place_holds = 0;
2056 my $item_onhold_count = 0;
2057 my $notforloan_count = 0;
2058 my $items_count = scalar(@fields);
2059 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2060 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2061 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2062
2063 # loop through every item
2064 foreach my $field (@fields) {
2065 my $item;
2066
2067 # populate the items hash
2068 foreach my $code ( keys %subfieldstosearch ) {
2069 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2070 }
2071 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2072
2073 # OPAC hidden items
2074 if ($is_opac) {
2075 # hidden because lost
2076 if ($hidelostitems && $item->{itemlost}) {
2077 $hideatopac_count++;
2078 next;
2079 }
2080 # hidden based on OpacHiddenItems syspref
2081 my @hi = C4::Items::GetHiddenItemnumbers({ items=> [ $item ], borcat => $search_context->{category} });
2082 if (scalar @hi) {
2083 push @hiddenitems, @hi;
2084 $hideatopac_count++;
2085 next;
2086 }
2087 }
2088
2089 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2090 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2091
2092 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2093 if ($item->{$hbranch}) {
2094 $item->{'branchname'} = $branches{$item->{$hbranch}};
2095 }
2096 elsif ($item->{$otherbranch}) { # Last resort
2097 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2098 }
2099
2100 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2101 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2102 if ( $item->{onloan}
2103 and $logged_in_user
2104 and !( $patron_category_hide_lost_items and $item->{itemlost} ) )
2105 {
2106 $onloan_count++;
2107 my $key = $prefix . $item->{onloan} . $item->{barcode};
2108 $onloan_items->{$key}->{due_date} = $item->{onloan};
2109 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2110 $onloan_items->{$key}->{branchname} = $item->{branchname};
2111 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2112 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2113 $onloan_items->{$key}->{description} = $item->{description};
2114 $onloan_items->{$key}->{imageurl} =
2115 getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2116
2117 # if something's checked out and lost, mark it as 'long overdue'
2118 if ( $item->{itemlost} ) {
2119 $onloan_items->{$key}->{longoverdue}++;
2120 $longoverdue_count++;
2121 }
2122 else { # can place holds as long as item isn't lost
2123 $can_place_holds = 1;
2124 }
2125 }
2126
2127 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2128 else {
2129
2130 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2131
2132 # item is on order
2133 if ( $item->{notforloan} < 0 ) {
2134 $ordered_count++;
2135 } elsif ( $item->{notforloan} > 0 ) {
2136 $notforloan_count++;
2137 }
2138
2139 # is item in transit?
2140 my $transfertwhen = '';
2141 my ($transfertfrom, $transfertto);
2142
2143 # is item on the reserve shelf?
2144 my $reservestatus = '';
2145
2146 unless ($item->{withdrawn}
2147 || $item->{itemlost}
2148 || $item->{damaged}
2149 || $item->{notforloan}
2150 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2151 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2152
2153 # A couple heuristics to limit how many times
2154 # we query the database for item transfer information, sacrificing
2155 # accuracy in some cases for speed;
2156 #
2157 # 1. don't query if item has one of the other statuses
2158 # 2. don't check transit status if the bib has
2159 # more than 20 items
2160 #
2161 # FIXME: to avoid having the query the database like this, and to make
2162 # the in transit status count as unavailable for search limiting,
2163 # should map transit status to record indexed in Zebra.
2164 #
2165 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2166 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2167 }
2168
2169 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2170 if ( $item->{withdrawn}
2171 || $item->{itemlost}
2172 || $item->{damaged}
2173 || $item->{notforloan}
2174 || $reservestatus eq 'Waiting'
2175 || ($transfertwhen ne ''))
2176 {
2177 $withdrawn_count++ if $item->{withdrawn};
2178 $itemlost_count++ if $item->{itemlost};
2179 $itemdamaged_count++ if $item->{damaged};
2180 $item_in_transit_count++ if $transfertwhen ne '';
2181 $item_onhold_count++ if $reservestatus eq 'Waiting';
2182 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2183
2184 # can place a hold on a item if
2185 # not lost nor withdrawn
2186 # not damaged unless AllowHoldsOnDamagedItems is true
2187 # item is either for loan or on order (notforloan < 0)
2188 $can_place_holds = 1
2189 if (
2190 !$item->{itemlost}
2191 && !$item->{withdrawn}
2192 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2193 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2194 );
2195
2196 $other_count++;
2197
2198 my $key = $prefix . $item->{status};
2199 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2200 $other_items->{$key}->{$_} = $item->{$_};
2201 }
2202 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2203 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2204 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2205 $other_items->{$key}->{count}++ if $item->{$hbranch};
2206 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2207 $other_items->{$key}->{description} = $item->{description};
2208 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2209 }
2210 # item is available
2211 else {
2212 $can_place_holds = 1;
2213 $available_count++;
2214 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2215 foreach (qw(branchname itemcallnumber description)) {
2216 $available_items->{$prefix}->{$_} = $item->{$_};
2217 }
2218 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2219 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} );
2220 }
2221 }
2222 } # notforloan, item level and biblioitem level
2223
2224 # if all items are hidden, do not show the record
2225 if ($items_count > 0 && $hideatopac_count == $items_count) {
2226 next;
2227 }
2228
2229 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2230 for my $key ( sort keys %$onloan_items ) {
2231 (++$onloanitemscount > $maxitems) and last;
2232 push @onloan_items_loop, $onloan_items->{$key};
2233 }
2234 for my $key ( sort keys %$other_items ) {
2235 (++$otheritemscount > $maxitems) and last;
2236 push @other_items_loop, $other_items->{$key};
2237 }
2238 for my $key ( sort keys %$available_items ) {
2239 (++$availableitemscount > $maxitems) and last;
2240 push @available_items_loop, $available_items->{$key}
2241 }
2242
2243 # XSLT processing of some stuff
2244 # we fetched the sysprefs already before the loop through all retrieved record!
2245 if (!$scan && $xslfile) {
2246 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2247 }
2248
2249 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2250 if (!C4::Context->preference("item-level_itypes")) {
2251 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2252 $can_place_holds = 0;
2253 }
2254 }
2255 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2256 $oldbiblio->{items_count} = $items_count;
2257 $oldbiblio->{available_items_loop} = \@available_items_loop;
2258 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2259 $oldbiblio->{other_items_loop} = \@other_items_loop;
2260 $oldbiblio->{availablecount} = $available_count;
2261 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2262 $oldbiblio->{onloancount} = $onloan_count;
2263 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2264 $oldbiblio->{othercount} = $other_count;
2265 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2266 $oldbiblio->{withdrawncount} = $withdrawn_count;
2267 $oldbiblio->{itemlostcount} = $itemlost_count;
2268 $oldbiblio->{damagedcount} = $itemdamaged_count;
2269 $oldbiblio->{intransitcount} = $item_in_transit_count;
2270 $oldbiblio->{onholdcount} = $item_onhold_count;
2271 $oldbiblio->{orderedcount} = $ordered_count;
2272 $oldbiblio->{notforloancount} = $notforloan_count;
2273
2274 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2275 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2276 my $subfields = substr $fieldspec, 3;
2277 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2278 my @alternateholdingsinfo = ();
2279 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2280 my $alternateholdingscount = 0;
2281
2282 for my $field (@holdingsfields) {
2283 my %holding = ( holding => '' );
2284 my $havesubfield = 0;
2285 for my $subfield ($field->subfields()) {
2286 if ((index $subfields, $$subfield[0]) >= 0) {
2287 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2288 $holding{'holding'} .= $$subfield[1];
2289 $havesubfield++;
2290 }
2291 }
2292 if ($havesubfield) {
2293 push(@alternateholdingsinfo, \%holding);
2294 $alternateholdingscount++;
2295 }
2296 }
2297
2298 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2299 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2300 }
2301
2302 push( @newresults, $oldbiblio );
2303 }
2304
2305 return @newresults;
2306 }
2307
2308 =head2 enabled_staff_search_views
2309
2310 %hash = enabled_staff_search_views()
2311
2312 This function returns a hash that contains three flags obtained from the system
2313 preferences, used to determine whether a particular staff search results view
2314 is enabled.
2315
2316 =over 2
2317
2318 =item C<Output arg:>
2319
2320 * $hash{can_view_MARC} is true only if the MARC view is enabled
2321 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2322 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2323
2324 =item C<usage in the script:>
2325
2326 =back
2327
2328 $template->param ( C4::Search::enabled_staff_search_views );
2329
2330 =cut
2331
2332 sub enabled_staff_search_views
2333 {
2334 return (
2335 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2336 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2337 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2338 );
2339 }
2340
2341 =head2 z3950_search_args
2342
2343 $arrayref = z3950_search_args($matchpoints)
2344
2345 This function returns an array reference that contains the search parameters to be
2346 passed to the Z39.50 search script (z3950_search.pl). The array elements
2347 are hash refs whose keys are name and value, and whose values are the
2348 name of a search parameter, the value of that search parameter and the URL encoded
2349 value of that parameter.
2350
2351 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2352
2353 The search parameter values are obtained from the bibliographic record whose
2354 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2355
2356 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2357 a general purpose search argument. In this case, the returned array contains only
2358 entry: the key is 'title' and the value is derived from $matchpoints.
2359
2360 If a search parameter value is undefined or empty, it is not included in the returned
2361 array.
2362
2363 The returned array reference may be passed directly to the template parameters.
2364
2365 =over 2
2366
2367 =item C<Output arg:>
2368
2369 * $array containing hash refs as described above
2370
2371 =item C<usage in the script:>
2372
2373 =back
2374
2375 $data = Biblio::GetBiblioData($bibno);
2376 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2377
2378 *OR*
2379
2380 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2381
2382 =cut
2383
2384 sub z3950_search_args {
2385 my $bibrec = shift;
2386
2387 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2388 my $isbn = Business::ISBN->new( $isbn_string );
2389
2390 if (defined $isbn && $isbn->is_valid)
2391 {
2392 if ( ref($bibrec) ) {
2393 $bibrec->{isbn} = $isbn_string;
2394 $bibrec->{title} = undef;
2395 } else {
2396 $bibrec = { isbn => $isbn_string };
2397 }
2398 }
2399 else {
2400 $bibrec = { title => $bibrec } if !ref $bibrec;
2401 }
2402 my $array = [];
2403 for my $field (qw/ lccn isbn issn title author dewey subject /)
2404 {
2405 push @$array, { name => $field, value => $bibrec->{$field} }
2406 if defined $bibrec->{$field};
2407 }
2408 return $array;
2409 }
2410
2411 =head2 GetDistinctValues($field);
2412
2413 C<$field> is a reference to the fields array
2414
2415 =cut
2416
2417 sub GetDistinctValues {
2418 my ($fieldname,$string)=@_;
2419 # returns a reference to a hash of references to branches...
2420 if ($fieldname=~/\./){
2421 my ($table,$column)=split /\./, $fieldname;
2422 my $dbh = C4::Context->dbh;
2423 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2424 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 ");
2425 $sth->execute;
2426 my $elements=$sth->fetchall_arrayref({});
2427 return $elements;
2428 }
2429 else {
2430 $string||= qq("");
2431 my @servers=qw<biblioserver authorityserver>;
2432 my (@zconns,@results);
2433 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2434 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2435 $results[$i] =
2436 $zconns[$i]->scan(
2437 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2438 );
2439 }
2440 # The big moment: asynchronously retrieve results from all servers
2441 my @elements;
2442 _ZOOM_event_loop(
2443 \@zconns,
2444 \@results,
2445 sub {
2446 my ( $i, $size ) = @_;
2447 for ( my $j = 0 ; $j < $size ; $j++ ) {
2448 my %hashscan;
2449 @hashscan{qw(value cnt)} =
2450 $results[ $i - 1 ]->display_term($j);
2451 push @elements, \%hashscan;
2452 }
2453 }
2454 );
2455 return \@elements;
2456 }
2457 }
2458
2459 =head2 _ZOOM_event_loop
2460
2461 _ZOOM_event_loop(\@zconns, \@results, sub {
2462 my ( $i, $size ) = @_;
2463 ....
2464 } );
2465
2466 Processes a ZOOM event loop and passes control to a closure for
2467 processing the results, and destroying the resultsets.
2468
2469 =cut
2470
2471 sub _ZOOM_event_loop {
2472 my ($zconns, $results, $callback) = @_;
2473 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2474 my $ev = $zconns->[ $i - 1 ]->last_event();
2475 if ( $ev == ZOOM::Event::ZEND ) {
2476 next unless $results->[ $i - 1 ];
2477 my $size = $results->[ $i - 1 ]->size();
2478 if ( $size > 0 ) {
2479 $callback->($i, $size);
2480 }
2481 }
2482 }
2483
2484 foreach my $result (@$results) {
2485 $result->destroy();
2486 }
2487 }
2488
2489 =head2 new_record_from_zebra
2490
2491 Given raw data from a searchengine result set, return a MARC::Record object
2492
2493 This helper function is needed to take into account all the involved
2494 system preferences and configuration variables to properly create the
2495 MARC::Record object.
2496
2497 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2498 data. If we are using DOM, then it has to be MARCXML.
2499
2500 If we are using elasticsearch, it'll already be a MARC::Record and this
2501 function needs a new name.
2502
2503 =cut
2504
2505 sub new_record_from_zebra {
2506
2507 my $server = shift;
2508 my $raw_data = shift;
2509 # Set the default indexing modes
2510 my $search_engine = C4::Context->preference("SearchEngine");
2511 if ($search_engine eq 'Elasticsearch') {
2512 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2513 }
2514 my $index_mode = ( $server eq 'biblioserver' )
2515 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2516 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2517
2518 my $marc_record = eval {
2519 if ( $index_mode eq 'dom' ) {
2520 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2521 } else {
2522 MARC::Record->new_from_usmarc( $raw_data );
2523 }
2524 };
2525
2526 if ($@) {
2527 return;
2528 } else {
2529 return $marc_record;
2530 }
2531
2532 }
2533
2534 END { } # module clean-up code here (global destructor)
2535
2536 1;
2537 __END__
2538
2539 =head1 AUTHOR
2540
2541 Koha Development Team <http://koha-community.org/>
2542
2543 =cut
Koha ILS