?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 10306: (QA follow-up) More feedback for admins in koha2marclinks
[koha.git] / C4 / Search.pm
blob9278f262e2e0bb2ad966b4cd26f638eb81e53f5c
1 package C4::Search;
2
3 # This file is part of Koha.
4 #
5 # Koha is free software; you can redistribute it and/or modify it
6 # under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3 of the License, or
8 # (at your option) any later version.
9 #
10 # Koha is distributed in the hope that it will be useful, but
11 # WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
14 #
15 # You should have received a copy of the GNU General Public License
16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
17
18 use strict;
19 #use warnings; FIXME - Bug 2505
20 require Exporter;
21 use C4::Context;
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
24 use Koha::DateUtils;
25 use Koha::Libraries;
26 use Lingua::Stem;
27 use C4::Search::PazPar2;
28 use XML::Simple;
29 use C4::XSLT;
30 use C4::Reserves; # GetReserveStatus
31 use C4::Debug;
32 use C4::Charset;
33 use Koha::AuthorisedValues;
34 use Koha::ItemTypes;
35 use Koha::Libraries;
36 use Koha::Patrons;
37 use YAML;
38 use URI::Escape;
39 use Business::ISBN;
40 use MARC::Record;
41 use MARC::Field;
42 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
43
44 BEGIN {
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
46 }
47
48 =head1 NAME
49
50 C4::Search - Functions for searching the Koha catalog.
51
52 =head1 SYNOPSIS
53
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
55
56 =head1 DESCRIPTION
57
58 This module provides searching functions for Koha's bibliographic databases
59
60 =head1 FUNCTIONS
61
62 =cut
63
64 @ISA = qw(Exporter);
65 @EXPORT = qw(
66 &FindDuplicate
67 &SimpleSearch
68 &searchResults
69 &getRecords
70 &buildQuery
71 &GetDistinctValues
72 &enabled_staff_search_views
73 );
74
75 # make all your functions, whether exported or not;
76
77 =head2 FindDuplicate
78
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
80
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
82
83 =cut
84
85 sub FindDuplicate {
86 my ($record) = @_;
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $record, '' );
89 my $sth;
90 my $query;
91 my $search;
92 my $type;
93 my ( $biblionumber, $title );
94
95 # search duplicate on ISBN, easy and fast..
96 # ... normalize first
97 if ( $result->{isbn} ) {
98 $result->{isbn} =~ s/\(.*$//;
99 $result->{isbn} =~ s/\s+$//;
100 $query = "isbn:$result->{isbn}";
101 }
102 else {
103 my $QParser;
104 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
105 my $titleindex;
106 my $authorindex;
107 my $op;
108
109 if ($QParser) {
110 $titleindex = 'title|exact';
111 $authorindex = 'author|exact';
112 $op = '&&';
113 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
114 } else {
115 $titleindex = 'ti,ext';
116 $authorindex = 'au,ext';
117 $op = 'and';
118 }
119
120 $result->{title} =~ s /\\//g;
121 $result->{title} =~ s /\"//g;
122 $result->{title} =~ s /\(//g;
123 $result->{title} =~ s /\)//g;
124
125 # FIXME: instead of removing operators, could just do
126 # quotes around the value
127 $result->{title} =~ s/(and|or|not)//g;
128 $query = "$titleindex:\"$result->{title}\"";
129 if ( $result->{author} ) {
130 $result->{author} =~ s /\\//g;
131 $result->{author} =~ s /\"//g;
132 $result->{author} =~ s /\(//g;
133 $result->{author} =~ s /\)//g;
134
135 # remove valid operators
136 $result->{author} =~ s/(and|or|not)//g;
137 $query .= " $op $authorindex:\"$result->{author}\"";
138 }
139 }
140
141 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
142 my @results;
143 if (!defined $error) {
144 foreach my $possible_duplicate_record (@{$searchresults}) {
145 my $marcrecord = new_record_from_zebra(
146 'biblioserver',
147 $possible_duplicate_record
148 );
149
150 my $result = TransformMarcToKoha( $marcrecord, '' );
151
152 # FIXME :: why 2 $biblionumber ?
153 if ($result) {
154 push @results, $result->{'biblionumber'};
155 push @results, $result->{'title'};
156 }
157 }
158 }
159 return @results;
160 }
161
162 =head2 SimpleSearch
163
164 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
165
166 This function provides a simple search API on the bibliographic catalog
167
168 =over 2
169
170 =item C<input arg:>
171
172 * $query can be a simple keyword or a complete CCL query
173 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
174 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
175 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
176 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
177
178
179 =item C<Return:>
180
181 Returns an array consisting of three elements
182 * $error is undefined unless an error is detected
183 * $results is a reference to an array of records.
184 * $total_hits is the number of hits that would have been returned with no limit
185
186 If an error is returned the two other return elements are undefined. If error itself is undefined
187 the other two elements are always defined
188
189 =item C<usage in the script:>
190
191 =back
192
193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
194
195 if (defined $error) {
196 $template->param(query_error => $error);
197 warn "error: ".$error;
198 output_html_with_http_headers $input, $cookie, $template->output;
199 exit;
200 }
201
202 my $hits = @{$marcresults};
203 my @results;
204
205 for my $r ( @{$marcresults} ) {
206 my $marcrecord = MARC::File::USMARC::decode($r);
207 my $biblio = TransformMarcToKoha($marcrecord,q{});
208
209 #build the iarray of hashs for the template.
210 push @results, {
211 title => $biblio->{'title'},
212 subtitle => $biblio->{'subtitle'},
213 biblionumber => $biblio->{'biblionumber'},
214 author => $biblio->{'author'},
215 publishercode => $biblio->{'publishercode'},
216 publicationyear => $biblio->{'publicationyear'},
217 };
218
219 }
220
221 $template->param(result=>\@results);
222
223 =cut
224
225 sub SimpleSearch {
226 my ( $query, $offset, $max_results, $servers, %options ) = @_;
227
228 return ( 'No query entered', undef, undef ) unless $query;
229 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
230 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
231 my @zoom_queries;
232 my @tmpresults;
233 my @zconns;
234 my $results = [];
235 my $total_hits = 0;
236
237 my $QParser;
238 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
239 if ($QParser) {
240 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
241 }
242
243 # Initialize & Search Zebra
244 for ( my $i = 0 ; $i < @servers ; $i++ ) {
245 eval {
246 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
247 if ($QParser) {
248 $query =~ s/=/:/g unless $options{skip_normalize};
249 $QParser->parse( $query );
250 $query = $QParser->target_syntax($servers[$i]);
251 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
252 } else {
253 $query =~ s/:/=/g unless $options{skip_normalize};
254 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
255 }
256 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
257
258 # error handling
259 my $error =
260 $zconns[$i]->errmsg() . " ("
261 . $zconns[$i]->errcode() . ") "
262 . $zconns[$i]->addinfo() . " "
263 . $zconns[$i]->diagset();
264
265 return ( $error, undef, undef ) if $zconns[$i]->errcode();
266 };
267 if ($@) {
268
269 # caught a ZOOM::Exception
270 my $error =
271 $@->message() . " ("
272 . $@->code() . ") "
273 . $@->addinfo() . " "
274 . $@->diagset();
275 warn $error." for query: $query";
276 return ( $error, undef, undef );
277 }
278 }
279
280 _ZOOM_event_loop(
281 \@zconns,
282 \@tmpresults,
283 sub {
284 my ($i, $size) = @_;
285 my $first_record = defined($offset) ? $offset + 1 : 1;
286 my $hits = $tmpresults[ $i - 1 ]->size();
287 $total_hits += $hits;
288 my $last_record = $hits;
289 if ( defined $max_results && $offset + $max_results < $hits ) {
290 $last_record = $offset + $max_results;
291 }
292
293 for my $j ( $first_record .. $last_record ) {
294 my $record = eval {
295 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
296 ; # 0 indexed
297 };
298 push @{$results}, $record if defined $record;
299 }
300 }
301 );
302
303 foreach my $zoom_query (@zoom_queries) {
304 $zoom_query->destroy();
305 }
306
307 return ( undef, $results, $total_hits );
308 }
309
310 =head2 getRecords
311
312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
313
314 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
315 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
316 $query_type, $scan
317 );
318
319 The all singing, all dancing, multi-server, asynchronous, scanning,
320 searching, record nabbing, facet-building
321
322 See verbse embedded documentation.
323
324 =cut
325
326 sub getRecords {
327 my (
328 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
329 $results_per_page, $offset, $expanded_facet, $branches,
330 $itemtypes, $query_type, $scan, $opac
331 ) = @_;
332
333 my @servers = @$servers_ref;
334 my @sort_by = @$sort_by_ref;
335 $offset = 0 if $offset < 0;
336
337 # Initialize variables for the ZOOM connection and results object
338 my $zconn;
339 my @zconns;
340 my @results;
341 my $results_hashref = ();
342
343 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
344 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
345
346 # Initialize variables for the faceted results objects
347 my $facets_counter = {};
348 my $facets_info = {};
349 my $facets = getFacets();
350
351 my @facets_loop; # stores the ref to array of hashes for template facets loop
352
353 ### LOOP THROUGH THE SERVERS
354 for ( my $i = 0 ; $i < @servers ; $i++ ) {
355 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
356
357 # perform the search, create the results objects
358 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
359 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
360
361 #$query_to_use = $simple_query if $scan;
362 warn $simple_query if ( $scan and $DEBUG );
363
364 # Check if we've got a query_type defined, if so, use it
365 eval {
366 if ($query_type) {
367 if ($query_type =~ /^ccl/) {
368 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
370 } elsif ($query_type =~ /^cql/) {
371 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
372 } elsif ($query_type =~ /^pqf/) {
373 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
374 } else {
375 warn "Unknown query_type '$query_type'. Results undetermined.";
376 }
377 } elsif ($scan) {
378 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
379 } else {
380 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
381 }
382 };
383 if ($@) {
384 warn "WARNING: query problem with $query_to_use " . $@;
385 }
386
387 # Concatenate the sort_by limits and pass them to the results object
388 # Note: sort will override rank
389 my $sort_by;
390 foreach my $sort (@sort_by) {
391 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
392 $sort_by .= "1=1003 <i ";
393 }
394 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
395 $sort_by .= "1=1003 >i ";
396 }
397 elsif ( $sort eq "popularity_asc" ) {
398 $sort_by .= "1=9003 <i ";
399 }
400 elsif ( $sort eq "popularity_dsc" ) {
401 $sort_by .= "1=9003 >i ";
402 }
403 elsif ( $sort eq "call_number_asc" ) {
404 $sort_by .= "1=8007 <i ";
405 }
406 elsif ( $sort eq "call_number_dsc" ) {
407 $sort_by .= "1=8007 >i ";
408 }
409 elsif ( $sort eq "pubdate_asc" ) {
410 $sort_by .= "1=31 <i ";
411 }
412 elsif ( $sort eq "pubdate_dsc" ) {
413 $sort_by .= "1=31 >i ";
414 }
415 elsif ( $sort eq "acqdate_asc" ) {
416 $sort_by .= "1=32 <i ";
417 }
418 elsif ( $sort eq "acqdate_dsc" ) {
419 $sort_by .= "1=32 >i ";
420 }
421 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
422 $sort_by .= "1=4 <i ";
423 }
424 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
425 $sort_by .= "1=4 >i ";
426 }
427 else {
428 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
429 }
430 }
431 if ( $sort_by && !$scan && $results[$i] ) {
432 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
433 warn "WARNING sort $sort_by failed";
434 }
435 }
436 } # finished looping through servers
437
438 # The big moment: asynchronously retrieve results from all servers
439 _ZOOM_event_loop(
440 \@zconns,
441 \@results,
442 sub {
443 my ( $i, $size ) = @_;
444 my $results_hash;
445
446 # loop through the results
447 $results_hash->{'hits'} = $size;
448 my $times;
449 if ( $offset + $results_per_page <= $size ) {
450 $times = $offset + $results_per_page;
451 }
452 else {
453 $times = $size;
454 }
455
456 for ( my $j = $offset ; $j < $times ; $j++ ) {
457 my $records_hash;
458 my $record;
459
460 ## Check if it's an index scan
461 if ($scan) {
462 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
463
464 # here we create a minimal MARC record and hand it off to the
465 # template just like a normal result ... perhaps not ideal, but
466 # it works for now
467 my $tmprecord = MARC::Record->new();
468 $tmprecord->encoding('UTF-8');
469 my $tmptitle;
470 my $tmpauthor;
471
472 # the minimal record in author/title (depending on MARC flavour)
473 if ( C4::Context->preference("marcflavour") eq
474 "UNIMARC" )
475 {
476 $tmptitle = MARC::Field->new(
477 '200', ' ', ' ',
478 a => $term,
479 f => $occ
480 );
481 $tmprecord->append_fields($tmptitle);
482 }
483 else {
484 $tmptitle =
485 MARC::Field->new( '245', ' ', ' ', a => $term, );
486 $tmpauthor =
487 MARC::Field->new( '100', ' ', ' ', a => $occ, );
488 $tmprecord->append_fields($tmptitle);
489 $tmprecord->append_fields($tmpauthor);
490 }
491 $results_hash->{'RECORDS'}[$j] =
492 $tmprecord->as_usmarc();
493 }
494
495 # not an index scan
496 else {
497 $record = $results[ $i - 1 ]->record($j)->raw();
498 # warn "RECORD $j:".$record;
499 $results_hash->{'RECORDS'}[$j] = $record;
500 }
501
502 }
503 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
504
505 # Fill the facets while we're looping, but only for the
506 # biblioserver and not for a scan
507 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
508 $facets_counter = GetFacets( $results[ $i - 1 ] );
509 $facets_info = _get_facets_info( $facets );
510 }
511
512 # BUILD FACETS
513 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
514 for my $link_value (
515 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
516 keys %$facets_counter
517 )
518 {
519 my $expandable;
520 my $number_of_facets;
521 my @this_facets_array;
522 for my $one_facet (
523 sort {
524 $facets_counter->{$link_value}
525 ->{$b} <=> $facets_counter->{$link_value}
526 ->{$a}
527 } keys %{ $facets_counter->{$link_value} }
528 )
529 {
530 $number_of_facets++;
531 if ( ( $number_of_facets <= 5 )
532 || ( $expanded_facet eq $link_value )
533 || ( $facets_info->{$link_value}->{'expanded'} )
534 )
535 {
536
537 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
538 my $facet_link_value = $one_facet;
539 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
540
541 # fix the length that will display in the label,
542 my $facet_label_value = $one_facet;
543 my $facet_max_length = C4::Context->preference(
544 'FacetLabelTruncationLength')
545 || 20;
546 $facet_label_value =
547 substr( $one_facet, 0, $facet_max_length )
548 . "..."
549 if length($facet_label_value) >
550 $facet_max_length;
551
552 # if it's a branch, label by the name, not the code,
553 if ( $link_value =~ /branch/ ) {
554 if ( defined $branches
555 && ref($branches) eq "HASH"
556 && defined $branches->{$one_facet}
557 && ref( $branches->{$one_facet} ) eq
558 "HASH" )
559 {
560 $facet_label_value =
561 $branches->{$one_facet}
562 ->{'branchname'};
563 }
564 else {
565 $facet_label_value = "*";
566 }
567 }
568
569 # if it's a itemtype, label by the name, not the code,
570 if ( $link_value =~ /itype/ ) {
571 if ( defined $itemtypes
572 && ref($itemtypes) eq "HASH"
573 && defined $itemtypes->{$one_facet}
574 && ref( $itemtypes->{$one_facet} ) eq
575 "HASH" )
576 {
577 $facet_label_value =
578 $itemtypes->{$one_facet}
579 ->{translated_description};
580 }
581 }
582
583 # also, if it's a location code, use the name instead of the code
584 if ( $link_value =~ /location/ ) {
585 # TODO Retrieve all authorised values at once, instead of 1 query per entry
586 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
587 $facet_label_value = $av->count ? $av->next->opac_description : '';
588 }
589
590 # but we're down with the whole label being in the link's title.
591 push @this_facets_array,
592 {
593 facet_count =>
594 $facets_counter->{$link_value}
595 ->{$one_facet},
596 facet_label_value => $facet_label_value,
597 facet_title_value => $one_facet,
598 facet_link_value => $facet_link_value,
599 type_link_value => $link_value,
600 }
601 if ($facet_label_value);
602 }
603 }
604
605 # handle expanded option
606 unless ( $facets_info->{$link_value}->{'expanded'} ) {
607 $expandable = 1
608 if ( ( $number_of_facets > 5 )
609 && ( $expanded_facet ne $link_value ) );
610 }
611 push @facets_loop,
612 {
613 type_link_value => $link_value,
614 type_id => $link_value . "_id",
615 "type_label_"
616 . $facets_info->{$link_value}->{'label_value'} =>
617 1,
618 facets => \@this_facets_array,
619 expandable => $expandable,
620 expand => $link_value,
621 }
622 unless (
623 (
624 $facets_info->{$link_value}->{'label_value'} =~
625 /Libraries/
626 )
627 and ( Koha::Libraries->search->count == 1 )
628 );
629 }
630 }
631 }
632 );
633
634 # This sorts the facets into alphabetical order
635 if (@facets_loop) {
636 foreach my $f (@facets_loop) {
637 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
638 }
639 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
640 }
641
642 return ( undef, $results_hashref, \@facets_loop );
643 }
644
645 sub GetFacets {
646
647 my $rs = shift;
648 my $facets;
649
650 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
651 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
652
653 if ( $indexing_mode eq 'dom' &&
654 $use_zebra_facets ) {
655 $facets = _get_facets_from_zebra( $rs );
656 } else {
657 $facets = _get_facets_from_records( $rs );
658 }
659
660 return $facets;
661 }
662
663 sub _get_facets_from_records {
664
665 my $rs = shift;
666
667 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
668 my $facets_config = getFacets();
669 my $facets = {};
670 my $size = $rs->size();
671 my $jmax = $size > $facets_maxrecs
672 ? $facets_maxrecs
673 : $size;
674
675 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
676
677 my $marc_record = new_record_from_zebra (
678 'biblioserver',
679 $rs->record( $j )->raw()
680 );
681
682 if ( ! defined $marc_record ) {
683 warn "ERROR DECODING RECORD - $@: " .
684 $rs->record( $j )->raw();
685 next;
686 }
687
688 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
689 }
690
691 return $facets;
692 }
693
694 =head2 _get_facets_data_from_record
695
696 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
697
698 Internal function that extracts facets information from a MARC::Record object
699 and populates $facets_counter for using in getRecords.
700
701 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
702 facets for Zebra).
703
704 =cut
705
706 sub _get_facets_data_from_record {
707
708 my ( $marc_record, $facets, $facets_counter ) = @_;
709
710 for my $facet (@$facets) {
711
712 my @used_datas = ();
713
714 foreach my $tag ( @{ $facet->{ tags } } ) {
715
716 # tag number is the first three digits
717 my $tag_num = substr( $tag, 0, 3 );
718 # subfields are the remainder
719 my $subfield_letters = substr( $tag, 3 );
720
721 my @fields = $marc_record->field( $tag_num );
722 foreach my $field (@fields) {
723 # If $field->indicator(1) eq 'z', it means it is a 'see from'
724 # field introduced because of IncludeSeeFromInSearches, so skip it
725 next if $field->indicator(1) eq 'z';
726
727 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
728
729 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
730 push @used_datas, $data;
731 $facets_counter->{ $facet->{ idx } }->{ $data }++;
732 }
733 }
734 }
735 }
736 }
737
738 =head2 _get_facets_from_zebra
739
740 my $facets = _get_facets_from_zebra( $result_set )
741
742 Retrieves facets for a specified result set. It loops through the facets defined
743 in C4::Koha::getFacets and returns a hash with the following structure:
744
745 { facet_idx => {
746 facet_value => count
747 },
748 ...
749 }
750
751 =cut
752
753 sub _get_facets_from_zebra {
754
755 my $rs = shift;
756
757 # save current elementSetName
758 my $elementSetName = $rs->option( 'elementSetName' );
759
760 my $facets_loop = getFacets();
761 my $facets_data = {};
762 # loop through defined facets and fill the facets hashref
763 foreach my $facet ( @$facets_loop ) {
764
765 my $idx = $facet->{ idx };
766 my $sep = $facet->{ sep };
767 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
768 if ( $facet_values ) {
769 # we've actually got a result
770 $facets_data->{ $idx } = $facet_values;
771 }
772 }
773 # set elementSetName to its previous value to avoid side effects
774 $rs->option( elementSetName => $elementSetName );
775
776 return $facets_data;
777 }
778
779 =head2 _get_facet_from_result_set
780
781 my $facet_values =
782 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
783
784 Internal function that extracts facet information for a specific index ($facet_idx) and
785 returns a hash containing facet values and count:
786
787 {
788 $facet_value => $count ,
789 ...
790 }
791
792 Warning: this function has the side effect of changing the elementSetName for the result
793 set. It is a helper function for the main loop, which takes care of backing it up for
794 restoring.
795
796 =cut
797
798 sub _get_facet_from_result_set {
799
800 my $facet_idx = shift;
801 my $rs = shift;
802 my $sep = shift;
803
804 my $internal_sep = '<*>';
805 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
806
807 return if ( ! defined $facet_idx || ! defined $rs );
808 # zebra's facet element, untokenized index
809 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
810 # configure zebra results for retrieving the desired facet
811 $rs->option( elementSetName => $facet_element );
812 # get the facet record from result set
813 my $facet = $rs->record( 0 )->raw;
814 # if the facet has no restuls...
815 return if !defined $facet;
816 # TODO: benchmark DOM vs. SAX performance
817 my $facet_dom = XML::LibXML->load_xml(
818 string => ($facet)
819 );
820 my @terms = $facet_dom->getElementsByTagName('term');
821 return if ! @terms;
822
823 my $facets = {};
824 foreach my $term ( @terms ) {
825 my $facet_value = $term->textContent;
826 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
827 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
828 }
829
830 return $facets;
831 }
832
833 =head2 _get_facets_info
834
835 my $facets_info = C4::Search::_get_facets_info( $facets )
836
837 Internal function that extracts facets information and properly builds
838 the data structure needed to render facet labels.
839
840 =cut
841
842 sub _get_facets_info {
843
844 my $facets = shift;
845
846 my $facets_info = {};
847
848 for my $facet ( @$facets ) {
849 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
850 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
851 }
852
853 return $facets_info;
854 }
855
856 sub pazGetRecords {
857 my (
858 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
859 $results_per_page, $offset, $expanded_facet, $branches,
860 $query_type, $scan
861 ) = @_;
862 #NOTE: Parameter $branches is not used here !
863
864 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
865 $paz->init();
866 $paz->search($simple_query);
867 sleep 1; # FIXME: WHY?
868
869 # do results
870 my $results_hashref = {};
871 my $stats = XMLin($paz->stat);
872 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
873
874 # for a grouped search result, the number of hits
875 # is the number of groups returned; 'bib_hits' will have
876 # the total number of bibs.
877 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
878 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
879
880 HIT: foreach my $hit (@{ $results->{'hit'} }) {
881 my $recid = $hit->{recid}->[0];
882
883 my $work_title = $hit->{'md-work-title'}->[0];
884 my $work_author;
885 if (exists $hit->{'md-work-author'}) {
886 $work_author = $hit->{'md-work-author'}->[0];
887 }
888 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
889
890 my $result_group = {};
891 $result_group->{'group_label'} = $group_label;
892 $result_group->{'group_merge_key'} = $recid;
893
894 my $count = 1;
895 if (exists $hit->{count}) {
896 $count = $hit->{count}->[0];
897 }
898 $result_group->{'group_count'} = $count;
899
900 for (my $i = 0; $i < $count; $i++) {
901 # FIXME -- may need to worry about diacritics here
902 my $rec = $paz->record($recid, $i);
903 push @{ $result_group->{'RECORDS'} }, $rec;
904 }
905
906 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
907 }
908
909 # pass through facets
910 my $termlist_xml = $paz->termlist('author,subject');
911 my $terms = XMLin($termlist_xml, forcearray => 1);
912 my @facets_loop = ();
913 #die Dumper($results);
914 # foreach my $list (sort keys %{ $terms->{'list'} }) {
915 # my @facets = ();
916 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
917 # push @facets, {
918 # facet_label_value => $facet->{'name'}->[0],
919 # };
920 # }
921 # push @facets_loop, ( {
922 # type_label => $list,
923 # facets => \@facets,
924 # } );
925 # }
926
927 return ( undef, $results_hashref, \@facets_loop );
928 }
929
930 # TRUNCATION
931 sub _detect_truncation {
932 my ( $operand, $index ) = @_;
933 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
934 @regexpr );
935 $operand =~ s/^ //g;
936 my @wordlist = split( /\s/, $operand );
937 foreach my $word (@wordlist) {
938 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
939 push @rightlefttruncated, $word;
940 }
941 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
942 push @lefttruncated, $word;
943 }
944 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
945 push @righttruncated, $word;
946 }
947 elsif ( index( $word, "*" ) < 0 ) {
948 push @nontruncated, $word;
949 }
950 else {
951 push @regexpr, $word;
952 }
953 }
954 return (
955 \@nontruncated, \@righttruncated, \@lefttruncated,
956 \@rightlefttruncated, \@regexpr
957 );
958 }
959
960 # STEMMING
961 sub _build_stemmed_operand {
962 my ($operand,$lang) = @_;
963 require Lingua::Stem::Snowball ;
964 my $stemmed_operand=q{};
965
966 # If operand contains a digit, it is almost certainly an identifier, and should
967 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
968 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
969 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
970 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
971 return $operand if $operand =~ /\d/;
972
973 # FIXME: the locale should be set based on the user's language and/or search choice
974 #warn "$lang";
975 # Make sure we only use the first two letters from the language code
976 $lang = lc(substr($lang, 0, 2));
977 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
978 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
979 if ($lang eq 'nb' || $lang eq 'nn') {
980 $lang = 'no';
981 }
982 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
983 encoding => "UTF-8" );
984
985 my @words = split( / /, $operand );
986 my @stems = $stemmer->stem(\@words);
987 for my $stem (@stems) {
988 $stemmed_operand .= "$stem";
989 $stemmed_operand .= "?"
990 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
991 $stemmed_operand .= " ";
992 }
993 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
994 return $stemmed_operand;
995 }
996
997 # FIELD WEIGHTING
998 sub _build_weighted_query {
999
1000 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1001 # pretty well but could work much better if we had a smarter query parser
1002 my ( $operand, $stemmed_operand, $index ) = @_;
1003 my $stemming = C4::Context->preference("QueryStemming") || 0;
1004 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1005 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1006 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1007
1008 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1009
1010 # Keyword, or, no index specified
1011 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1012 $weighted_query .=
1013 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1014 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1015 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1016 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1017 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1018 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1019 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1020 if $fuzzy_enabled; # add fuzzy, word list
1021 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1022 if ( $stemming and $stemmed_operand )
1023 ; # add stemming, right truncation
1024 $weighted_query .= " or wrdl,r9=\"$operand\"";
1025
1026 # embedded sorting: 0 a-z; 1 z-a
1027 # $weighted_query .= ") or (sort1,aut=1";
1028 }
1029
1030 # Barcode searches should skip this process
1031 elsif ( $index eq 'bc' ) {
1032 $weighted_query .= "bc=\"$operand\"";
1033 }
1034
1035 # Authority-number searches should skip this process
1036 elsif ( $index eq 'an' ) {
1037 $weighted_query .= "an=\"$operand\"";
1038 }
1039
1040 # If the index is numeric, don't autoquote it.
1041 elsif ( $index =~ /,st-numeric$/ ) {
1042 $weighted_query .= " $index=$operand";
1043 }
1044
1045 # If the index already has more than one qualifier, wrap the operand
1046 # in quotes and pass it back (assumption is that the user knows what they
1047 # are doing and won't appreciate us mucking up their query
1048 elsif ( $index =~ ',' ) {
1049 $weighted_query .= " $index=\"$operand\"";
1050 }
1051
1052 #TODO: build better cases based on specific search indexes
1053 else {
1054 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1055 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1056 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1057 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1058 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1059 if $fuzzy_enabled; # add fuzzy, word list
1060 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1061 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1062 }
1063
1064 $weighted_query .= "))"; # close rank specification
1065 return $weighted_query;
1066 }
1067
1068 =head2 getIndexes
1069
1070 Return an array with available indexes.
1071
1072 =cut
1073
1074 sub getIndexes{
1075 my @indexes = (
1076 # biblio indexes
1077 'ab',
1078 'Abstract',
1079 'acqdate',
1080 'allrecords',
1081 'an',
1082 'Any',
1083 'at',
1084 'au',
1085 'aub',
1086 'aud',
1087 'audience',
1088 'auo',
1089 'aut',
1090 'Author',
1091 'Author-in-order ',
1092 'Author-personal-bibliography',
1093 'Authority-Number',
1094 'authtype',
1095 'bc',
1096 'Bib-level',
1097 'biblionumber',
1098 'bio',
1099 'biography',
1100 'callnum',
1101 'cfn',
1102 'Chronological-subdivision',
1103 'cn-bib-source',
1104 'cn-bib-sort',
1105 'cn-class',
1106 'cn-item',
1107 'cn-prefix',
1108 'cn-suffix',
1109 'cpn',
1110 'Code-institution',
1111 'Conference-name',
1112 'Conference-name-heading',
1113 'Conference-name-see',
1114 'Conference-name-seealso',
1115 'Content-type',
1116 'Control-number',
1117 'copydate',
1118 'Corporate-name',
1119 'Corporate-name-heading',
1120 'Corporate-name-see',
1121 'Corporate-name-seealso',
1122 'Country-publication',
1123 'ctype',
1124 'curriculum',
1125 'date-entered-on-file',
1126 'Date-of-acquisition',
1127 'Date-of-publication',
1128 'Date-time-last-modified',
1129 'Dewey-classification',
1130 'Dissertation-information',
1131 'diss',
1132 'dtlm',
1133 'EAN',
1134 'extent',
1135 'fic',
1136 'fiction',
1137 'Form-subdivision',
1138 'format',
1139 'Geographic-subdivision',
1140 'he',
1141 'Heading',
1142 'Heading-use-main-or-added-entry',
1143 'Heading-use-series-added-entry ',
1144 'Heading-use-subject-added-entry',
1145 'Host-item',
1146 'id-other',
1147 'ident',
1148 'Identifier-standard',
1149 'Illustration-code',
1150 'Index-term-genre',
1151 'Index-term-uncontrolled',
1152 'Interest-age-level',
1153 'Interest-grade-level',
1154 'ISBN',
1155 'isbn',
1156 'ISSN',
1157 'issn',
1158 'itemtype',
1159 'kw',
1160 'Koha-Auth-Number',
1161 'l-format',
1162 'language',
1163 'language-original',
1164 'lc-card',
1165 'LC-card-number',
1166 'lcn',
1167 'lex',
1168 'lexile-number',
1169 'llength',
1170 'ln',
1171 'ln-audio',
1172 'ln-subtitle',
1173 'Local-classification',
1174 'Local-number',
1175 'Match-heading',
1176 'Match-heading-see-from',
1177 'Material-type',
1178 'mc-itemtype',
1179 'mc-rtype',
1180 'mus',
1181 'name',
1182 'Music-number',
1183 'Name-geographic',
1184 'Name-geographic-heading',
1185 'Name-geographic-see',
1186 'Name-geographic-seealso',
1187 'nb',
1188 'Note',
1189 'notes',
1190 'ns',
1191 'nt',
1192 'Other-control-number',
1193 'pb',
1194 'Personal-name',
1195 'Personal-name-heading',
1196 'Personal-name-see',
1197 'Personal-name-seealso',
1198 'pl',
1199 'Place-publication',
1200 'pn',
1201 'popularity',
1202 'pubdate',
1203 'Publisher',
1204 'Provider',
1205 'pv',
1206 'Reading-grade-level',
1207 'Record-control-number',
1208 'rcn',
1209 'Record-type',
1210 'rtype',
1211 'se',
1212 'See',
1213 'See-also',
1214 'sn',
1215 'Stock-number',
1216 'su',
1217 'Subject',
1218 'Subject-heading-thesaurus',
1219 'Subject-name-personal',
1220 'Subject-subdivision',
1221 'Summary',
1222 'Suppress',
1223 'su-geo',
1224 'su-na',
1225 'su-to',
1226 'su-ut',
1227 'ut',
1228 'Term-genre-form',
1229 'Term-genre-form-heading',
1230 'Term-genre-form-see',
1231 'Term-genre-form-seealso',
1232 'ti',
1233 'Title',
1234 'Title-cover',
1235 'Title-series',
1236 'Title-uniform',
1237 'Title-uniform-heading',
1238 'Title-uniform-see',
1239 'Title-uniform-seealso',
1240 'totalissues',
1241 'yr',
1242
1243 # items indexes
1244 'acqsource',
1245 'barcode',
1246 'bc',
1247 'branch',
1248 'ccode',
1249 'classification-source',
1250 'cn-sort',
1251 'coded-location-qualifier',
1252 'copynumber',
1253 'damaged',
1254 'datelastborrowed',
1255 'datelastseen',
1256 'holdingbranch',
1257 'homebranch',
1258 'issues',
1259 'item',
1260 'itemnumber',
1261 'itype',
1262 'Local-classification',
1263 'location',
1264 'lost',
1265 'materials-specified',
1266 'mc-ccode',
1267 'mc-itype',
1268 'mc-loc',
1269 'notforloan',
1270 'Number-local-acquisition',
1271 'onloan',
1272 'price',
1273 'renewals',
1274 'replacementprice',
1275 'replacementpricedate',
1276 'reserves',
1277 'restricted',
1278 'stack',
1279 'stocknumber',
1280 'inv',
1281 'uri',
1282 'withdrawn',
1283
1284 # subject related
1285 );
1286
1287 return \@indexes;
1288 }
1289
1290 =head2 _handle_exploding_index
1291
1292 my $query = _handle_exploding_index($index, $term)
1293
1294 Callback routine to generate the search for "exploding" indexes (i.e.
1295 those indexes which are turned into multiple or-connected searches based
1296 on authority data).
1297
1298 =cut
1299
1300 sub _handle_exploding_index {
1301 my ($QParser, $filter, $params, $negate, $server) = @_;
1302 my $index = $filter;
1303 my $term = join(' ', @$params);
1304
1305 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1306
1307 my $marcflavour = C4::Context->preference('marcflavour');
1308
1309 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1310 my $wantedcodes = '';
1311 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1312 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1313 foreach my $auth (@$results) {
1314 my $record = MARC::Record->new_from_usmarc($auth);
1315 my @references = $record->field('5..');
1316 if (@references) {
1317 if ($index eq 'su-br') {
1318 $wantedcodes = 'g';
1319 } elsif ($index eq 'su-na') {
1320 $wantedcodes = 'h';
1321 } elsif ($index eq 'su-rl') {
1322 $wantedcodes = '';
1323 }
1324 foreach my $reference (@references) {
1325 my $codes = $reference->subfield($codesubfield);
1326 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1327 }
1328 }
1329 }
1330 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1331 return $query;
1332 }
1333
1334 =head2 parseQuery
1335
1336 ( $operators, $operands, $indexes, $limits,
1337 $sort_by, $scan, $lang ) =
1338 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1339
1340 Shim function to ease the transition from buildQuery to a new QueryParser.
1341 This function is called at the beginning of buildQuery, and modifies
1342 buildQuery's input. If it can handle the input, it returns a query that
1343 buildQuery will not try to parse.
1344
1345 =cut
1346
1347 sub parseQuery {
1348 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1349
1350 my @operators = $operators ? @$operators : ();
1351 my @indexes = $indexes ? @$indexes : ();
1352 my @operands = $operands ? @$operands : ();
1353 my @limits = $limits ? @$limits : ();
1354 my @sort_by = $sort_by ? @$sort_by : ();
1355
1356 my $query = $operands[0];
1357 my $index;
1358 my $term;
1359 my $query_desc;
1360
1361 my $QParser;
1362 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1363 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1364 undef $QParser if (scalar @limits > 0);
1365
1366 if ($QParser)
1367 {
1368 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1369 $query = '';
1370 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1371 next unless $operands[$ii];
1372 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1373 if ($query);
1374 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1375 $query .= $operands[$ii];
1376 }
1377 elsif ( $indexes[$ii] =~ m/su-/ ) {
1378 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1379 }
1380 else {
1381 $query .=
1382 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1383 }
1384 }
1385 foreach my $limit (@limits) {
1386 }
1387 if ( scalar(@sort_by) > 0 ) {
1388 my $modifier_re =
1389 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1390 $query =~ s/$modifier_re//g;
1391 foreach my $modifier (@sort_by) {
1392 $query .= " #$modifier";
1393 }
1394 }
1395
1396 $query_desc = $query;
1397 $query_desc =~ s/\s+/ /g;
1398 if ( C4::Context->preference("QueryWeightFields") ) {
1399 }
1400 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1401 { 'target_syntax_callback' => \&_handle_exploding_index } );
1402 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1403 { 'target_syntax_callback' => \&_handle_exploding_index } );
1404 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1405 { 'target_syntax_callback' => \&_handle_exploding_index } );
1406 $QParser->parse($query);
1407 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1408 }
1409 else {
1410 require Koha::QueryParser::Driver::PQF;
1411 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1412 s/$modifier_re//g for @operands;
1413 }
1414
1415 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1416 }
1417
1418 =head2 buildQuery
1419
1420 ( $error, $query,
1421 $simple_query, $query_cgi,
1422 $query_desc, $limit,
1423 $limit_cgi, $limit_desc,
1424 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1425
1426 Build queries and limits in CCL, CGI, Human,
1427 handle truncation, stemming, field weighting, fuzziness, etc.
1428
1429 See verbose embedded documentation.
1430
1431
1432 =cut
1433
1434 sub buildQuery {
1435 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1436
1437 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1438
1439 my $query_desc;
1440 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1441
1442 # dereference
1443 my @operators = $operators ? @$operators : ();
1444 my @indexes = $indexes ? @$indexes : ();
1445 my @operands = $operands ? @$operands : ();
1446 my @limits = $limits ? @$limits : ();
1447 my @sort_by = $sort_by ? @$sort_by : ();
1448
1449 my $stemming = C4::Context->preference("QueryStemming") || 0;
1450 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1451 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1452 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1453
1454 my $query = $operands[0];
1455 my $simple_query = $operands[0];
1456
1457 # initialize the variables we're passing back
1458 my $query_cgi;
1459 my $query_type;
1460
1461 my $limit;
1462 my $limit_cgi;
1463 my $limit_desc;
1464
1465 my $cclq = 0;
1466 my $cclindexes = getIndexes();
1467 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1468 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1469 my $dx = lc($1);
1470 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1471 }
1472 $query = "ccl=$query" if $cclq;
1473 }
1474
1475 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1476 # DIAGNOSTIC ONLY!!
1477 if ( $query =~ /^ccl=/ ) {
1478 my $q=$';
1479 # This is needed otherwise ccl= and &limit won't work together, and
1480 # this happens when selecting a subject on the opac-detail page
1481 @limits = grep {!/^$/} @limits;
1482 my $original_q = $q; # without available part
1483 unless ( grep { /^available$/ } @limits ) {
1484 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1485 $original_q = $q;
1486 }
1487 if ( @limits ) {
1488 if ( grep { /^available$/ } @limits ) {
1489 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1490 delete $limits['available'];
1491 }
1492 $q .= ' and '.join(' and ', @limits) if @limits;
1493 }
1494 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1495 }
1496 if ( $query =~ /^cql=/ ) {
1497 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1498 }
1499 if ( $query =~ /^pqf=/ ) {
1500 if ($query_desc) {
1501 $query_cgi = "q=".uri_escape_utf8($query_desc);
1502 } else {
1503 $query_desc = $';
1504 $query_cgi = "q=pqf=".uri_escape_utf8($');
1505 }
1506 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1507 }
1508
1509 # pass nested queries directly
1510 # FIXME: need better handling of some of these variables in this case
1511 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1512 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1513 # if ( $query =~ /(\(|\))/ ) {
1514 # return (
1515 # undef, $query, $simple_query, $query_cgi,
1516 # $query, $limit, $limit_cgi, $limit_desc,
1517 # 'ccl'
1518 # );
1519 # }
1520
1521 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1522 # query operands and indexes and add stemming, truncation, field weighting, etc.
1523 # Once we do so, we'll end up with a value in $query, just like if we had an
1524 # incoming $query from the user
1525 else {
1526 $query = ""
1527 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1528 my $previous_operand
1529 ; # a flag used to keep track if there was a previous query
1530 # if there was, we can apply the current operator
1531 # for every operand
1532 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1533
1534 # COMBINE OPERANDS, INDEXES AND OPERATORS
1535 if ( $operands[$i] ) {
1536 $operands[$i]=~s/^\s+//;
1537
1538 # A flag to determine whether or not to add the index to the query
1539 my $indexes_set;
1540
1541 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1542 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1543 $weight_fields = 0;
1544 $stemming = 0;
1545 } else {
1546 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1547 }
1548 my $operand = $operands[$i];
1549 my $index = $indexes[$i];
1550
1551 # Add index-specific attributes
1552
1553 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1554 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1555 #which is processed higher up in this sub. Other than that, year searches are typically
1556 #handled as limits which are not processed her either.
1557
1558 # Search ranges: Date of Publication, st-numeric
1559 if ( $index =~ /(yr|st-numeric)/ ) {
1560 #weight_fields/relevance search causes errors with date ranges
1561 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1562 #In the case of YYYY-YYYY, it will return no results
1563 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1564 }
1565
1566 # Date of Acquisition
1567 elsif ( $index =~ /acqdate/ ) {
1568 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1569 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1570 #top of the results just because they have lots of item records matching that date.
1571 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1572 #irrelevant here
1573 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1574 }
1575 # ISBN,ISSN,Standard Number, don't need special treatment
1576 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1577 (
1578 $stemming, $auto_truncation,
1579 $weight_fields, $fuzzy_enabled
1580 ) = ( 0, 0, 0, 0 );
1581
1582 if ( $index eq 'nb' ) {
1583 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1584 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1585 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1586 $indexes[$i] = $index = '';
1587 }
1588 }
1589 }
1590
1591 if(not $index){
1592 $index = 'kw';
1593 }
1594
1595 # Set default structure attribute (word list)
1596 my $struct_attr = q{};
1597 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1598 $struct_attr = ",wrdl";
1599 }
1600
1601 # Some helpful index variants
1602 my $index_plus = $index . $struct_attr . ':';
1603 my $index_plus_comma = $index . $struct_attr . ',';
1604
1605 if ($auto_truncation){
1606 unless ( $index =~ /,(st-|phr|ext)/ ) {
1607 #FIXME only valid with LTR scripts
1608 $operand=join(" ",map{
1609 (index($_,"*")>0?"$_":"$_*")
1610 }split (/\s+/,$operand));
1611 warn $operand if $DEBUG;
1612 }
1613 }
1614
1615 # Detect Truncation
1616 my $truncated_operand;
1617 my( $nontruncated, $righttruncated, $lefttruncated,
1618 $rightlefttruncated, $regexpr
1619 ) = _detect_truncation( $operand, $index );
1620 warn
1621 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1622 if $DEBUG;
1623
1624 # Apply Truncation
1625 if (
1626 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1627 scalar(@$rightlefttruncated) > 0 )
1628 {
1629
1630 # Don't field weight or add the index to the query, we do it here
1631 $indexes_set = 1;
1632 undef $weight_fields;
1633 my $previous_truncation_operand;
1634 if (scalar @$nontruncated) {
1635 $truncated_operand .= "$index_plus @$nontruncated ";
1636 $previous_truncation_operand = 1;
1637 }
1638 if (scalar @$righttruncated) {
1639 $truncated_operand .= "and " if $previous_truncation_operand;
1640 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1641 $previous_truncation_operand = 1;
1642 }
1643 if (scalar @$lefttruncated) {
1644 $truncated_operand .= "and " if $previous_truncation_operand;
1645 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1646 $previous_truncation_operand = 1;
1647 }
1648 if (scalar @$rightlefttruncated) {
1649 $truncated_operand .= "and " if $previous_truncation_operand;
1650 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1651 $previous_truncation_operand = 1;
1652 }
1653 }
1654 $operand = $truncated_operand if $truncated_operand;
1655 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1656
1657 # Handle Stemming
1658 my $stemmed_operand;
1659 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1660 if $stemming;
1661
1662 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1663
1664 # Handle Field Weighting
1665 my $weighted_operand;
1666 if ($weight_fields) {
1667 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1668 $operand = $weighted_operand;
1669 $indexes_set = 1;
1670 }
1671
1672 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1673
1674 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1675 query => $query,
1676 query_cgi => $query_cgi,
1677 query_desc => $query_desc,
1678 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1679 parsed_operand => $operand,
1680 original_operand => ($operands[$i]) ? $operands[$i] : '',
1681 index => $index,
1682 index_plus => $index_plus,
1683 indexes_set => $indexes_set,
1684 previous_operand => $previous_operand,
1685 });
1686
1687 } #/if $operands
1688 } # /for
1689 }
1690 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1691
1692 # add limits
1693 my %group_OR_limits;
1694 my $availability_limit;
1695 foreach my $this_limit (@limits) {
1696 next unless $this_limit;
1697 if ( $this_limit =~ /available/ ) {
1698 #
1699 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1700 ## In English:
1701 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1702 $availability_limit .=
1703 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1704 $limit_cgi .= "&limit=available";
1705 $limit_desc .= "";
1706 }
1707
1708 # group_OR_limits, prefixed by mc-
1709 # OR every member of the group
1710 elsif ( $this_limit =~ /mc/ ) {
1711 my ($k,$v) = split(/:/, $this_limit,2);
1712 if ( $k !~ /mc-i(tem)?type/ ) {
1713 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1714 $this_limit =~ tr/"//d;
1715 $this_limit = $k.":'".$v."'";
1716 }
1717
1718 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1719 $limit_desc .= " or " if $group_OR_limits{$k};
1720 $group_OR_limits{$k} .= "$this_limit";
1721 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1722 $limit_desc .= " $this_limit";
1723 }
1724
1725 # Regular old limits
1726 else {
1727 $limit .= " and " if $limit || $query;
1728 $limit .= "$this_limit";
1729 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1730 if ($this_limit =~ /^branch:(.+)/) {
1731 my $branchcode = $1;
1732 my $library = Koha::Libraries->find( $branchcode );
1733 if (defined $library) {
1734 $limit_desc .= " branch:" . $library->branchname;
1735 } else {
1736 $limit_desc .= " $this_limit";
1737 }
1738 } else {
1739 $limit_desc .= " $this_limit";
1740 }
1741 }
1742 }
1743 foreach my $k (keys (%group_OR_limits)) {
1744 $limit .= " and " if ( $query || $limit );
1745 $limit .= "($group_OR_limits{$k})";
1746 }
1747 if ($availability_limit) {
1748 $limit .= " and " if ( $query || $limit );
1749 $limit .= "($availability_limit)";
1750 }
1751
1752 # Normalize the query and limit strings
1753 # This is flawed , means we can't search anything with : in it
1754 # if user wants to do ccl or cql, start the query with that
1755 # $query =~ s/:/=/g;
1756 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1757 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1758 $query =~ s/(?<=(wrdl)):/=/g;
1759 $query =~ s/(?<=(trn|phr)):/=/g;
1760 $query =~ s/(?<=(st-numeric)):/=/g;
1761 $query =~ s/(?<=(st-year)):/=/g;
1762 $query =~ s/(?<=(st-date-normalized)):/=/g;
1763 $limit =~ s/:/=/g;
1764 for ( $query, $query_desc, $limit, $limit_desc ) {
1765 s/ +/ /g; # remove extra spaces
1766 s/^ //g; # remove any beginning spaces
1767 s/ $//g; # remove any ending spaces
1768 s/==/=/g; # remove double == from query
1769 }
1770 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1771
1772 for ($query_cgi,$simple_query) {
1773 s/"//g;
1774 }
1775 # append the limit to the query
1776 $query .= " " . $limit;
1777
1778 # Warnings if DEBUG
1779 if ($DEBUG) {
1780 warn "QUERY:" . $query;
1781 warn "QUERY CGI:" . $query_cgi;
1782 warn "QUERY DESC:" . $query_desc;
1783 warn "LIMIT:" . $limit;
1784 warn "LIMIT CGI:" . $limit_cgi;
1785 warn "LIMIT DESC:" . $limit_desc;
1786 warn "---------\nLeave buildQuery\n---------";
1787 }
1788
1789 return (
1790 undef, $query, $simple_query, $query_cgi,
1791 $query_desc, $limit, $limit_cgi, $limit_desc,
1792 $query_type
1793 );
1794 }
1795
1796 =head2 _build_initial_query
1797
1798 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1799
1800 Build a section of the initial query containing indexes, operators, and operands.
1801
1802 =cut
1803
1804 sub _build_initial_query {
1805 my ($params) = @_;
1806
1807 my $operator = "";
1808 if ($params->{previous_operand}){
1809 #If there is a previous operand, add a supplied operator or the default 'and'
1810 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1811 }
1812
1813 #NOTE: indexes_set is typically set when doing truncation or field weighting
1814 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1815
1816 #e.g. "kw,wrdl:test"
1817 #e.g. " and kw,wrdl:test"
1818 $params->{query} .= $operator . $operand;
1819
1820 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1821 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1822 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1823
1824 #e.g. " and kw,wrdl: test"
1825 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1826
1827 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1828
1829 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1830 }
1831
1832 =head2 searchResults
1833
1834 my @search_results = searchResults($search_context, $searchdesc, $hits,
1835 $results_per_page, $offset, $scan,
1836 @marcresults);
1837
1838 Format results in a form suitable for passing to the template
1839
1840 =cut
1841
1842 # IMO this subroutine is pretty messy still -- it's responsible for
1843 # building the HTML output for the template
1844 sub searchResults {
1845 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1846 my $dbh = C4::Context->dbh;
1847 my @newresults;
1848
1849 require C4::Items;
1850
1851 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1852 my ($is_opac, $hidelostitems);
1853 if ($search_context eq 'opac') {
1854 $hidelostitems = C4::Context->preference('hidelostitems');
1855 $is_opac = 1;
1856 }
1857
1858 #Build branchnames hash
1859 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1860
1861 # FIXME - We build an authorised values hash here, using the default framework
1862 # though it is possible to have different authvals for different fws.
1863
1864 my $shelflocations =
1865 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1866
1867 # get notforloan authorised value list (see $shelflocations FIXME)
1868 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => { not => undef } });
1869 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1870
1871 #Get itemtype hash
1872 my $itemtypes = Koha::ItemTypes->search_with_localization;
1873 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1874
1875 #search item field code
1876 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1877
1878 ## find column names of items related to MARC
1879 my %subfieldstosearch;
1880 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1881 for my $column ( @columns ) {
1882 my ( $tagfield, $tagsubfield ) =
1883 &GetMarcFromKohaField( "items." . $column, "" );
1884 if ( defined $tagsubfield ) {
1885 $subfieldstosearch{$column} = $tagsubfield;
1886 }
1887 }
1888
1889 # handle which records to actually retrieve
1890 my $times;
1891 if ( $hits && $offset + $results_per_page <= $hits ) {
1892 $times = $offset + $results_per_page;
1893 }
1894 else {
1895 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1896 }
1897
1898 my $marcflavour = C4::Context->preference("marcflavour");
1899 # We get the biblionumber position in MARC
1900 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1901
1902 # set stuff for XSLT processing here once, not later again for every record we retrieved
1903 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1904 my $xslsyspref = $interface . "XSLTResultsDisplay";
1905 my $xslfile = C4::Context->preference($xslsyspref);
1906 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1907 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1908
1909 # loop through all of the records we've retrieved
1910 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1911
1912 my $marcrecord;
1913 if ($scan) {
1914 # For Scan searches we built USMARC data
1915 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1916 } else {
1917 # Normal search, render from Zebra's output
1918 $marcrecord = new_record_from_zebra(
1919 'biblioserver',
1920 $marcresults->[$i]
1921 );
1922
1923 if ( ! defined $marcrecord ) {
1924 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1925 next;
1926 }
1927 }
1928
1929 my $fw = $scan
1930 ? undef
1931 : $bibliotag < 10
1932 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1933 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1934
1935 SetUTF8Flag($marcrecord);
1936 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1937 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1938 $oldbiblio->{result_number} = $i + 1;
1939
1940 # add imageurl to itemtype if there is one
1941 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1942
1943 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1944 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1945 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1946 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1947 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1948
1949 # edition information, if any
1950 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1951 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1952 # Build summary if there is one (the summary is defined in the itemtypes table)
1953 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1954 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1955 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1956 my @fields = $marcrecord->fields();
1957
1958 my $newsummary;
1959 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1960 my $tags = {};
1961 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1962 $tag =~ /(.{3})(.)/;
1963 if($marcrecord->field($1)){
1964 my @abc = $marcrecord->field($1)->subfield($2);
1965 $tags->{$tag} = $#abc + 1 ;
1966 }
1967 }
1968
1969 # We catch how many times to repeat this line
1970 my $max = 0;
1971 foreach my $tag (keys(%$tags)){
1972 $max = $tags->{$tag} if($tags->{$tag} > $max);
1973 }
1974
1975 # we replace, and repeat each line
1976 for (my $i = 0 ; $i < $max ; $i++){
1977 my $newline = $line;
1978
1979 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1980 $tag =~ /(.{3})(.)/;
1981
1982 if($marcrecord->field($1)){
1983 my @repl = $marcrecord->field($1)->subfield($2);
1984 my $subfieldvalue = $repl[$i];
1985 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1986 }
1987 }
1988 $newsummary .= "$newline\n";
1989 }
1990 }
1991
1992 $newsummary =~ s/\[(.*?)]//g;
1993 $newsummary =~ s/\n/<br\/>/g;
1994 $oldbiblio->{summary} = $newsummary;
1995 }
1996
1997 # Pull out the items fields
1998 my @fields = $marcrecord->field($itemtag);
1999 my $marcflavor = C4::Context->preference("marcflavour");
2000 # adding linked items that belong to host records
2001 my $analyticsfield = '773';
2002 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2003 $analyticsfield = '773';
2004 } elsif ($marcflavor eq 'UNIMARC') {
2005 $analyticsfield = '461';
2006 }
2007 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2008 my $hostbiblionumber = $hostfield->subfield("0");
2009 my $linkeditemnumber = $hostfield->subfield("9");
2010 if(!$hostbiblionumber eq undef){
2011 my $hostbiblio = GetMarcBiblio({
2012 biblionumber => $hostbiblionumber,
2013 embed_items => 1 });
2014 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2015 if(!$hostbiblio eq undef){
2016 my @hostitems = $hostbiblio->field($itemfield);
2017 foreach my $hostitem (@hostitems){
2018 if ($hostitem->subfield("9") eq $linkeditemnumber){
2019 my $linkeditem =$hostitem;
2020 # append linked items if they exist
2021 if (!$linkeditem eq undef){
2022 push (@fields, $linkeditem);}
2023 }
2024 }
2025 }
2026 }
2027 }
2028
2029 # Setting item statuses for display
2030 my @available_items_loop;
2031 my @onloan_items_loop;
2032 my @other_items_loop;
2033
2034 my $available_items;
2035 my $onloan_items;
2036 my $other_items;
2037
2038 my $ordered_count = 0;
2039 my $available_count = 0;
2040 my $onloan_count = 0;
2041 my $longoverdue_count = 0;
2042 my $other_count = 0;
2043 my $withdrawn_count = 0;
2044 my $itemlost_count = 0;
2045 my $hideatopac_count = 0;
2046 my $itembinding_count = 0;
2047 my $itemdamaged_count = 0;
2048 my $item_in_transit_count = 0;
2049 my $can_place_holds = 0;
2050 my $item_onhold_count = 0;
2051 my $notforloan_count = 0;
2052 my $items_count = scalar(@fields);
2053 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2054 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2055 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2056
2057 # loop through every item
2058 foreach my $field (@fields) {
2059 my $item;
2060
2061 # populate the items hash
2062 foreach my $code ( keys %subfieldstosearch ) {
2063 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2064 }
2065 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2066
2067 # OPAC hidden items
2068 if ($is_opac) {
2069 # hidden because lost
2070 if ($hidelostitems && $item->{itemlost}) {
2071 $hideatopac_count++;
2072 next;
2073 }
2074 # hidden based on OpacHiddenItems syspref
2075 my @hi = C4::Items::GetHiddenItemnumbers($item);
2076 if (scalar @hi) {
2077 push @hiddenitems, @hi;
2078 $hideatopac_count++;
2079 next;
2080 }
2081 }
2082
2083 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2084 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2085
2086 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2087 if ($item->{$hbranch}) {
2088 $item->{'branchname'} = $branches{$item->{$hbranch}};
2089 }
2090 elsif ($item->{$otherbranch}) { # Last resort
2091 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2092 }
2093
2094 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2095 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2096 my $userenv = C4::Context->userenv;
2097 if ( $item->{onloan}
2098 && $userenv
2099 && $userenv->{number}
2100 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2101 {
2102 $onloan_count++;
2103 my $key = $prefix . $item->{onloan} . $item->{barcode};
2104 $onloan_items->{$key}->{due_date} = $item->{onloan};
2105 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2106 $onloan_items->{$key}->{branchname} = $item->{branchname};
2107 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2108 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2109 $onloan_items->{$key}->{description} = $item->{description};
2110 $onloan_items->{$key}->{imageurl} =
2111 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2112
2113 # if something's checked out and lost, mark it as 'long overdue'
2114 if ( $item->{itemlost} ) {
2115 $onloan_items->{$key}->{longoverdue}++;
2116 $longoverdue_count++;
2117 }
2118 else { # can place holds as long as item isn't lost
2119 $can_place_holds = 1;
2120 }
2121 }
2122
2123 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2124 else {
2125
2126 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2127
2128 # item is on order
2129 if ( $item->{notforloan} < 0 ) {
2130 $ordered_count++;
2131 } elsif ( $item->{notforloan} > 0 ) {
2132 $notforloan_count++;
2133 }
2134
2135 # is item in transit?
2136 my $transfertwhen = '';
2137 my ($transfertfrom, $transfertto);
2138
2139 # is item on the reserve shelf?
2140 my $reservestatus = '';
2141
2142 unless ($item->{withdrawn}
2143 || $item->{itemlost}
2144 || $item->{damaged}
2145 || $item->{notforloan}
2146 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2147 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2148
2149 # A couple heuristics to limit how many times
2150 # we query the database for item transfer information, sacrificing
2151 # accuracy in some cases for speed;
2152 #
2153 # 1. don't query if item has one of the other statuses
2154 # 2. don't check transit status if the bib has
2155 # more than 20 items
2156 #
2157 # FIXME: to avoid having the query the database like this, and to make
2158 # the in transit status count as unavailable for search limiting,
2159 # should map transit status to record indexed in Zebra.
2160 #
2161 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2162 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2163 }
2164
2165 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2166 if ( $item->{withdrawn}
2167 || $item->{itemlost}
2168 || $item->{damaged}
2169 || $item->{notforloan}
2170 || $reservestatus eq 'Waiting'
2171 || ($transfertwhen ne ''))
2172 {
2173 $withdrawn_count++ if $item->{withdrawn};
2174 $itemlost_count++ if $item->{itemlost};
2175 $itemdamaged_count++ if $item->{damaged};
2176 $item_in_transit_count++ if $transfertwhen ne '';
2177 $item_onhold_count++ if $reservestatus eq 'Waiting';
2178 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2179
2180 # can place a hold on a item if
2181 # not lost nor withdrawn
2182 # not damaged unless AllowHoldsOnDamagedItems is true
2183 # item is either for loan or on order (notforloan < 0)
2184 $can_place_holds = 1
2185 if (
2186 !$item->{itemlost}
2187 && !$item->{withdrawn}
2188 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2189 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2190 );
2191
2192 $other_count++;
2193
2194 my $key = $prefix . $item->{status};
2195 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2196 $other_items->{$key}->{$_} = $item->{$_};
2197 }
2198 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2199 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2200 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2201 $other_items->{$key}->{count}++ if $item->{$hbranch};
2202 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2203 $other_items->{$key}->{description} = $item->{description};
2204 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2205 }
2206 # item is available
2207 else {
2208 $can_place_holds = 1;
2209 $available_count++;
2210 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2211 foreach (qw(branchname itemcallnumber description)) {
2212 $available_items->{$prefix}->{$_} = $item->{$_};
2213 }
2214 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2215 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2216 }
2217 }
2218 } # notforloan, item level and biblioitem level
2219
2220 # if all items are hidden, do not show the record
2221 if ($items_count > 0 && $hideatopac_count == $items_count) {
2222 next;
2223 }
2224
2225 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2226 for my $key ( sort keys %$onloan_items ) {
2227 (++$onloanitemscount > $maxitems) and last;
2228 push @onloan_items_loop, $onloan_items->{$key};
2229 }
2230 for my $key ( sort keys %$other_items ) {
2231 (++$otheritemscount > $maxitems) and last;
2232 push @other_items_loop, $other_items->{$key};
2233 }
2234 for my $key ( sort keys %$available_items ) {
2235 (++$availableitemscount > $maxitems) and last;
2236 push @available_items_loop, $available_items->{$key}
2237 }
2238
2239 # XSLT processing of some stuff
2240 # we fetched the sysprefs already before the loop through all retrieved record!
2241 if (!$scan && $xslfile) {
2242 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2243 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2244 }
2245
2246 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2247 if (!C4::Context->preference("item-level_itypes")) {
2248 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2249 $can_place_holds = 0;
2250 }
2251 }
2252 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2253 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2254 $oldbiblio->{items_count} = $items_count;
2255 $oldbiblio->{available_items_loop} = \@available_items_loop;
2256 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2257 $oldbiblio->{other_items_loop} = \@other_items_loop;
2258 $oldbiblio->{availablecount} = $available_count;
2259 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2260 $oldbiblio->{onloancount} = $onloan_count;
2261 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2262 $oldbiblio->{othercount} = $other_count;
2263 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2264 $oldbiblio->{withdrawncount} = $withdrawn_count;
2265 $oldbiblio->{itemlostcount} = $itemlost_count;
2266 $oldbiblio->{damagedcount} = $itemdamaged_count;
2267 $oldbiblio->{intransitcount} = $item_in_transit_count;
2268 $oldbiblio->{onholdcount} = $item_onhold_count;
2269 $oldbiblio->{orderedcount} = $ordered_count;
2270 $oldbiblio->{notforloancount} = $notforloan_count;
2271
2272 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2273 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2274 my $subfields = substr $fieldspec, 3;
2275 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2276 my @alternateholdingsinfo = ();
2277 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2278 my $alternateholdingscount = 0;
2279
2280 for my $field (@holdingsfields) {
2281 my %holding = ( holding => '' );
2282 my $havesubfield = 0;
2283 for my $subfield ($field->subfields()) {
2284 if ((index $subfields, $$subfield[0]) >= 0) {
2285 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2286 $holding{'holding'} .= $$subfield[1];
2287 $havesubfield++;
2288 }
2289 }
2290 if ($havesubfield) {
2291 push(@alternateholdingsinfo, \%holding);
2292 $alternateholdingscount++;
2293 }
2294 }
2295
2296 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2297 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2298 }
2299
2300 push( @newresults, $oldbiblio );
2301 }
2302
2303 return @newresults;
2304 }
2305
2306 =head2 enabled_staff_search_views
2307
2308 %hash = enabled_staff_search_views()
2309
2310 This function returns a hash that contains three flags obtained from the system
2311 preferences, used to determine whether a particular staff search results view
2312 is enabled.
2313
2314 =over 2
2315
2316 =item C<Output arg:>
2317
2318 * $hash{can_view_MARC} is true only if the MARC view is enabled
2319 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2320 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2321
2322 =item C<usage in the script:>
2323
2324 =back
2325
2326 $template->param ( C4::Search::enabled_staff_search_views );
2327
2328 =cut
2329
2330 sub enabled_staff_search_views
2331 {
2332 return (
2333 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2334 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2335 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2336 );
2337 }
2338
2339 =head2 z3950_search_args
2340
2341 $arrayref = z3950_search_args($matchpoints)
2342
2343 This function returns an array reference that contains the search parameters to be
2344 passed to the Z39.50 search script (z3950_search.pl). The array elements
2345 are hash refs whose keys are name and value, and whose values are the
2346 name of a search parameter, the value of that search parameter and the URL encoded
2347 value of that parameter.
2348
2349 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2350
2351 The search parameter values are obtained from the bibliographic record whose
2352 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2353
2354 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2355 a general purpose search argument. In this case, the returned array contains only
2356 entry: the key is 'title' and the value is derived from $matchpoints.
2357
2358 If a search parameter value is undefined or empty, it is not included in the returned
2359 array.
2360
2361 The returned array reference may be passed directly to the template parameters.
2362
2363 =over 2
2364
2365 =item C<Output arg:>
2366
2367 * $array containing hash refs as described above
2368
2369 =item C<usage in the script:>
2370
2371 =back
2372
2373 $data = Biblio::GetBiblioData($bibno);
2374 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2375
2376 *OR*
2377
2378 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2379
2380 =cut
2381
2382 sub z3950_search_args {
2383 my $bibrec = shift;
2384
2385 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2386 my $isbn = Business::ISBN->new( $isbn_string );
2387
2388 if (defined $isbn && $isbn->is_valid)
2389 {
2390 if ( ref($bibrec) ) {
2391 $bibrec->{isbn} = $isbn_string;
2392 $bibrec->{title} = undef;
2393 } else {
2394 $bibrec = { isbn => $isbn_string };
2395 }
2396 }
2397 else {
2398 $bibrec = { title => $bibrec } if !ref $bibrec;
2399 }
2400 my $array = [];
2401 for my $field (qw/ lccn isbn issn title author dewey subject /)
2402 {
2403 push @$array, { name => $field, value => $bibrec->{$field} }
2404 if defined $bibrec->{$field};
2405 }
2406 return $array;
2407 }
2408
2409 =head2 GetDistinctValues($field);
2410
2411 C<$field> is a reference to the fields array
2412
2413 =cut
2414
2415 sub GetDistinctValues {
2416 my ($fieldname,$string)=@_;
2417 # returns a reference to a hash of references to branches...
2418 if ($fieldname=~/\./){
2419 my ($table,$column)=split /\./, $fieldname;
2420 my $dbh = C4::Context->dbh;
2421 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2422 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 ");
2423 $sth->execute;
2424 my $elements=$sth->fetchall_arrayref({});
2425 return $elements;
2426 }
2427 else {
2428 $string||= qq("");
2429 my @servers=qw<biblioserver authorityserver>;
2430 my (@zconns,@results);
2431 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2432 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2433 $results[$i] =
2434 $zconns[$i]->scan(
2435 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2436 );
2437 }
2438 # The big moment: asynchronously retrieve results from all servers
2439 my @elements;
2440 _ZOOM_event_loop(
2441 \@zconns,
2442 \@results,
2443 sub {
2444 my ( $i, $size ) = @_;
2445 for ( my $j = 0 ; $j < $size ; $j++ ) {
2446 my %hashscan;
2447 @hashscan{qw(value cnt)} =
2448 $results[ $i - 1 ]->display_term($j);
2449 push @elements, \%hashscan;
2450 }
2451 }
2452 );
2453 return \@elements;
2454 }
2455 }
2456
2457 =head2 _ZOOM_event_loop
2458
2459 _ZOOM_event_loop(\@zconns, \@results, sub {
2460 my ( $i, $size ) = @_;
2461 ....
2462 } );
2463
2464 Processes a ZOOM event loop and passes control to a closure for
2465 processing the results, and destroying the resultsets.
2466
2467 =cut
2468
2469 sub _ZOOM_event_loop {
2470 my ($zconns, $results, $callback) = @_;
2471 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2472 my $ev = $zconns->[ $i - 1 ]->last_event();
2473 if ( $ev == ZOOM::Event::ZEND ) {
2474 next unless $results->[ $i - 1 ];
2475 my $size = $results->[ $i - 1 ]->size();
2476 if ( $size > 0 ) {
2477 $callback->($i, $size);
2478 }
2479 }
2480 }
2481
2482 foreach my $result (@$results) {
2483 $result->destroy();
2484 }
2485 }
2486
2487 =head2 new_record_from_zebra
2488
2489 Given raw data from a searchengine result set, return a MARC::Record object
2490
2491 This helper function is needed to take into account all the involved
2492 system preferences and configuration variables to properly create the
2493 MARC::Record object.
2494
2495 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2496 data. If we are using DOM, then it has to be MARCXML.
2497
2498 If we are using elasticsearch, it'll already be a MARC::Record and this
2499 function needs a new name.
2500
2501 =cut
2502
2503 sub new_record_from_zebra {
2504
2505 my $server = shift;
2506 my $raw_data = shift;
2507 # Set the default indexing modes
2508 my $search_engine = C4::Context->preference("SearchEngine");
2509 if ($search_engine eq 'Elasticsearch') {
2510 return $raw_data;
2511 }
2512 my $index_mode = ( $server eq 'biblioserver' )
2513 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2514 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2515
2516 my $marc_record = eval {
2517 if ( $index_mode eq 'dom' ) {
2518 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2519 } else {
2520 MARC::Record->new_from_usmarc( $raw_data );
2521 }
2522 };
2523
2524 if ($@) {
2525 return;
2526 } else {
2527 return $marc_record;
2528 }
2529
2530 }
2531
2532 END { } # module clean-up code here (global destructor)
2533
2534 1;
2535 __END__
2536
2537 =head1 AUTHOR
2538
2539 Koha Development Team <http://koha-community.org/>
2540
2541 =cut
Koha ILS