?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 19817: use https instead of http
[koha.git] / C4 / Search.pm
blob06e837de03cb810519447c2f6c3bc7c9579f8a42
1 package C4::Search;
2
3 # This file is part of Koha.
4 #
5 # Koha is free software; you can redistribute it and/or modify it
6 # under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3 of the License, or
8 # (at your option) any later version.
9 #
10 # Koha is distributed in the hope that it will be useful, but
11 # WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
14 #
15 # You should have received a copy of the GNU General Public License
16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
17
18 use strict;
19 #use warnings; FIXME - Bug 2505
20 require Exporter;
21 use C4::Context;
22 use C4::Biblio; # GetMarcFromKohaField, GetBiblioData
23 use C4::Koha; # getFacets
24 use Koha::DateUtils;
25 use Koha::Libraries;
26 use Lingua::Stem;
27 use C4::Search::PazPar2;
28 use XML::Simple;
29 use C4::XSLT;
30 use C4::Reserves; # GetReserveStatus
31 use C4::Debug;
32 use C4::Charset;
33 use Koha::AuthorisedValues;
34 use Koha::ItemTypes;
35 use Koha::Libraries;
36 use Koha::Patrons;
37 use YAML;
38 use URI::Escape;
39 use Business::ISBN;
40 use MARC::Record;
41 use MARC::Field;
42 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
43
44 BEGIN {
45 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
46 }
47
48 =head1 NAME
49
50 C4::Search - Functions for searching the Koha catalog.
51
52 =head1 SYNOPSIS
53
54 See opac/opac-search.pl or catalogue/search.pl for example of usage
55
56 =head1 DESCRIPTION
57
58 This module provides searching functions for Koha's bibliographic databases
59
60 =head1 FUNCTIONS
61
62 =cut
63
64 @ISA = qw(Exporter);
65 @EXPORT = qw(
66 &FindDuplicate
67 &SimpleSearch
68 &searchResults
69 &getRecords
70 &buildQuery
71 &GetDistinctValues
72 &enabled_staff_search_views
73 );
74
75 # make all your functions, whether exported or not;
76
77 =head2 FindDuplicate
78
79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
80
81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
82
83 =cut
84
85 sub FindDuplicate {
86 my ($record) = @_;
87 my $dbh = C4::Context->dbh;
88 my $result = TransformMarcToKoha( $record, '' );
89 my $sth;
90 my $query;
91 my $search;
92 my $type;
93 my ( $biblionumber, $title );
94
95 # search duplicate on ISBN, easy and fast..
96 # ... normalize first
97 if ( $result->{isbn} ) {
98 $result->{isbn} =~ s/\(.*$//;
99 $result->{isbn} =~ s/\s+$//;
100 $query = "isbn:$result->{isbn}";
101 }
102 else {
103 my $QParser;
104 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
105 my $titleindex;
106 my $authorindex;
107 my $op;
108
109 if ($QParser) {
110 $titleindex = 'title|exact';
111 $authorindex = 'author|exact';
112 $op = '&&';
113 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
114 } else {
115 $titleindex = 'ti,ext';
116 $authorindex = 'au,ext';
117 $op = 'and';
118 }
119
120 $result->{title} =~ s /\\//g;
121 $result->{title} =~ s /\"//g;
122 $result->{title} =~ s /\(//g;
123 $result->{title} =~ s /\)//g;
124
125 # FIXME: instead of removing operators, could just do
126 # quotes around the value
127 $result->{title} =~ s/(and|or|not)//g;
128 $query = "$titleindex:\"$result->{title}\"";
129 if ( $result->{author} ) {
130 $result->{author} =~ s /\\//g;
131 $result->{author} =~ s /\"//g;
132 $result->{author} =~ s /\(//g;
133 $result->{author} =~ s /\)//g;
134
135 # remove valid operators
136 $result->{author} =~ s/(and|or|not)//g;
137 $query .= " $op $authorindex:\"$result->{author}\"";
138 }
139 }
140
141 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
142 my @results;
143 if (!defined $error) {
144 foreach my $possible_duplicate_record (@{$searchresults}) {
145 my $marcrecord = new_record_from_zebra(
146 'biblioserver',
147 $possible_duplicate_record
148 );
149
150 my $result = TransformMarcToKoha( $marcrecord, '' );
151
152 # FIXME :: why 2 $biblionumber ?
153 if ($result) {
154 push @results, $result->{'biblionumber'};
155 push @results, $result->{'title'};
156 }
157 }
158 }
159 return @results;
160 }
161
162 =head2 SimpleSearch
163
164 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] );
165
166 This function provides a simple search API on the bibliographic catalog
167
168 =over 2
169
170 =item C<input arg:>
171
172 * $query can be a simple keyword or a complete CCL query
173 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
174 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
175 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
176 * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = )
177
178
179 =item C<Return:>
180
181 Returns an array consisting of three elements
182 * $error is undefined unless an error is detected
183 * $results is a reference to an array of records.
184 * $total_hits is the number of hits that would have been returned with no limit
185
186 If an error is returned the two other return elements are undefined. If error itself is undefined
187 the other two elements are always defined
188
189 =item C<usage in the script:>
190
191 =back
192
193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
194
195 if (defined $error) {
196 $template->param(query_error => $error);
197 warn "error: ".$error;
198 output_html_with_http_headers $input, $cookie, $template->output;
199 exit;
200 }
201
202 my $hits = @{$marcresults};
203 my @results;
204
205 for my $r ( @{$marcresults} ) {
206 my $marcrecord = MARC::File::USMARC::decode($r);
207 my $biblio = TransformMarcToKoha($marcrecord,q{});
208
209 #build the iarray of hashs for the template.
210 push @results, {
211 title => $biblio->{'title'},
212 subtitle => $biblio->{'subtitle'},
213 biblionumber => $biblio->{'biblionumber'},
214 author => $biblio->{'author'},
215 publishercode => $biblio->{'publishercode'},
216 publicationyear => $biblio->{'publicationyear'},
217 };
218
219 }
220
221 $template->param(result=>\@results);
222
223 =cut
224
225 sub SimpleSearch {
226 my ( $query, $offset, $max_results, $servers, %options ) = @_;
227
228 return ( 'No query entered', undef, undef ) unless $query;
229 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
230 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
231 my @zoom_queries;
232 my @tmpresults;
233 my @zconns;
234 my $results = [];
235 my $total_hits = 0;
236
237 my $QParser;
238 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
239 if ($QParser) {
240 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
241 }
242
243 # Initialize & Search Zebra
244 for ( my $i = 0 ; $i < @servers ; $i++ ) {
245 eval {
246 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
247 if ($QParser) {
248 $query =~ s/=/:/g unless $options{skip_normalize};
249 $QParser->parse( $query );
250 $query = $QParser->target_syntax($servers[$i]);
251 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
252 } else {
253 $query =~ s/:/=/g unless $options{skip_normalize};
254 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
255 }
256 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
257
258 # error handling
259 my $error =
260 $zconns[$i]->errmsg() . " ("
261 . $zconns[$i]->errcode() . ") "
262 . $zconns[$i]->addinfo() . " "
263 . $zconns[$i]->diagset();
264
265 return ( $error, undef, undef ) if $zconns[$i]->errcode();
266 };
267 if ($@) {
268
269 # caught a ZOOM::Exception
270 my $error =
271 $@->message() . " ("
272 . $@->code() . ") "
273 . $@->addinfo() . " "
274 . $@->diagset();
275 warn $error." for query: $query";
276 return ( $error, undef, undef );
277 }
278 }
279
280 _ZOOM_event_loop(
281 \@zconns,
282 \@tmpresults,
283 sub {
284 my ($i, $size) = @_;
285 my $first_record = defined($offset) ? $offset + 1 : 1;
286 my $hits = $tmpresults[ $i - 1 ]->size();
287 $total_hits += $hits;
288 my $last_record = $hits;
289 if ( defined $max_results && $offset + $max_results < $hits ) {
290 $last_record = $offset + $max_results;
291 }
292
293 for my $j ( $first_record .. $last_record ) {
294 my $record = eval {
295 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
296 ; # 0 indexed
297 };
298 push @{$results}, $record if defined $record;
299 }
300 }
301 );
302
303 foreach my $zoom_query (@zoom_queries) {
304 $zoom_query->destroy();
305 }
306
307 return ( undef, $results, $total_hits );
308 }
309
310 =head2 getRecords
311
312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
313
314 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
315 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
316 $query_type, $scan
317 );
318
319 The all singing, all dancing, multi-server, asynchronous, scanning,
320 searching, record nabbing, facet-building
321
322 See verbse embedded documentation.
323
324 =cut
325
326 sub getRecords {
327 my (
328 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
329 $results_per_page, $offset, $expanded_facet, $branches,
330 $itemtypes, $query_type, $scan, $opac
331 ) = @_;
332
333 my @servers = @$servers_ref;
334 my @sort_by = @$sort_by_ref;
335 $offset = 0 if $offset < 0;
336
337 # Initialize variables for the ZOOM connection and results object
338 my $zconn;
339 my @zconns;
340 my @results;
341 my $results_hashref = ();
342
343 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
344 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
345
346 # Initialize variables for the faceted results objects
347 my $facets_counter = {};
348 my $facets_info = {};
349 my $facets = getFacets();
350
351 my @facets_loop; # stores the ref to array of hashes for template facets loop
352
353 ### LOOP THROUGH THE SERVERS
354 for ( my $i = 0 ; $i < @servers ; $i++ ) {
355 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
356
357 # perform the search, create the results objects
358 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
359 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
360
361 #$query_to_use = $simple_query if $scan;
362 warn $simple_query if ( $scan and $DEBUG );
363
364 # Check if we've got a query_type defined, if so, use it
365 eval {
366 if ($query_type) {
367 if ($query_type =~ /^ccl/) {
368 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
370 } elsif ($query_type =~ /^cql/) {
371 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
372 } elsif ($query_type =~ /^pqf/) {
373 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
374 } else {
375 warn "Unknown query_type '$query_type'. Results undetermined.";
376 }
377 } elsif ($scan) {
378 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
379 } else {
380 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
381 }
382 };
383 if ($@) {
384 warn "WARNING: query problem with $query_to_use " . $@;
385 }
386
387 # Concatenate the sort_by limits and pass them to the results object
388 # Note: sort will override rank
389 my $sort_by;
390 foreach my $sort (@sort_by) {
391 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
392 $sort_by .= "1=1003 <i ";
393 }
394 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
395 $sort_by .= "1=1003 >i ";
396 }
397 elsif ( $sort eq "popularity_asc" ) {
398 $sort_by .= "1=9003 <i ";
399 }
400 elsif ( $sort eq "popularity_dsc" ) {
401 $sort_by .= "1=9003 >i ";
402 }
403 elsif ( $sort eq "call_number_asc" ) {
404 $sort_by .= "1=8007 <i ";
405 }
406 elsif ( $sort eq "call_number_dsc" ) {
407 $sort_by .= "1=8007 >i ";
408 }
409 elsif ( $sort eq "pubdate_asc" ) {
410 $sort_by .= "1=31 <i ";
411 }
412 elsif ( $sort eq "pubdate_dsc" ) {
413 $sort_by .= "1=31 >i ";
414 }
415 elsif ( $sort eq "acqdate_asc" ) {
416 $sort_by .= "1=32 <i ";
417 }
418 elsif ( $sort eq "acqdate_dsc" ) {
419 $sort_by .= "1=32 >i ";
420 }
421 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
422 $sort_by .= "1=4 <i ";
423 }
424 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
425 $sort_by .= "1=4 >i ";
426 }
427 else {
428 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
429 }
430 }
431 if ( $sort_by && !$scan && $results[$i] ) {
432 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
433 warn "WARNING sort $sort_by failed";
434 }
435 }
436 } # finished looping through servers
437
438 # The big moment: asynchronously retrieve results from all servers
439 _ZOOM_event_loop(
440 \@zconns,
441 \@results,
442 sub {
443 my ( $i, $size ) = @_;
444 my $results_hash;
445
446 # loop through the results
447 $results_hash->{'hits'} = $size;
448 my $times;
449 if ( $offset + $results_per_page <= $size ) {
450 $times = $offset + $results_per_page;
451 }
452 else {
453 $times = $size;
454 }
455
456 for ( my $j = $offset ; $j < $times ; $j++ ) {
457 my $records_hash;
458 my $record;
459
460 ## Check if it's an index scan
461 if ($scan) {
462 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
463
464 # here we create a minimal MARC record and hand it off to the
465 # template just like a normal result ... perhaps not ideal, but
466 # it works for now
467 my $tmprecord = MARC::Record->new();
468 $tmprecord->encoding('UTF-8');
469 my $tmptitle;
470 my $tmpauthor;
471
472 # the minimal record in author/title (depending on MARC flavour)
473 if ( C4::Context->preference("marcflavour") eq
474 "UNIMARC" )
475 {
476 $tmptitle = MARC::Field->new(
477 '200', ' ', ' ',
478 a => $term,
479 f => $occ
480 );
481 $tmprecord->append_fields($tmptitle);
482 }
483 else {
484 $tmptitle =
485 MARC::Field->new( '245', ' ', ' ', a => $term, );
486 $tmpauthor =
487 MARC::Field->new( '100', ' ', ' ', a => $occ, );
488 $tmprecord->append_fields($tmptitle);
489 $tmprecord->append_fields($tmpauthor);
490 }
491 $results_hash->{'RECORDS'}[$j] =
492 $tmprecord->as_usmarc();
493 }
494
495 # not an index scan
496 else {
497 $record = $results[ $i - 1 ]->record($j)->raw();
498 # warn "RECORD $j:".$record;
499 $results_hash->{'RECORDS'}[$j] = $record;
500 }
501
502 }
503 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
504
505 # Fill the facets while we're looping, but only for the
506 # biblioserver and not for a scan
507 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
508 $facets_counter = GetFacets( $results[ $i - 1 ] );
509 $facets_info = _get_facets_info( $facets );
510 }
511
512 # BUILD FACETS
513 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
514 for my $link_value (
515 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
516 keys %$facets_counter
517 )
518 {
519 my $expandable;
520 my $number_of_facets;
521 my @this_facets_array;
522 for my $one_facet (
523 sort {
524 $facets_counter->{$link_value}
525 ->{$b} <=> $facets_counter->{$link_value}
526 ->{$a}
527 } keys %{ $facets_counter->{$link_value} }
528 )
529 {
530 $number_of_facets++;
531 if ( ( $number_of_facets <= 5 )
532 || ( $expanded_facet eq $link_value )
533 || ( $facets_info->{$link_value}->{'expanded'} )
534 )
535 {
536
537 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
538 my $facet_link_value = $one_facet;
539 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
540
541 # fix the length that will display in the label,
542 my $facet_label_value = $one_facet;
543 my $facet_max_length = C4::Context->preference(
544 'FacetLabelTruncationLength')
545 || 20;
546 $facet_label_value =
547 substr( $one_facet, 0, $facet_max_length )
548 . "..."
549 if length($facet_label_value) >
550 $facet_max_length;
551
552 # if it's a branch, label by the name, not the code,
553 if ( $link_value =~ /branch/ ) {
554 if ( defined $branches
555 && ref($branches) eq "HASH"
556 && defined $branches->{$one_facet}
557 && ref( $branches->{$one_facet} ) eq
558 "HASH" )
559 {
560 $facet_label_value =
561 $branches->{$one_facet}
562 ->{'branchname'};
563 }
564 else {
565 $facet_label_value = "*";
566 }
567 }
568
569 # if it's a itemtype, label by the name, not the code,
570 if ( $link_value =~ /itype/ ) {
571 if ( defined $itemtypes
572 && ref($itemtypes) eq "HASH"
573 && defined $itemtypes->{$one_facet}
574 && ref( $itemtypes->{$one_facet} ) eq
575 "HASH" )
576 {
577 $facet_label_value =
578 $itemtypes->{$one_facet}
579 ->{translated_description};
580 }
581 }
582
583 # also, if it's a location code, use the name instead of the code
584 if ( $link_value =~ /location/ ) {
585 # TODO Retrieve all authorised values at once, instead of 1 query per entry
586 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
587 $facet_label_value = $av->count ? $av->next->opac_description : '';
588 }
589
590 # also, if it's a collection code, use the name instead of the code
591 if ( $link_value =~ /ccode/ ) {
592 # TODO Retrieve all authorised values at once, instead of 1 query per entry
593 my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet });
594 $facet_label_value = $av->count ? $av->next->opac_description : '';
595 }
596
597 # but we're down with the whole label being in the link's title.
598 push @this_facets_array,
599 {
600 facet_count =>
601 $facets_counter->{$link_value}
602 ->{$one_facet},
603 facet_label_value => $facet_label_value,
604 facet_title_value => $one_facet,
605 facet_link_value => $facet_link_value,
606 type_link_value => $link_value,
607 }
608 if ($facet_label_value);
609 }
610 }
611
612 # handle expanded option
613 unless ( $facets_info->{$link_value}->{'expanded'} ) {
614 $expandable = 1
615 if ( ( $number_of_facets > 5 )
616 && ( $expanded_facet ne $link_value ) );
617 }
618 push @facets_loop,
619 {
620 type_link_value => $link_value,
621 type_id => $link_value . "_id",
622 "type_label_"
623 . $facets_info->{$link_value}->{'label_value'} =>
624 1,
625 facets => \@this_facets_array,
626 expandable => $expandable,
627 expand => $link_value,
628 }
629 unless (
630 (
631 $facets_info->{$link_value}->{'label_value'} =~
632 /Libraries/
633 )
634 and ( Koha::Libraries->search->count == 1 )
635 );
636 }
637 }
638 }
639 );
640
641 # This sorts the facets into alphabetical order
642 if (@facets_loop) {
643 foreach my $f (@facets_loop) {
644 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
645 }
646 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
647 }
648
649 return ( undef, $results_hashref, \@facets_loop );
650 }
651
652 sub GetFacets {
653
654 my $rs = shift;
655 my $facets;
656
657 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
658
659 if ( $use_zebra_facets ) {
660 $facets = _get_facets_from_zebra( $rs );
661 } else {
662 $facets = _get_facets_from_records( $rs );
663 }
664
665 return $facets;
666 }
667
668 sub _get_facets_from_records {
669
670 my $rs = shift;
671
672 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
673 my $facets_config = getFacets();
674 my $facets = {};
675 my $size = $rs->size();
676 my $jmax = $size > $facets_maxrecs
677 ? $facets_maxrecs
678 : $size;
679
680 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
681
682 my $marc_record = new_record_from_zebra (
683 'biblioserver',
684 $rs->record( $j )->raw()
685 );
686
687 if ( ! defined $marc_record ) {
688 warn "ERROR DECODING RECORD - $@: " .
689 $rs->record( $j )->raw();
690 next;
691 }
692
693 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
694 }
695
696 return $facets;
697 }
698
699 =head2 _get_facets_data_from_record
700
701 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
702
703 Internal function that extracts facets information from a MARC::Record object
704 and populates $facets_counter for using in getRecords.
705
706 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
707 facets for Zebra).
708
709 =cut
710
711 sub _get_facets_data_from_record {
712
713 my ( $marc_record, $facets, $facets_counter ) = @_;
714
715 for my $facet (@$facets) {
716
717 my @used_datas = ();
718
719 foreach my $tag ( @{ $facet->{ tags } } ) {
720
721 # tag number is the first three digits
722 my $tag_num = substr( $tag, 0, 3 );
723 # subfields are the remainder
724 my $subfield_letters = substr( $tag, 3 );
725
726 my @fields = $marc_record->field( $tag_num );
727 foreach my $field (@fields) {
728 # If $field->indicator(1) eq 'z', it means it is a 'see from'
729 # field introduced because of IncludeSeeFromInSearches, so skip it
730 next if $field->indicator(1) eq 'z';
731
732 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
733
734 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
735 push @used_datas, $data;
736 $facets_counter->{ $facet->{ idx } }->{ $data }++;
737 }
738 }
739 }
740 }
741 }
742
743 =head2 _get_facets_from_zebra
744
745 my $facets = _get_facets_from_zebra( $result_set )
746
747 Retrieves facets for a specified result set. It loops through the facets defined
748 in C4::Koha::getFacets and returns a hash with the following structure:
749
750 { facet_idx => {
751 facet_value => count
752 },
753 ...
754 }
755
756 =cut
757
758 sub _get_facets_from_zebra {
759
760 my $rs = shift;
761
762 # save current elementSetName
763 my $elementSetName = $rs->option( 'elementSetName' );
764
765 my $facets_loop = getFacets();
766 my $facets_data = {};
767 # loop through defined facets and fill the facets hashref
768 foreach my $facet ( @$facets_loop ) {
769
770 my $idx = $facet->{ idx };
771 my $sep = $facet->{ sep };
772 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
773 if ( $facet_values ) {
774 # we've actually got a result
775 $facets_data->{ $idx } = $facet_values;
776 }
777 }
778 # set elementSetName to its previous value to avoid side effects
779 $rs->option( elementSetName => $elementSetName );
780
781 return $facets_data;
782 }
783
784 =head2 _get_facet_from_result_set
785
786 my $facet_values =
787 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
788
789 Internal function that extracts facet information for a specific index ($facet_idx) and
790 returns a hash containing facet values and count:
791
792 {
793 $facet_value => $count ,
794 ...
795 }
796
797 Warning: this function has the side effect of changing the elementSetName for the result
798 set. It is a helper function for the main loop, which takes care of backing it up for
799 restoring.
800
801 =cut
802
803 sub _get_facet_from_result_set {
804
805 my $facet_idx = shift;
806 my $rs = shift;
807 my $sep = shift;
808
809 my $internal_sep = '<*>';
810 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
811
812 return if ( ! defined $facet_idx || ! defined $rs );
813 # zebra's facet element, untokenized index
814 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
815 # configure zebra results for retrieving the desired facet
816 $rs->option( elementSetName => $facet_element );
817 # get the facet record from result set
818 my $facet = $rs->record( 0 )->raw;
819 # if the facet has no restuls...
820 return if !defined $facet;
821 # TODO: benchmark DOM vs. SAX performance
822 my $facet_dom = XML::LibXML->load_xml(
823 string => ($facet)
824 );
825 my @terms = $facet_dom->getElementsByTagName('term');
826 return if ! @terms;
827
828 my $facets = {};
829 foreach my $term ( @terms ) {
830 my $facet_value = $term->textContent;
831 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
832 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
833 }
834
835 return $facets;
836 }
837
838 =head2 _get_facets_info
839
840 my $facets_info = C4::Search::_get_facets_info( $facets )
841
842 Internal function that extracts facets information and properly builds
843 the data structure needed to render facet labels.
844
845 =cut
846
847 sub _get_facets_info {
848
849 my $facets = shift;
850
851 my $facets_info = {};
852
853 for my $facet ( @$facets ) {
854 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
855 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
856 }
857
858 return $facets_info;
859 }
860
861 sub pazGetRecords {
862 my (
863 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
864 $results_per_page, $offset, $expanded_facet, $branches,
865 $query_type, $scan
866 ) = @_;
867 #NOTE: Parameter $branches is not used here !
868
869 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
870 $paz->init();
871 $paz->search($simple_query);
872 sleep 1; # FIXME: WHY?
873
874 # do results
875 my $results_hashref = {};
876 my $stats = XMLin($paz->stat);
877 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
878
879 # for a grouped search result, the number of hits
880 # is the number of groups returned; 'bib_hits' will have
881 # the total number of bibs.
882 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
883 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
884
885 HIT: foreach my $hit (@{ $results->{'hit'} }) {
886 my $recid = $hit->{recid}->[0];
887
888 my $work_title = $hit->{'md-work-title'}->[0];
889 my $work_author;
890 if (exists $hit->{'md-work-author'}) {
891 $work_author = $hit->{'md-work-author'}->[0];
892 }
893 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
894
895 my $result_group = {};
896 $result_group->{'group_label'} = $group_label;
897 $result_group->{'group_merge_key'} = $recid;
898
899 my $count = 1;
900 if (exists $hit->{count}) {
901 $count = $hit->{count}->[0];
902 }
903 $result_group->{'group_count'} = $count;
904
905 for (my $i = 0; $i < $count; $i++) {
906 # FIXME -- may need to worry about diacritics here
907 my $rec = $paz->record($recid, $i);
908 push @{ $result_group->{'RECORDS'} }, $rec;
909 }
910
911 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
912 }
913
914 # pass through facets
915 my $termlist_xml = $paz->termlist('author,subject');
916 my $terms = XMLin($termlist_xml, forcearray => 1);
917 my @facets_loop = ();
918 #die Dumper($results);
919 # foreach my $list (sort keys %{ $terms->{'list'} }) {
920 # my @facets = ();
921 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
922 # push @facets, {
923 # facet_label_value => $facet->{'name'}->[0],
924 # };
925 # }
926 # push @facets_loop, ( {
927 # type_label => $list,
928 # facets => \@facets,
929 # } );
930 # }
931
932 return ( undef, $results_hashref, \@facets_loop );
933 }
934
935 # TRUNCATION
936 sub _detect_truncation {
937 my ( $operand, $index ) = @_;
938 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
939 @regexpr );
940 $operand =~ s/^ //g;
941 my @wordlist = split( /\s/, $operand );
942 foreach my $word (@wordlist) {
943 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
944 push @rightlefttruncated, $word;
945 }
946 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
947 push @lefttruncated, $word;
948 }
949 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
950 push @righttruncated, $word;
951 }
952 elsif ( index( $word, "*" ) < 0 ) {
953 push @nontruncated, $word;
954 }
955 else {
956 push @regexpr, $word;
957 }
958 }
959 return (
960 \@nontruncated, \@righttruncated, \@lefttruncated,
961 \@rightlefttruncated, \@regexpr
962 );
963 }
964
965 # STEMMING
966 sub _build_stemmed_operand {
967 my ($operand,$lang) = @_;
968 require Lingua::Stem::Snowball ;
969 my $stemmed_operand=q{};
970
971 # If operand contains a digit, it is almost certainly an identifier, and should
972 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
973 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
974 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
975 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
976 return $operand if $operand =~ /\d/;
977
978 # FIXME: the locale should be set based on the user's language and/or search choice
979 #warn "$lang";
980 # Make sure we only use the first two letters from the language code
981 $lang = lc(substr($lang, 0, 2));
982 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
983 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
984 if ($lang eq 'nb' || $lang eq 'nn') {
985 $lang = 'no';
986 }
987 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
988 encoding => "UTF-8" );
989
990 my @words = split( / /, $operand );
991 my @stems = $stemmer->stem(\@words);
992 for my $stem (@stems) {
993 $stemmed_operand .= "$stem";
994 $stemmed_operand .= "?"
995 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
996 $stemmed_operand .= " ";
997 }
998 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
999 return $stemmed_operand;
1000 }
1001
1002 # FIELD WEIGHTING
1003 sub _build_weighted_query {
1004
1005 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1006 # pretty well but could work much better if we had a smarter query parser
1007 my ( $operand, $stemmed_operand, $index ) = @_;
1008 my $stemming = C4::Context->preference("QueryStemming") || 0;
1009 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1010 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1011 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1012
1013 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1014
1015 # Keyword, or, no index specified
1016 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1017 $weighted_query .=
1018 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1019 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1020 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1021 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1022 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1023 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1024 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1025 if $fuzzy_enabled; # add fuzzy, word list
1026 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1027 if ( $stemming and $stemmed_operand )
1028 ; # add stemming, right truncation
1029 $weighted_query .= " or wrdl,r9=\"$operand\"";
1030
1031 # embedded sorting: 0 a-z; 1 z-a
1032 # $weighted_query .= ") or (sort1,aut=1";
1033 }
1034
1035 # Barcode searches should skip this process
1036 elsif ( $index eq 'bc' ) {
1037 $weighted_query .= "bc=\"$operand\"";
1038 }
1039
1040 # Authority-number searches should skip this process
1041 elsif ( $index eq 'an' ) {
1042 $weighted_query .= "an=\"$operand\"";
1043 }
1044
1045 # If the index is numeric, don't autoquote it.
1046 elsif ( $index =~ /,st-numeric$/ ) {
1047 $weighted_query .= " $index=$operand";
1048 }
1049
1050 # If the index already has more than one qualifier, wrap the operand
1051 # in quotes and pass it back (assumption is that the user knows what they
1052 # are doing and won't appreciate us mucking up their query
1053 elsif ( $index =~ ',' ) {
1054 $weighted_query .= " $index=\"$operand\"";
1055 }
1056
1057 #TODO: build better cases based on specific search indexes
1058 else {
1059 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1060 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1061 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1062 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1063 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1064 if $fuzzy_enabled; # add fuzzy, word list
1065 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1066 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1067 }
1068
1069 $weighted_query .= "))"; # close rank specification
1070 return $weighted_query;
1071 }
1072
1073 =head2 getIndexes
1074
1075 Return an array with available indexes.
1076
1077 =cut
1078
1079 sub getIndexes{
1080 my @indexes = (
1081 # biblio indexes
1082 'ab',
1083 'Abstract',
1084 'acqdate',
1085 'allrecords',
1086 'an',
1087 'Any',
1088 'at',
1089 'au',
1090 'aub',
1091 'aud',
1092 'audience',
1093 'auo',
1094 'aut',
1095 'Author',
1096 'Author-in-order ',
1097 'Author-personal-bibliography',
1098 'Authority-Number',
1099 'authtype',
1100 'bc',
1101 'Bib-level',
1102 'biblionumber',
1103 'bio',
1104 'biography',
1105 'callnum',
1106 'cfn',
1107 'Chronological-subdivision',
1108 'cn-bib-source',
1109 'cn-bib-sort',
1110 'cn-class',
1111 'cn-item',
1112 'cn-prefix',
1113 'cn-suffix',
1114 'cpn',
1115 'Code-institution',
1116 'Conference-name',
1117 'Conference-name-heading',
1118 'Conference-name-see',
1119 'Conference-name-seealso',
1120 'Content-type',
1121 'Control-number',
1122 'copydate',
1123 'Corporate-name',
1124 'Corporate-name-heading',
1125 'Corporate-name-see',
1126 'Corporate-name-seealso',
1127 'Country-publication',
1128 'ctype',
1129 'curriculum',
1130 'date-entered-on-file',
1131 'Date-of-acquisition',
1132 'Date-of-publication',
1133 'Date-time-last-modified',
1134 'Dewey-classification',
1135 'Dissertation-information',
1136 'diss',
1137 'dtlm',
1138 'EAN',
1139 'extent',
1140 'fic',
1141 'fiction',
1142 'Form-subdivision',
1143 'format',
1144 'Geographic-subdivision',
1145 'he',
1146 'Heading',
1147 'Heading-use-main-or-added-entry',
1148 'Heading-use-series-added-entry ',
1149 'Heading-use-subject-added-entry',
1150 'Host-item',
1151 'id-other',
1152 'ident',
1153 'Identifier-standard',
1154 'Illustration-code',
1155 'Index-term-genre',
1156 'Index-term-uncontrolled',
1157 'Interest-age-level',
1158 'Interest-grade-level',
1159 'ISBN',
1160 'isbn',
1161 'ISSN',
1162 'issn',
1163 'itemtype',
1164 'kw',
1165 'Koha-Auth-Number',
1166 'l-format',
1167 'language',
1168 'language-original',
1169 'lc-card',
1170 'LC-card-number',
1171 'lcn',
1172 'lex',
1173 'lexile-number',
1174 'llength',
1175 'ln',
1176 'ln-audio',
1177 'ln-subtitle',
1178 'Local-classification',
1179 'Local-number',
1180 'Match-heading',
1181 'Match-heading-see-from',
1182 'Material-type',
1183 'mc-itemtype',
1184 'mc-rtype',
1185 'mus',
1186 'name',
1187 'Music-number',
1188 'Name-geographic',
1189 'Name-geographic-heading',
1190 'Name-geographic-see',
1191 'Name-geographic-seealso',
1192 'nb',
1193 'Note',
1194 'notes',
1195 'ns',
1196 'nt',
1197 'Other-control-number',
1198 'pb',
1199 'Personal-name',
1200 'Personal-name-heading',
1201 'Personal-name-see',
1202 'Personal-name-seealso',
1203 'pl',
1204 'Place-publication',
1205 'pn',
1206 'popularity',
1207 'pubdate',
1208 'Publisher',
1209 'Provider',
1210 'pv',
1211 'Reading-grade-level',
1212 'Record-control-number',
1213 'rcn',
1214 'Record-type',
1215 'rtype',
1216 'se',
1217 'See',
1218 'See-also',
1219 'sn',
1220 'Stock-number',
1221 'su',
1222 'Subject',
1223 'Subject-heading-thesaurus',
1224 'Subject-name-personal',
1225 'Subject-subdivision',
1226 'Summary',
1227 'Suppress',
1228 'su-geo',
1229 'su-na',
1230 'su-to',
1231 'su-ut',
1232 'ut',
1233 'Term-genre-form',
1234 'Term-genre-form-heading',
1235 'Term-genre-form-see',
1236 'Term-genre-form-seealso',
1237 'ti',
1238 'Title',
1239 'Title-cover',
1240 'Title-series',
1241 'Title-uniform',
1242 'Title-uniform-heading',
1243 'Title-uniform-see',
1244 'Title-uniform-seealso',
1245 'totalissues',
1246 'yr',
1247
1248 # items indexes
1249 'acqsource',
1250 'barcode',
1251 'bc',
1252 'branch',
1253 'ccode',
1254 'classification-source',
1255 'cn-sort',
1256 'coded-location-qualifier',
1257 'copynumber',
1258 'damaged',
1259 'datelastborrowed',
1260 'datelastseen',
1261 'holdingbranch',
1262 'homebranch',
1263 'issues',
1264 'item',
1265 'itemnumber',
1266 'itype',
1267 'Local-classification',
1268 'location',
1269 'lost',
1270 'materials-specified',
1271 'mc-ccode',
1272 'mc-itype',
1273 'mc-loc',
1274 'notforloan',
1275 'Number-local-acquisition',
1276 'onloan',
1277 'price',
1278 'renewals',
1279 'replacementprice',
1280 'replacementpricedate',
1281 'reserves',
1282 'restricted',
1283 'stack',
1284 'stocknumber',
1285 'inv',
1286 'uri',
1287 'withdrawn',
1288
1289 # subject related
1290 );
1291
1292 return \@indexes;
1293 }
1294
1295 =head2 _handle_exploding_index
1296
1297 my $query = _handle_exploding_index($index, $term)
1298
1299 Callback routine to generate the search for "exploding" indexes (i.e.
1300 those indexes which are turned into multiple or-connected searches based
1301 on authority data).
1302
1303 =cut
1304
1305 sub _handle_exploding_index {
1306 my ($QParser, $filter, $params, $negate, $server) = @_;
1307 my $index = $filter;
1308 my $term = join(' ', @$params);
1309
1310 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1311
1312 my $marcflavour = C4::Context->preference('marcflavour');
1313
1314 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1315 my $wantedcodes = '';
1316 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1317 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1318 foreach my $auth (@$results) {
1319 my $record = MARC::Record->new_from_usmarc($auth);
1320 my @references = $record->field('5..');
1321 if (@references) {
1322 if ($index eq 'su-br') {
1323 $wantedcodes = 'g';
1324 } elsif ($index eq 'su-na') {
1325 $wantedcodes = 'h';
1326 } elsif ($index eq 'su-rl') {
1327 $wantedcodes = '';
1328 }
1329 foreach my $reference (@references) {
1330 my $codes = $reference->subfield($codesubfield);
1331 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1332 }
1333 }
1334 }
1335 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1336 return $query;
1337 }
1338
1339 =head2 parseQuery
1340
1341 ( $operators, $operands, $indexes, $limits,
1342 $sort_by, $scan, $lang ) =
1343 parseQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1344
1345 Shim function to ease the transition from buildQuery to a new QueryParser.
1346 This function is called at the beginning of buildQuery, and modifies
1347 buildQuery's input. If it can handle the input, it returns a query that
1348 buildQuery will not try to parse.
1349
1350 =cut
1351
1352 sub parseQuery {
1353 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1354
1355 my @operators = $operators ? @$operators : ();
1356 my @indexes = $indexes ? @$indexes : ();
1357 my @operands = $operands ? @$operands : ();
1358 my @limits = $limits ? @$limits : ();
1359 my @sort_by = $sort_by ? @$sort_by : ();
1360
1361 my $query = $operands[0];
1362 my $index;
1363 my $term;
1364 my $query_desc;
1365
1366 my $QParser;
1367 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1368 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1369 undef $QParser if (scalar @limits > 0);
1370
1371 if ($QParser)
1372 {
1373 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1374 $query = '';
1375 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1376 next unless $operands[$ii];
1377 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1378 if ($query);
1379 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1380 $query .= $operands[$ii];
1381 }
1382 elsif ( $indexes[$ii] =~ m/su-/ ) {
1383 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1384 }
1385 else {
1386 $query .=
1387 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1388 }
1389 }
1390 foreach my $limit (@limits) {
1391 }
1392 if ( scalar(@sort_by) > 0 ) {
1393 my $modifier_re =
1394 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1395 $query =~ s/$modifier_re//g;
1396 foreach my $modifier (@sort_by) {
1397 $query .= " #$modifier";
1398 }
1399 }
1400
1401 $query_desc = $query;
1402 $query_desc =~ s/\s+/ /g;
1403 if ( C4::Context->preference("QueryWeightFields") ) {
1404 }
1405 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1406 { 'target_syntax_callback' => \&_handle_exploding_index } );
1407 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1408 { 'target_syntax_callback' => \&_handle_exploding_index } );
1409 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1410 { 'target_syntax_callback' => \&_handle_exploding_index } );
1411 $QParser->parse($query);
1412 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1413 }
1414 else {
1415 require Koha::QueryParser::Driver::PQF;
1416 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1417 s/$modifier_re//g for @operands;
1418 }
1419
1420 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1421 }
1422
1423 =head2 buildQuery
1424
1425 ( $error, $query,
1426 $simple_query, $query_cgi,
1427 $query_desc, $limit,
1428 $limit_cgi, $limit_desc,
1429 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1430
1431 Build queries and limits in CCL, CGI, Human,
1432 handle truncation, stemming, field weighting, fuzziness, etc.
1433
1434 See verbose embedded documentation.
1435
1436
1437 =cut
1438
1439 sub buildQuery {
1440 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1441
1442 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1443
1444 my $query_desc;
1445 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1446
1447 # dereference
1448 my @operators = $operators ? @$operators : ();
1449 my @indexes = $indexes ? @$indexes : ();
1450 my @operands = $operands ? @$operands : ();
1451 my @limits = $limits ? @$limits : ();
1452 my @sort_by = $sort_by ? @$sort_by : ();
1453
1454 my $stemming = C4::Context->preference("QueryStemming") || 0;
1455 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1456 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1457 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1458
1459 my $query = $operands[0];
1460 my $simple_query = $operands[0];
1461
1462 # initialize the variables we're passing back
1463 my $query_cgi;
1464 my $query_type;
1465
1466 my $limit;
1467 my $limit_cgi;
1468 my $limit_desc;
1469
1470 my $cclq = 0;
1471 my $cclindexes = getIndexes();
1472 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1473 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1474 my $dx = lc($1);
1475 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1476 }
1477 $query = "ccl=$query" if $cclq;
1478 }
1479
1480 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1481 # DIAGNOSTIC ONLY!!
1482 if ( $query =~ /^ccl=/ ) {
1483 my $q=$';
1484 # This is needed otherwise ccl= and &limit won't work together, and
1485 # this happens when selecting a subject on the opac-detail page
1486 @limits = grep {!/^$/} @limits;
1487 my $original_q = $q; # without available part
1488 unless ( grep { /^available$/ } @limits ) {
1489 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1490 $original_q = $q;
1491 }
1492 if ( @limits ) {
1493 if ( grep { /^available$/ } @limits ) {
1494 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1495 delete $limits['available'];
1496 }
1497 $q .= ' and '.join(' and ', @limits) if @limits;
1498 }
1499 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1500 }
1501 if ( $query =~ /^cql=/ ) {
1502 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1503 }
1504 if ( $query =~ /^pqf=/ ) {
1505 if ($query_desc) {
1506 $query_cgi = "q=".uri_escape_utf8($query_desc);
1507 } else {
1508 $query_desc = $';
1509 $query_cgi = "q=pqf=".uri_escape_utf8($');
1510 }
1511 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1512 }
1513
1514 # pass nested queries directly
1515 # FIXME: need better handling of some of these variables in this case
1516 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1517 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1518 # if ( $query =~ /(\(|\))/ ) {
1519 # return (
1520 # undef, $query, $simple_query, $query_cgi,
1521 # $query, $limit, $limit_cgi, $limit_desc,
1522 # 'ccl'
1523 # );
1524 # }
1525
1526 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1527 # query operands and indexes and add stemming, truncation, field weighting, etc.
1528 # Once we do so, we'll end up with a value in $query, just like if we had an
1529 # incoming $query from the user
1530 else {
1531 $query = ""
1532 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1533 my $previous_operand
1534 ; # a flag used to keep track if there was a previous query
1535 # if there was, we can apply the current operator
1536 # for every operand
1537 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1538
1539 # COMBINE OPERANDS, INDEXES AND OPERATORS
1540 if ( ($operands[$i] // '') ne '' ) {
1541 $operands[$i]=~s/^\s+//;
1542
1543 # A flag to determine whether or not to add the index to the query
1544 my $indexes_set;
1545
1546 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1547 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1548 $weight_fields = 0;
1549 $stemming = 0;
1550 } else {
1551 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1552 }
1553 my $operand = $operands[$i];
1554 my $index = $indexes[$i];
1555
1556 # Add index-specific attributes
1557
1558 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1559 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1560 #which is processed higher up in this sub. Other than that, year searches are typically
1561 #handled as limits which are not processed her either.
1562
1563 # Search ranges: Date of Publication, st-numeric
1564 if ( $index =~ /(yr|st-numeric)/ ) {
1565 #weight_fields/relevance search causes errors with date ranges
1566 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1567 #In the case of YYYY-YYYY, it will return no results
1568 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1569 }
1570
1571 # Date of Acquisition
1572 elsif ( $index =~ /acqdate/ ) {
1573 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1574 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1575 #top of the results just because they have lots of item records matching that date.
1576 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1577 #irrelevant here
1578 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1579 }
1580 # ISBN,ISSN,Standard Number, don't need special treatment
1581 elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) {
1582 (
1583 $stemming, $auto_truncation,
1584 $weight_fields, $fuzzy_enabled
1585 ) = ( 0, 0, 0, 0 );
1586
1587 if ( $index eq 'nb' ) {
1588 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1589 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1590 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1591 $indexes[$i] = $index = '';
1592 }
1593 }
1594 }
1595
1596 if(not $index){
1597 $index = 'kw';
1598 }
1599
1600 # Set default structure attribute (word list)
1601 my $struct_attr = q{};
1602 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1603 $struct_attr = ",wrdl";
1604 }
1605
1606 # Some helpful index variants
1607 my $index_plus = $index . $struct_attr . ':';
1608 my $index_plus_comma = $index . $struct_attr . ',';
1609
1610 if ($auto_truncation){
1611 unless ( $index =~ /,(st-|phr|ext)/ ) {
1612 #FIXME only valid with LTR scripts
1613 $operand=join(" ",map{
1614 (index($_,"*")>0?"$_":"$_*")
1615 }split (/\s+/,$operand));
1616 warn $operand if $DEBUG;
1617 }
1618 }
1619
1620 # Detect Truncation
1621 my $truncated_operand;
1622 my( $nontruncated, $righttruncated, $lefttruncated,
1623 $rightlefttruncated, $regexpr
1624 ) = _detect_truncation( $operand, $index );
1625 warn
1626 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1627 if $DEBUG;
1628
1629 # Apply Truncation
1630 if (
1631 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1632 scalar(@$rightlefttruncated) > 0 )
1633 {
1634
1635 # Don't field weight or add the index to the query, we do it here
1636 $indexes_set = 1;
1637 undef $weight_fields;
1638 my $previous_truncation_operand;
1639 if (scalar @$nontruncated) {
1640 $truncated_operand .= "$index_plus @$nontruncated ";
1641 $previous_truncation_operand = 1;
1642 }
1643 if (scalar @$righttruncated) {
1644 $truncated_operand .= "and " if $previous_truncation_operand;
1645 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1646 $previous_truncation_operand = 1;
1647 }
1648 if (scalar @$lefttruncated) {
1649 $truncated_operand .= "and " if $previous_truncation_operand;
1650 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1651 $previous_truncation_operand = 1;
1652 }
1653 if (scalar @$rightlefttruncated) {
1654 $truncated_operand .= "and " if $previous_truncation_operand;
1655 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1656 $previous_truncation_operand = 1;
1657 }
1658 }
1659 $operand = $truncated_operand if $truncated_operand;
1660 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1661
1662 # Handle Stemming
1663 my $stemmed_operand;
1664 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1665 if $stemming;
1666
1667 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1668
1669 # Handle Field Weighting
1670 my $weighted_operand;
1671 if ($weight_fields) {
1672 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1673 $operand = $weighted_operand;
1674 $indexes_set = 1;
1675 }
1676
1677 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1678
1679 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1680 query => $query,
1681 query_cgi => $query_cgi,
1682 query_desc => $query_desc,
1683 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1684 parsed_operand => $operand,
1685 original_operand => $operands[$i] // '',
1686 index => $index,
1687 index_plus => $index_plus,
1688 indexes_set => $indexes_set,
1689 previous_operand => $previous_operand,
1690 });
1691
1692 } #/if $operands
1693 } # /for
1694 }
1695 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1696
1697 # add limits
1698 my %group_OR_limits;
1699 my $availability_limit;
1700 foreach my $this_limit (@limits) {
1701 next unless $this_limit;
1702 if ( $this_limit =~ /available/ ) {
1703 #
1704 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1705 ## In English:
1706 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1707 $availability_limit .=
1708 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1709 $limit_cgi .= "&limit=available";
1710 $limit_desc .= "";
1711 }
1712
1713 # group_OR_limits, prefixed by mc-
1714 # OR every member of the group
1715 elsif ( $this_limit =~ /mc/ ) {
1716 my ($k,$v) = split(/:/, $this_limit,2);
1717 if ( $k !~ /mc-i(tem)?type/ ) {
1718 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1719 $this_limit =~ tr/"//d;
1720 $this_limit = $k.":'".$v."'";
1721 }
1722
1723 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1724 $limit_desc .= " or " if $group_OR_limits{$k};
1725 $group_OR_limits{$k} .= "$this_limit";
1726 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1727 $limit_desc .= " $this_limit";
1728 }
1729
1730 # Regular old limits
1731 else {
1732 $limit .= " and " if $limit || $query;
1733 $limit .= "$this_limit";
1734 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1735 if ($this_limit =~ /^branch:(.+)/) {
1736 my $branchcode = $1;
1737 my $library = Koha::Libraries->find( $branchcode );
1738 if (defined $library) {
1739 $limit_desc .= " branch:" . $library->branchname;
1740 } else {
1741 $limit_desc .= " $this_limit";
1742 }
1743 } else {
1744 $limit_desc .= " $this_limit";
1745 }
1746 }
1747 }
1748 foreach my $k (keys (%group_OR_limits)) {
1749 $limit .= " and " if ( $query || $limit );
1750 $limit .= "($group_OR_limits{$k})";
1751 }
1752 if ($availability_limit) {
1753 $limit .= " and " if ( $query || $limit );
1754 $limit .= "($availability_limit)";
1755 }
1756
1757 # Normalize the query and limit strings
1758 # This is flawed , means we can't search anything with : in it
1759 # if user wants to do ccl or cql, start the query with that
1760 # $query =~ s/:/=/g;
1761 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1762 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1763 $query =~ s/(?<=(wrdl)):/=/g;
1764 $query =~ s/(?<=(trn|phr)):/=/g;
1765 $query =~ s/(?<=(st-numeric)):/=/g;
1766 $query =~ s/(?<=(st-year)):/=/g;
1767 $query =~ s/(?<=(st-date-normalized)):/=/g;
1768 $limit =~ s/:/=/g;
1769 for ( $query, $query_desc, $limit, $limit_desc ) {
1770 s/ +/ /g; # remove extra spaces
1771 s/^ //g; # remove any beginning spaces
1772 s/ $//g; # remove any ending spaces
1773 s/==/=/g; # remove double == from query
1774 }
1775 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1776
1777 for ($query_cgi,$simple_query) {
1778 s/"//g;
1779 }
1780 # append the limit to the query
1781 $query .= " " . $limit;
1782
1783 # Warnings if DEBUG
1784 if ($DEBUG) {
1785 warn "QUERY:" . $query;
1786 warn "QUERY CGI:" . $query_cgi;
1787 warn "QUERY DESC:" . $query_desc;
1788 warn "LIMIT:" . $limit;
1789 warn "LIMIT CGI:" . $limit_cgi;
1790 warn "LIMIT DESC:" . $limit_desc;
1791 warn "---------\nLeave buildQuery\n---------";
1792 }
1793
1794 return (
1795 undef, $query, $simple_query, $query_cgi,
1796 $query_desc, $limit, $limit_cgi, $limit_desc,
1797 $query_type
1798 );
1799 }
1800
1801 =head2 _build_initial_query
1802
1803 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1804
1805 Build a section of the initial query containing indexes, operators, and operands.
1806
1807 =cut
1808
1809 sub _build_initial_query {
1810 my ($params) = @_;
1811
1812 my $operator = "";
1813 if ($params->{previous_operand}){
1814 #If there is a previous operand, add a supplied operator or the default 'and'
1815 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1816 }
1817
1818 #NOTE: indexes_set is typically set when doing truncation or field weighting
1819 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1820
1821 #e.g. "kw,wrdl:test"
1822 #e.g. " and kw,wrdl:test"
1823 $params->{query} .= $operator . $operand;
1824
1825 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1826 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1827 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1828
1829 #e.g. " and kw,wrdl: test"
1830 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1831
1832 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1833
1834 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1835 }
1836
1837 =head2 searchResults
1838
1839 my @search_results = searchResults($search_context, $searchdesc, $hits,
1840 $results_per_page, $offset, $scan,
1841 @marcresults);
1842
1843 Format results in a form suitable for passing to the template
1844
1845 =cut
1846
1847 # IMO this subroutine is pretty messy still -- it's responsible for
1848 # building the HTML output for the template
1849 sub searchResults {
1850 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1851 my $dbh = C4::Context->dbh;
1852 my @newresults;
1853
1854 require C4::Items;
1855
1856 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1857 my ($is_opac, $hidelostitems);
1858 if ($search_context eq 'opac') {
1859 $hidelostitems = C4::Context->preference('hidelostitems');
1860 $is_opac = 1;
1861 }
1862
1863 #Build branchnames hash
1864 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1865
1866 # FIXME - We build an authorised values hash here, using the default framework
1867 # though it is possible to have different authvals for different fws.
1868
1869 my $shelflocations =
1870 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1871
1872 # get notforloan authorised value list (see $shelflocations FIXME)
1873 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] });
1874 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1875
1876 #Get itemtype hash
1877 my $itemtypes = Koha::ItemTypes->search_with_localization;
1878 my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed };
1879
1880 #search item field code
1881 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1882
1883 ## find column names of items related to MARC
1884 my %subfieldstosearch;
1885 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1886 for my $column ( @columns ) {
1887 my ( $tagfield, $tagsubfield ) =
1888 &GetMarcFromKohaField( "items." . $column, "" );
1889 if ( defined $tagsubfield ) {
1890 $subfieldstosearch{$column} = $tagsubfield;
1891 }
1892 }
1893
1894 # handle which records to actually retrieve
1895 my $times;
1896 if ( $hits && $offset + $results_per_page <= $hits ) {
1897 $times = $offset + $results_per_page;
1898 }
1899 else {
1900 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1901 }
1902
1903 my $marcflavour = C4::Context->preference("marcflavour");
1904 # We get the biblionumber position in MARC
1905 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1906
1907 # set stuff for XSLT processing here once, not later again for every record we retrieved
1908 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1909 my $xslsyspref = $interface . "XSLTResultsDisplay";
1910 my $xslfile = C4::Context->preference($xslsyspref);
1911 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1912 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1913
1914 # loop through all of the records we've retrieved
1915 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1916
1917 my $marcrecord;
1918 if ($scan) {
1919 # For Scan searches we built USMARC data
1920 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1921 } else {
1922 # Normal search, render from Zebra's output
1923 $marcrecord = new_record_from_zebra(
1924 'biblioserver',
1925 $marcresults->[$i]
1926 );
1927
1928 if ( ! defined $marcrecord ) {
1929 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1930 next;
1931 }
1932 }
1933
1934 my $fw = $scan
1935 ? undef
1936 : $bibliotag < 10
1937 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1938 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1939
1940 SetUTF8Flag($marcrecord);
1941 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1942 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1943 $oldbiblio->{result_number} = $i + 1;
1944
1945 # add imageurl to itemtype if there is one
1946 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1947
1948 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1949 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1950 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1951 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1952 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1953
1954 # edition information, if any
1955 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1956 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1957 # Build summary if there is one (the summary is defined in the itemtypes table)
1958 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1959 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1960 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1961 my @fields = $marcrecord->fields();
1962
1963 my $newsummary;
1964 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1965 my $tags = {};
1966 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1967 $tag =~ /(.{3})(.)/;
1968 if($marcrecord->field($1)){
1969 my @abc = $marcrecord->field($1)->subfield($2);
1970 $tags->{$tag} = $#abc + 1 ;
1971 }
1972 }
1973
1974 # We catch how many times to repeat this line
1975 my $max = 0;
1976 foreach my $tag (keys(%$tags)){
1977 $max = $tags->{$tag} if($tags->{$tag} > $max);
1978 }
1979
1980 # we replace, and repeat each line
1981 for (my $i = 0 ; $i < $max ; $i++){
1982 my $newline = $line;
1983
1984 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1985 $tag =~ /(.{3})(.)/;
1986
1987 if($marcrecord->field($1)){
1988 my @repl = $marcrecord->field($1)->subfield($2);
1989 my $subfieldvalue = $repl[$i];
1990 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1991 }
1992 }
1993 $newsummary .= "$newline\n";
1994 }
1995 }
1996
1997 $newsummary =~ s/\[(.*?)]//g;
1998 $newsummary =~ s/\n/<br\/>/g;
1999 $oldbiblio->{summary} = $newsummary;
2000 }
2001
2002 # Pull out the items fields
2003 my @fields = $marcrecord->field($itemtag);
2004 my $marcflavor = C4::Context->preference("marcflavour");
2005 # adding linked items that belong to host records
2006 my $analyticsfield = '773';
2007 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2008 $analyticsfield = '773';
2009 } elsif ($marcflavor eq 'UNIMARC') {
2010 $analyticsfield = '461';
2011 }
2012 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2013 my $hostbiblionumber = $hostfield->subfield("0");
2014 my $linkeditemnumber = $hostfield->subfield("9");
2015 if( $hostbiblionumber ) {
2016 my $hostbiblio = GetMarcBiblio({
2017 biblionumber => $hostbiblionumber,
2018 embed_items => 1 });
2019 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2020 if( $hostbiblio ) {
2021 my @hostitems = $hostbiblio->field($itemfield);
2022 foreach my $hostitem (@hostitems){
2023 if ($hostitem->subfield("9") eq $linkeditemnumber){
2024 my $linkeditem =$hostitem;
2025 # append linked items if they exist
2026 push @fields, $linkeditem if $linkeditem;
2027 }
2028 }
2029 }
2030 }
2031 }
2032
2033 # Setting item statuses for display
2034 my @available_items_loop;
2035 my @onloan_items_loop;
2036 my @other_items_loop;
2037
2038 my $available_items;
2039 my $onloan_items;
2040 my $other_items;
2041
2042 my $ordered_count = 0;
2043 my $available_count = 0;
2044 my $onloan_count = 0;
2045 my $longoverdue_count = 0;
2046 my $other_count = 0;
2047 my $withdrawn_count = 0;
2048 my $itemlost_count = 0;
2049 my $hideatopac_count = 0;
2050 my $itembinding_count = 0;
2051 my $itemdamaged_count = 0;
2052 my $item_in_transit_count = 0;
2053 my $can_place_holds = 0;
2054 my $item_onhold_count = 0;
2055 my $notforloan_count = 0;
2056 my $items_count = scalar(@fields);
2057 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2058 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2059 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2060
2061 # loop through every item
2062 foreach my $field (@fields) {
2063 my $item;
2064
2065 # populate the items hash
2066 foreach my $code ( keys %subfieldstosearch ) {
2067 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2068 }
2069 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2070
2071 # OPAC hidden items
2072 if ($is_opac) {
2073 # hidden because lost
2074 if ($hidelostitems && $item->{itemlost}) {
2075 $hideatopac_count++;
2076 next;
2077 }
2078 # hidden based on OpacHiddenItems syspref
2079 my @hi = C4::Items::GetHiddenItemnumbers($item);
2080 if (scalar @hi) {
2081 push @hiddenitems, @hi;
2082 $hideatopac_count++;
2083 next;
2084 }
2085 }
2086
2087 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2088 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2089
2090 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2091 if ($item->{$hbranch}) {
2092 $item->{'branchname'} = $branches{$item->{$hbranch}};
2093 }
2094 elsif ($item->{$otherbranch}) { # Last resort
2095 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2096 }
2097
2098 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2099 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2100 my $userenv = C4::Context->userenv;
2101 if ( $item->{onloan}
2102 && $userenv
2103 && $userenv->{number}
2104 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $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, $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, $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, $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, $lang);
2247 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2248 }
2249
2250 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2251 if (!C4::Context->preference("item-level_itypes")) {
2252 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2253 $can_place_holds = 0;
2254 }
2255 }
2256 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2257 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2258 $oldbiblio->{items_count} = $items_count;
2259 $oldbiblio->{available_items_loop} = \@available_items_loop;
2260 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2261 $oldbiblio->{other_items_loop} = \@other_items_loop;
2262 $oldbiblio->{availablecount} = $available_count;
2263 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2264 $oldbiblio->{onloancount} = $onloan_count;
2265 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2266 $oldbiblio->{othercount} = $other_count;
2267 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2268 $oldbiblio->{withdrawncount} = $withdrawn_count;
2269 $oldbiblio->{itemlostcount} = $itemlost_count;
2270 $oldbiblio->{damagedcount} = $itemdamaged_count;
2271 $oldbiblio->{intransitcount} = $item_in_transit_count;
2272 $oldbiblio->{onholdcount} = $item_onhold_count;
2273 $oldbiblio->{orderedcount} = $ordered_count;
2274 $oldbiblio->{notforloancount} = $notforloan_count;
2275
2276 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2277 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2278 my $subfields = substr $fieldspec, 3;
2279 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2280 my @alternateholdingsinfo = ();
2281 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2282 my $alternateholdingscount = 0;
2283
2284 for my $field (@holdingsfields) {
2285 my %holding = ( holding => '' );
2286 my $havesubfield = 0;
2287 for my $subfield ($field->subfields()) {
2288 if ((index $subfields, $$subfield[0]) >= 0) {
2289 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2290 $holding{'holding'} .= $$subfield[1];
2291 $havesubfield++;
2292 }
2293 }
2294 if ($havesubfield) {
2295 push(@alternateholdingsinfo, \%holding);
2296 $alternateholdingscount++;
2297 }
2298 }
2299
2300 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2301 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2302 }
2303
2304 push( @newresults, $oldbiblio );
2305 }
2306
2307 return @newresults;
2308 }
2309
2310 =head2 enabled_staff_search_views
2311
2312 %hash = enabled_staff_search_views()
2313
2314 This function returns a hash that contains three flags obtained from the system
2315 preferences, used to determine whether a particular staff search results view
2316 is enabled.
2317
2318 =over 2
2319
2320 =item C<Output arg:>
2321
2322 * $hash{can_view_MARC} is true only if the MARC view is enabled
2323 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2324 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2325
2326 =item C<usage in the script:>
2327
2328 =back
2329
2330 $template->param ( C4::Search::enabled_staff_search_views );
2331
2332 =cut
2333
2334 sub enabled_staff_search_views
2335 {
2336 return (
2337 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2338 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2339 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2340 );
2341 }
2342
2343 =head2 z3950_search_args
2344
2345 $arrayref = z3950_search_args($matchpoints)
2346
2347 This function returns an array reference that contains the search parameters to be
2348 passed to the Z39.50 search script (z3950_search.pl). The array elements
2349 are hash refs whose keys are name and value, and whose values are the
2350 name of a search parameter, the value of that search parameter and the URL encoded
2351 value of that parameter.
2352
2353 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2354
2355 The search parameter values are obtained from the bibliographic record whose
2356 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2357
2358 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2359 a general purpose search argument. In this case, the returned array contains only
2360 entry: the key is 'title' and the value is derived from $matchpoints.
2361
2362 If a search parameter value is undefined or empty, it is not included in the returned
2363 array.
2364
2365 The returned array reference may be passed directly to the template parameters.
2366
2367 =over 2
2368
2369 =item C<Output arg:>
2370
2371 * $array containing hash refs as described above
2372
2373 =item C<usage in the script:>
2374
2375 =back
2376
2377 $data = Biblio::GetBiblioData($bibno);
2378 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2379
2380 *OR*
2381
2382 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2383
2384 =cut
2385
2386 sub z3950_search_args {
2387 my $bibrec = shift;
2388
2389 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2390 my $isbn = Business::ISBN->new( $isbn_string );
2391
2392 if (defined $isbn && $isbn->is_valid)
2393 {
2394 if ( ref($bibrec) ) {
2395 $bibrec->{isbn} = $isbn_string;
2396 $bibrec->{title} = undef;
2397 } else {
2398 $bibrec = { isbn => $isbn_string };
2399 }
2400 }
2401 else {
2402 $bibrec = { title => $bibrec } if !ref $bibrec;
2403 }
2404 my $array = [];
2405 for my $field (qw/ lccn isbn issn title author dewey subject /)
2406 {
2407 push @$array, { name => $field, value => $bibrec->{$field} }
2408 if defined $bibrec->{$field};
2409 }
2410 return $array;
2411 }
2412
2413 =head2 GetDistinctValues($field);
2414
2415 C<$field> is a reference to the fields array
2416
2417 =cut
2418
2419 sub GetDistinctValues {
2420 my ($fieldname,$string)=@_;
2421 # returns a reference to a hash of references to branches...
2422 if ($fieldname=~/\./){
2423 my ($table,$column)=split /\./, $fieldname;
2424 my $dbh = C4::Context->dbh;
2425 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2426 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 ");
2427 $sth->execute;
2428 my $elements=$sth->fetchall_arrayref({});
2429 return $elements;
2430 }
2431 else {
2432 $string||= qq("");
2433 my @servers=qw<biblioserver authorityserver>;
2434 my (@zconns,@results);
2435 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2436 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2437 $results[$i] =
2438 $zconns[$i]->scan(
2439 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2440 );
2441 }
2442 # The big moment: asynchronously retrieve results from all servers
2443 my @elements;
2444 _ZOOM_event_loop(
2445 \@zconns,
2446 \@results,
2447 sub {
2448 my ( $i, $size ) = @_;
2449 for ( my $j = 0 ; $j < $size ; $j++ ) {
2450 my %hashscan;
2451 @hashscan{qw(value cnt)} =
2452 $results[ $i - 1 ]->display_term($j);
2453 push @elements, \%hashscan;
2454 }
2455 }
2456 );
2457 return \@elements;
2458 }
2459 }
2460
2461 =head2 _ZOOM_event_loop
2462
2463 _ZOOM_event_loop(\@zconns, \@results, sub {
2464 my ( $i, $size ) = @_;
2465 ....
2466 } );
2467
2468 Processes a ZOOM event loop and passes control to a closure for
2469 processing the results, and destroying the resultsets.
2470
2471 =cut
2472
2473 sub _ZOOM_event_loop {
2474 my ($zconns, $results, $callback) = @_;
2475 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2476 my $ev = $zconns->[ $i - 1 ]->last_event();
2477 if ( $ev == ZOOM::Event::ZEND ) {
2478 next unless $results->[ $i - 1 ];
2479 my $size = $results->[ $i - 1 ]->size();
2480 if ( $size > 0 ) {
2481 $callback->($i, $size);
2482 }
2483 }
2484 }
2485
2486 foreach my $result (@$results) {
2487 $result->destroy();
2488 }
2489 }
2490
2491 =head2 new_record_from_zebra
2492
2493 Given raw data from a searchengine result set, return a MARC::Record object
2494
2495 This helper function is needed to take into account all the involved
2496 system preferences and configuration variables to properly create the
2497 MARC::Record object.
2498
2499 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2500 data. If we are using DOM, then it has to be MARCXML.
2501
2502 If we are using elasticsearch, it'll already be a MARC::Record and this
2503 function needs a new name.
2504
2505 =cut
2506
2507 sub new_record_from_zebra {
2508
2509 my $server = shift;
2510 my $raw_data = shift;
2511 # Set the default indexing modes
2512 my $search_engine = C4::Context->preference("SearchEngine");
2513 if ($search_engine eq 'Elasticsearch') {
2514 return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2515 }
2516 my $index_mode = ( $server eq 'biblioserver' )
2517 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2518 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2519
2520 my $marc_record = eval {
2521 if ( $index_mode eq 'dom' ) {
2522 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2523 } else {
2524 MARC::Record->new_from_usmarc( $raw_data );
2525 }
2526 };
2527
2528 if ($@) {
2529 return;
2530 } else {
2531 return $marc_record;
2532 }
2533
2534 }
2535
2536 END { } # module clean-up code here (global destructor)
2537
2538 1;
2539 __END__
2540
2541 =head1 AUTHOR
2542
2543 Koha Development Team <http://koha-community.org/>
2544
2545 =cut
Koha ILS