?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 17903: Fix possible SQL injection in serial claims
[koha.git] / C4 / Search.pm
blob4270c826217a5cf6d2252ec2708d3e3fd93def66
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::Libraries;
35 use Koha::Patrons;
36 use YAML;
37 use URI::Escape;
38 use Business::ISBN;
39 use MARC::Record;
40 use MARC::Field;
41 use vars qw(@ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
42
43 BEGIN {
44 $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
45 }
46
47 =head1 NAME
48
49 C4::Search - Functions for searching the Koha catalog.
50
51 =head1 SYNOPSIS
52
53 See opac/opac-search.pl or catalogue/search.pl for example of usage
54
55 =head1 DESCRIPTION
56
57 This module provides searching functions for Koha's bibliographic databases
58
59 =head1 FUNCTIONS
60
61 =cut
62
63 @ISA = qw(Exporter);
64 @EXPORT = qw(
65 &FindDuplicate
66 &SimpleSearch
67 &searchResults
68 &getRecords
69 &buildQuery
70 &GetDistinctValues
71 &enabled_staff_search_views
72 );
73
74 # make all your functions, whether exported or not;
75
76 =head2 FindDuplicate
77
78 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
79
80 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
81
82 =cut
83
84 sub FindDuplicate {
85 my ($record) = @_;
86 my $dbh = C4::Context->dbh;
87 my $result = TransformMarcToKoha( $record, '' );
88 my $sth;
89 my $query;
90 my $search;
91 my $type;
92 my ( $biblionumber, $title );
93
94 # search duplicate on ISBN, easy and fast..
95 # ... normalize first
96 if ( $result->{isbn} ) {
97 $result->{isbn} =~ s/\(.*$//;
98 $result->{isbn} =~ s/\s+$//;
99 $query = "isbn:$result->{isbn}";
100 }
101 else {
102 my $QParser;
103 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
104 my $titleindex;
105 my $authorindex;
106 my $op;
107
108 if ($QParser) {
109 $titleindex = 'title|exact';
110 $authorindex = 'author|exact';
111 $op = '&&';
112 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
113 } else {
114 $titleindex = 'ti,ext';
115 $authorindex = 'au,ext';
116 $op = 'and';
117 }
118
119 $result->{title} =~ s /\\//g;
120 $result->{title} =~ s /\"//g;
121 $result->{title} =~ s /\(//g;
122 $result->{title} =~ s /\)//g;
123
124 # FIXME: instead of removing operators, could just do
125 # quotes around the value
126 $result->{title} =~ s/(and|or|not)//g;
127 $query = "$titleindex:\"$result->{title}\"";
128 if ( $result->{author} ) {
129 $result->{author} =~ s /\\//g;
130 $result->{author} =~ s /\"//g;
131 $result->{author} =~ s /\(//g;
132 $result->{author} =~ s /\)//g;
133
134 # remove valid operators
135 $result->{author} =~ s/(and|or|not)//g;
136 $query .= " $op $authorindex:\"$result->{author}\"";
137 }
138 }
139
140 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
141 my @results;
142 if (!defined $error) {
143 foreach my $possible_duplicate_record (@{$searchresults}) {
144 my $marcrecord = new_record_from_zebra(
145 'biblioserver',
146 $possible_duplicate_record
147 );
148
149 my $result = TransformMarcToKoha( $marcrecord, '' );
150
151 # FIXME :: why 2 $biblionumber ?
152 if ($result) {
153 push @results, $result->{'biblionumber'};
154 push @results, $result->{'title'};
155 }
156 }
157 }
158 return @results;
159 }
160
161 =head2 SimpleSearch
162
163 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
164
165 This function provides a simple search API on the bibliographic catalog
166
167 =over 2
168
169 =item C<input arg:>
170
171 * $query can be a simple keyword or a complete CCL query
172 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
173 * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
174 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
175
176
177 =item C<Return:>
178
179 Returns an array consisting of three elements
180 * $error is undefined unless an error is detected
181 * $results is a reference to an array of records.
182 * $total_hits is the number of hits that would have been returned with no limit
183
184 If an error is returned the two other return elements are undefined. If error itself is undefined
185 the other two elements are always defined
186
187 =item C<usage in the script:>
188
189 =back
190
191 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
192
193 if (defined $error) {
194 $template->param(query_error => $error);
195 warn "error: ".$error;
196 output_html_with_http_headers $input, $cookie, $template->output;
197 exit;
198 }
199
200 my $hits = @{$marcresults};
201 my @results;
202
203 for my $r ( @{$marcresults} ) {
204 my $marcrecord = MARC::File::USMARC::decode($r);
205 my $biblio = TransformMarcToKoha($marcrecord,q{});
206
207 #build the iarray of hashs for the template.
208 push @results, {
209 title => $biblio->{'title'},
210 subtitle => $biblio->{'subtitle'},
211 biblionumber => $biblio->{'biblionumber'},
212 author => $biblio->{'author'},
213 publishercode => $biblio->{'publishercode'},
214 publicationyear => $biblio->{'publicationyear'},
215 };
216
217 }
218
219 $template->param(result=>\@results);
220
221 =cut
222
223 sub SimpleSearch {
224 my ( $query, $offset, $max_results, $servers ) = @_;
225
226 return ( 'No query entered', undef, undef ) unless $query;
227 # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
228 my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
229 my @zoom_queries;
230 my @tmpresults;
231 my @zconns;
232 my $results = [];
233 my $total_hits = 0;
234
235 my $QParser;
236 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
237 if ($QParser) {
238 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
239 }
240
241 # Initialize & Search Zebra
242 for ( my $i = 0 ; $i < @servers ; $i++ ) {
243 eval {
244 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
245 if ($QParser) {
246 $query =~ s/=/:/g;
247 $QParser->parse( $query );
248 $query = $QParser->target_syntax($servers[$i]);
249 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
250 } else {
251 $query =~ s/:/=/g;
252 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
253 }
254 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
255
256 # error handling
257 my $error =
258 $zconns[$i]->errmsg() . " ("
259 . $zconns[$i]->errcode() . ") "
260 . $zconns[$i]->addinfo() . " "
261 . $zconns[$i]->diagset();
262
263 return ( $error, undef, undef ) if $zconns[$i]->errcode();
264 };
265 if ($@) {
266
267 # caught a ZOOM::Exception
268 my $error =
269 $@->message() . " ("
270 . $@->code() . ") "
271 . $@->addinfo() . " "
272 . $@->diagset();
273 warn $error." for query: $query";
274 return ( $error, undef, undef );
275 }
276 }
277
278 _ZOOM_event_loop(
279 \@zconns,
280 \@tmpresults,
281 sub {
282 my ($i, $size) = @_;
283 my $first_record = defined($offset) ? $offset + 1 : 1;
284 my $hits = $tmpresults[ $i - 1 ]->size();
285 $total_hits += $hits;
286 my $last_record = $hits;
287 if ( defined $max_results && $offset + $max_results < $hits ) {
288 $last_record = $offset + $max_results;
289 }
290
291 for my $j ( $first_record .. $last_record ) {
292 my $record = eval {
293 $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
294 ; # 0 indexed
295 };
296 push @{$results}, $record if defined $record;
297 }
298 }
299 );
300
301 foreach my $zoom_query (@zoom_queries) {
302 $zoom_query->destroy();
303 }
304
305 return ( undef, $results, $total_hits );
306 }
307
308 =head2 getRecords
309
310 ( undef, $results_hashref, \@facets_loop ) = getRecords (
311
312 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
313 $results_per_page, $offset, $expanded_facet, $branches,$itemtypes,
314 $query_type, $scan
315 );
316
317 The all singing, all dancing, multi-server, asynchronous, scanning,
318 searching, record nabbing, facet-building
319
320 See verbse embedded documentation.
321
322 =cut
323
324 sub getRecords {
325 my (
326 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
327 $results_per_page, $offset, $expanded_facet, $branches,
328 $itemtypes, $query_type, $scan, $opac
329 ) = @_;
330
331 my @servers = @$servers_ref;
332 my @sort_by = @$sort_by_ref;
333
334 # Initialize variables for the ZOOM connection and results object
335 my $zconn;
336 my @zconns;
337 my @results;
338 my $results_hashref = ();
339
340 # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter
341 $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search };
342
343 # Initialize variables for the faceted results objects
344 my $facets_counter = {};
345 my $facets_info = {};
346 my $facets = getFacets();
347
348 my @facets_loop; # stores the ref to array of hashes for template facets loop
349
350 ### LOOP THROUGH THE SERVERS
351 for ( my $i = 0 ; $i < @servers ; $i++ ) {
352 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
353
354 # perform the search, create the results objects
355 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
356 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
357
358 #$query_to_use = $simple_query if $scan;
359 warn $simple_query if ( $scan and $DEBUG );
360
361 # Check if we've got a query_type defined, if so, use it
362 eval {
363 if ($query_type) {
364 if ($query_type =~ /^ccl/) {
365 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
366 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
367 } elsif ($query_type =~ /^cql/) {
368 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
369 } elsif ($query_type =~ /^pqf/) {
370 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
371 } else {
372 warn "Unknown query_type '$query_type'. Results undetermined.";
373 }
374 } elsif ($scan) {
375 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
376 } else {
377 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
378 }
379 };
380 if ($@) {
381 warn "WARNING: query problem with $query_to_use " . $@;
382 }
383
384 # Concatenate the sort_by limits and pass them to the results object
385 # Note: sort will override rank
386 my $sort_by;
387 foreach my $sort (@sort_by) {
388 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
389 $sort_by .= "1=1003 <i ";
390 }
391 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
392 $sort_by .= "1=1003 >i ";
393 }
394 elsif ( $sort eq "popularity_asc" ) {
395 $sort_by .= "1=9003 <i ";
396 }
397 elsif ( $sort eq "popularity_dsc" ) {
398 $sort_by .= "1=9003 >i ";
399 }
400 elsif ( $sort eq "call_number_asc" ) {
401 $sort_by .= "1=8007 <i ";
402 }
403 elsif ( $sort eq "call_number_dsc" ) {
404 $sort_by .= "1=8007 >i ";
405 }
406 elsif ( $sort eq "pubdate_asc" ) {
407 $sort_by .= "1=31 <i ";
408 }
409 elsif ( $sort eq "pubdate_dsc" ) {
410 $sort_by .= "1=31 >i ";
411 }
412 elsif ( $sort eq "acqdate_asc" ) {
413 $sort_by .= "1=32 <i ";
414 }
415 elsif ( $sort eq "acqdate_dsc" ) {
416 $sort_by .= "1=32 >i ";
417 }
418 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
419 $sort_by .= "1=4 <i ";
420 }
421 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
422 $sort_by .= "1=4 >i ";
423 }
424 else {
425 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
426 }
427 }
428 if ( $sort_by && !$scan && $results[$i] ) {
429 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
430 warn "WARNING sort $sort_by failed";
431 }
432 }
433 } # finished looping through servers
434
435 # The big moment: asynchronously retrieve results from all servers
436 _ZOOM_event_loop(
437 \@zconns,
438 \@results,
439 sub {
440 my ( $i, $size ) = @_;
441 my $results_hash;
442
443 # loop through the results
444 $results_hash->{'hits'} = $size;
445 my $times;
446 if ( $offset + $results_per_page <= $size ) {
447 $times = $offset + $results_per_page;
448 }
449 else {
450 $times = $size;
451 }
452
453 for ( my $j = $offset ; $j < $times ; $j++ ) {
454 my $records_hash;
455 my $record;
456
457 ## Check if it's an index scan
458 if ($scan) {
459 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
460
461 # here we create a minimal MARC record and hand it off to the
462 # template just like a normal result ... perhaps not ideal, but
463 # it works for now
464 my $tmprecord = MARC::Record->new();
465 $tmprecord->encoding('UTF-8');
466 my $tmptitle;
467 my $tmpauthor;
468
469 # the minimal record in author/title (depending on MARC flavour)
470 if ( C4::Context->preference("marcflavour") eq
471 "UNIMARC" )
472 {
473 $tmptitle = MARC::Field->new(
474 '200', ' ', ' ',
475 a => $term,
476 f => $occ
477 );
478 $tmprecord->append_fields($tmptitle);
479 }
480 else {
481 $tmptitle =
482 MARC::Field->new( '245', ' ', ' ', a => $term, );
483 $tmpauthor =
484 MARC::Field->new( '100', ' ', ' ', a => $occ, );
485 $tmprecord->append_fields($tmptitle);
486 $tmprecord->append_fields($tmpauthor);
487 }
488 $results_hash->{'RECORDS'}[$j] =
489 $tmprecord->as_usmarc();
490 }
491
492 # not an index scan
493 else {
494 $record = $results[ $i - 1 ]->record($j)->raw();
495 # warn "RECORD $j:".$record;
496 $results_hash->{'RECORDS'}[$j] = $record;
497 }
498
499 }
500 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
501
502 # Fill the facets while we're looping, but only for the
503 # biblioserver and not for a scan
504 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
505 $facets_counter = GetFacets( $results[ $i - 1 ] );
506 $facets_info = _get_facets_info( $facets );
507 }
508
509 # BUILD FACETS
510 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
511 for my $link_value (
512 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
513 keys %$facets_counter
514 )
515 {
516 my $expandable;
517 my $number_of_facets;
518 my @this_facets_array;
519 for my $one_facet (
520 sort {
521 $facets_counter->{$link_value}
522 ->{$b} <=> $facets_counter->{$link_value}
523 ->{$a}
524 } keys %{ $facets_counter->{$link_value} }
525 )
526 {
527 $number_of_facets++;
528 if ( ( $number_of_facets <= 5 )
529 || ( $expanded_facet eq $link_value )
530 || ( $facets_info->{$link_value}->{'expanded'} )
531 )
532 {
533
534 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
535 my $facet_link_value = $one_facet;
536 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
537
538 # fix the length that will display in the label,
539 my $facet_label_value = $one_facet;
540 my $facet_max_length = C4::Context->preference(
541 'FacetLabelTruncationLength')
542 || 20;
543 $facet_label_value =
544 substr( $one_facet, 0, $facet_max_length )
545 . "..."
546 if length($facet_label_value) >
547 $facet_max_length;
548
549 # if it's a branch, label by the name, not the code,
550 if ( $link_value =~ /branch/ ) {
551 if ( defined $branches
552 && ref($branches) eq "HASH"
553 && defined $branches->{$one_facet}
554 && ref( $branches->{$one_facet} ) eq
555 "HASH" )
556 {
557 $facet_label_value =
558 $branches->{$one_facet}
559 ->{'branchname'};
560 }
561 else {
562 $facet_label_value = "*";
563 }
564 }
565
566 # if it's a itemtype, label by the name, not the code,
567 if ( $link_value =~ /itype/ ) {
568 if ( defined $itemtypes
569 && ref($itemtypes) eq "HASH"
570 && defined $itemtypes->{$one_facet}
571 && ref( $itemtypes->{$one_facet} ) eq
572 "HASH" )
573 {
574 $facet_label_value =
575 $itemtypes->{$one_facet}
576 ->{translated_description};
577 }
578 }
579
580 # also, if it's a location code, use the name instead of the code
581 if ( $link_value =~ /location/ ) {
582 # TODO Retrieve all authorised values at once, instead of 1 query per entry
583 my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet });
584 $facet_label_value = $av->count ? $av->next->opac_description : '';
585 }
586
587 # but we're down with the whole label being in the link's title.
588 push @this_facets_array,
589 {
590 facet_count =>
591 $facets_counter->{$link_value}
592 ->{$one_facet},
593 facet_label_value => $facet_label_value,
594 facet_title_value => $one_facet,
595 facet_link_value => $facet_link_value,
596 type_link_value => $link_value,
597 }
598 if ($facet_label_value);
599 }
600 }
601
602 # handle expanded option
603 unless ( $facets_info->{$link_value}->{'expanded'} ) {
604 $expandable = 1
605 if ( ( $number_of_facets > 5 )
606 && ( $expanded_facet ne $link_value ) );
607 }
608 push @facets_loop,
609 {
610 type_link_value => $link_value,
611 type_id => $link_value . "_id",
612 "type_label_"
613 . $facets_info->{$link_value}->{'label_value'} =>
614 1,
615 facets => \@this_facets_array,
616 expandable => $expandable,
617 expand => $link_value,
618 }
619 unless (
620 (
621 $facets_info->{$link_value}->{'label_value'} =~
622 /Libraries/
623 )
624 and ( Koha::Libraries->search->count == 1 )
625 );
626 }
627 }
628 }
629 );
630
631 # This sorts the facets into alphabetical order
632 if (@facets_loop) {
633 foreach my $f (@facets_loop) {
634 $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ];
635 }
636 @facets_loop = sort {$a->{expand} cmp $b->{expand}} @facets_loop;
637 }
638
639 return ( undef, $results_hashref, \@facets_loop );
640 }
641
642 sub GetFacets {
643
644 my $rs = shift;
645 my $facets;
646
647 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
648 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
649
650 if ( $indexing_mode eq 'dom' &&
651 $use_zebra_facets ) {
652 $facets = _get_facets_from_zebra( $rs );
653 } else {
654 $facets = _get_facets_from_records( $rs );
655 }
656
657 return $facets;
658 }
659
660 sub _get_facets_from_records {
661
662 my $rs = shift;
663
664 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
665 my $facets_config = getFacets();
666 my $facets = {};
667 my $size = $rs->size();
668 my $jmax = $size > $facets_maxrecs
669 ? $facets_maxrecs
670 : $size;
671
672 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
673
674 my $marc_record = new_record_from_zebra (
675 'biblioserver',
676 $rs->record( $j )->raw()
677 );
678
679 if ( ! defined $marc_record ) {
680 warn "ERROR DECODING RECORD - $@: " .
681 $rs->record( $j )->raw();
682 next;
683 }
684
685 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
686 }
687
688 return $facets;
689 }
690
691 =head2 _get_facets_data_from_record
692
693 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
694
695 Internal function that extracts facets information from a MARC::Record object
696 and populates $facets_counter for using in getRecords.
697
698 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
699 facets for Zebra).
700
701 =cut
702
703 sub _get_facets_data_from_record {
704
705 my ( $marc_record, $facets, $facets_counter ) = @_;
706
707 for my $facet (@$facets) {
708
709 my @used_datas = ();
710
711 foreach my $tag ( @{ $facet->{ tags } } ) {
712
713 # tag number is the first three digits
714 my $tag_num = substr( $tag, 0, 3 );
715 # subfields are the remainder
716 my $subfield_letters = substr( $tag, 3 );
717
718 my @fields = $marc_record->field( $tag_num );
719 foreach my $field (@fields) {
720 # If $field->indicator(1) eq 'z', it means it is a 'see from'
721 # field introduced because of IncludeSeeFromInSearches, so skip it
722 next if $field->indicator(1) eq 'z';
723
724 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
725
726 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
727 push @used_datas, $data;
728 $facets_counter->{ $facet->{ idx } }->{ $data }++;
729 }
730 }
731 }
732 }
733 }
734
735 =head2 _get_facets_from_zebra
736
737 my $facets = _get_facets_from_zebra( $result_set )
738
739 Retrieves facets for a specified result set. It loops through the facets defined
740 in C4::Koha::getFacets and returns a hash with the following structure:
741
742 { facet_idx => {
743 facet_value => count
744 },
745 ...
746 }
747
748 =cut
749
750 sub _get_facets_from_zebra {
751
752 my $rs = shift;
753
754 # save current elementSetName
755 my $elementSetName = $rs->option( 'elementSetName' );
756
757 my $facets_loop = getFacets();
758 my $facets_data = {};
759 # loop through defined facets and fill the facets hashref
760 foreach my $facet ( @$facets_loop ) {
761
762 my $idx = $facet->{ idx };
763 my $sep = $facet->{ sep };
764 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
765 if ( $facet_values ) {
766 # we've actually got a result
767 $facets_data->{ $idx } = $facet_values;
768 }
769 }
770 # set elementSetName to its previous value to avoid side effects
771 $rs->option( elementSetName => $elementSetName );
772
773 return $facets_data;
774 }
775
776 =head2 _get_facet_from_result_set
777
778 my $facet_values =
779 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
780
781 Internal function that extracts facet information for a specific index ($facet_idx) and
782 returns a hash containing facet values and count:
783
784 {
785 $facet_value => $count ,
786 ...
787 }
788
789 Warning: this function has the side effect of changing the elementSetName for the result
790 set. It is a helper function for the main loop, which takes care of backing it up for
791 restoring.
792
793 =cut
794
795 sub _get_facet_from_result_set {
796
797 my $facet_idx = shift;
798 my $rs = shift;
799 my $sep = shift;
800
801 my $internal_sep = '<*>';
802 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
803
804 return if ( ! defined $facet_idx || ! defined $rs );
805 # zebra's facet element, untokenized index
806 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
807 # configure zebra results for retrieving the desired facet
808 $rs->option( elementSetName => $facet_element );
809 # get the facet record from result set
810 my $facet = $rs->record( 0 )->raw;
811 # if the facet has no restuls...
812 return if !defined $facet;
813 # TODO: benchmark DOM vs. SAX performance
814 my $facet_dom = XML::LibXML->load_xml(
815 string => ($facet)
816 );
817 my @terms = $facet_dom->getElementsByTagName('term');
818 return if ! @terms;
819
820 my $facets = {};
821 foreach my $term ( @terms ) {
822 my $facet_value = $term->textContent;
823 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
824 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
825 }
826
827 return $facets;
828 }
829
830 =head2 _get_facets_info
831
832 my $facets_info = C4::Search::_get_facets_info( $facets )
833
834 Internal function that extracts facets information and properly builds
835 the data structure needed to render facet labels.
836
837 =cut
838
839 sub _get_facets_info {
840
841 my $facets = shift;
842
843 my $facets_info = {};
844
845 for my $facet ( @$facets ) {
846 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
847 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
848 }
849
850 return $facets_info;
851 }
852
853 sub pazGetRecords {
854 my (
855 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
856 $results_per_page, $offset, $expanded_facet, $branches,
857 $query_type, $scan
858 ) = @_;
859 #NOTE: Parameter $branches is not used here !
860
861 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
862 $paz->init();
863 $paz->search($simple_query);
864 sleep 1; # FIXME: WHY?
865
866 # do results
867 my $results_hashref = {};
868 my $stats = XMLin($paz->stat);
869 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
870
871 # for a grouped search result, the number of hits
872 # is the number of groups returned; 'bib_hits' will have
873 # the total number of bibs.
874 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
875 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
876
877 HIT: foreach my $hit (@{ $results->{'hit'} }) {
878 my $recid = $hit->{recid}->[0];
879
880 my $work_title = $hit->{'md-work-title'}->[0];
881 my $work_author;
882 if (exists $hit->{'md-work-author'}) {
883 $work_author = $hit->{'md-work-author'}->[0];
884 }
885 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
886
887 my $result_group = {};
888 $result_group->{'group_label'} = $group_label;
889 $result_group->{'group_merge_key'} = $recid;
890
891 my $count = 1;
892 if (exists $hit->{count}) {
893 $count = $hit->{count}->[0];
894 }
895 $result_group->{'group_count'} = $count;
896
897 for (my $i = 0; $i < $count; $i++) {
898 # FIXME -- may need to worry about diacritics here
899 my $rec = $paz->record($recid, $i);
900 push @{ $result_group->{'RECORDS'} }, $rec;
901 }
902
903 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
904 }
905
906 # pass through facets
907 my $termlist_xml = $paz->termlist('author,subject');
908 my $terms = XMLin($termlist_xml, forcearray => 1);
909 my @facets_loop = ();
910 #die Dumper($results);
911 # foreach my $list (sort keys %{ $terms->{'list'} }) {
912 # my @facets = ();
913 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
914 # push @facets, {
915 # facet_label_value => $facet->{'name'}->[0],
916 # };
917 # }
918 # push @facets_loop, ( {
919 # type_label => $list,
920 # facets => \@facets,
921 # } );
922 # }
923
924 return ( undef, $results_hashref, \@facets_loop );
925 }
926
927 # TRUNCATION
928 sub _detect_truncation {
929 my ( $operand, $index ) = @_;
930 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
931 @regexpr );
932 $operand =~ s/^ //g;
933 my @wordlist = split( /\s/, $operand );
934 foreach my $word (@wordlist) {
935 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
936 push @rightlefttruncated, $word;
937 }
938 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
939 push @lefttruncated, $word;
940 }
941 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
942 push @righttruncated, $word;
943 }
944 elsif ( index( $word, "*" ) < 0 ) {
945 push @nontruncated, $word;
946 }
947 else {
948 push @regexpr, $word;
949 }
950 }
951 return (
952 \@nontruncated, \@righttruncated, \@lefttruncated,
953 \@rightlefttruncated, \@regexpr
954 );
955 }
956
957 # STEMMING
958 sub _build_stemmed_operand {
959 my ($operand,$lang) = @_;
960 require Lingua::Stem::Snowball ;
961 my $stemmed_operand=q{};
962
963 # If operand contains a digit, it is almost certainly an identifier, and should
964 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
965 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
966 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
967 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
968 return $operand if $operand =~ /\d/;
969
970 # FIXME: the locale should be set based on the user's language and/or search choice
971 #warn "$lang";
972 # Make sure we only use the first two letters from the language code
973 $lang = lc(substr($lang, 0, 2));
974 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
975 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
976 if ($lang eq 'nb' || $lang eq 'nn') {
977 $lang = 'no';
978 }
979 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
980 encoding => "UTF-8" );
981
982 my @words = split( / /, $operand );
983 my @stems = $stemmer->stem(\@words);
984 for my $stem (@stems) {
985 $stemmed_operand .= "$stem";
986 $stemmed_operand .= "?"
987 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
988 $stemmed_operand .= " ";
989 }
990 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
991 return $stemmed_operand;
992 }
993
994 # FIELD WEIGHTING
995 sub _build_weighted_query {
996
997 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
998 # pretty well but could work much better if we had a smarter query parser
999 my ( $operand, $stemmed_operand, $index ) = @_;
1000 my $stemming = C4::Context->preference("QueryStemming") || 0;
1001 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1002 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1003 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1004
1005 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1006
1007 # Keyword, or, no index specified
1008 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1009 $weighted_query .=
1010 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1011 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1012 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1013 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1014 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1015 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1016 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1017 if $fuzzy_enabled; # add fuzzy, word list
1018 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1019 if ( $stemming and $stemmed_operand )
1020 ; # add stemming, right truncation
1021 $weighted_query .= " or wrdl,r9=\"$operand\"";
1022
1023 # embedded sorting: 0 a-z; 1 z-a
1024 # $weighted_query .= ") or (sort1,aut=1";
1025 }
1026
1027 # Barcode searches should skip this process
1028 elsif ( $index eq 'bc' ) {
1029 $weighted_query .= "bc=\"$operand\"";
1030 }
1031
1032 # Authority-number searches should skip this process
1033 elsif ( $index eq 'an' ) {
1034 $weighted_query .= "an=\"$operand\"";
1035 }
1036
1037 # If the index is numeric, don't autoquote it.
1038 elsif ( $index =~ /,st-numeric$/ ) {
1039 $weighted_query .= " $index=$operand";
1040 }
1041
1042 # If the index already has more than one qualifier, wrap the operand
1043 # in quotes and pass it back (assumption is that the user knows what they
1044 # are doing and won't appreciate us mucking up their query
1045 elsif ( $index =~ ',' ) {
1046 $weighted_query .= " $index=\"$operand\"";
1047 }
1048
1049 #TODO: build better cases based on specific search indexes
1050 else {
1051 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1052 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1053 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1054 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1055 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1056 if $fuzzy_enabled; # add fuzzy, word list
1057 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1058 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1059 }
1060
1061 $weighted_query .= "))"; # close rank specification
1062 return $weighted_query;
1063 }
1064
1065 =head2 getIndexes
1066
1067 Return an array with available indexes.
1068
1069 =cut
1070
1071 sub getIndexes{
1072 my @indexes = (
1073 # biblio indexes
1074 'ab',
1075 'Abstract',
1076 'acqdate',
1077 'allrecords',
1078 'an',
1079 'Any',
1080 'at',
1081 'au',
1082 'aub',
1083 'aud',
1084 'audience',
1085 'auo',
1086 'aut',
1087 'Author',
1088 'Author-in-order ',
1089 'Author-personal-bibliography',
1090 'Authority-Number',
1091 'authtype',
1092 'bc',
1093 'Bib-level',
1094 'biblionumber',
1095 'bio',
1096 'biography',
1097 'callnum',
1098 'cfn',
1099 'Chronological-subdivision',
1100 'cn-bib-source',
1101 'cn-bib-sort',
1102 'cn-class',
1103 'cn-item',
1104 'cn-prefix',
1105 'cn-suffix',
1106 'cpn',
1107 'Code-institution',
1108 'Conference-name',
1109 'Conference-name-heading',
1110 'Conference-name-see',
1111 'Conference-name-seealso',
1112 'Content-type',
1113 'Control-number',
1114 'copydate',
1115 'Corporate-name',
1116 'Corporate-name-heading',
1117 'Corporate-name-see',
1118 'Corporate-name-seealso',
1119 'Country-publication',
1120 'ctype',
1121 'curriculum',
1122 'date-entered-on-file',
1123 'Date-of-acquisition',
1124 'Date-of-publication',
1125 'Date-time-last-modified',
1126 'Dewey-classification',
1127 'Dissertation-information',
1128 'diss',
1129 'dtlm',
1130 'EAN',
1131 'extent',
1132 'fic',
1133 'fiction',
1134 'Form-subdivision',
1135 'format',
1136 'Geographic-subdivision',
1137 'he',
1138 'Heading',
1139 'Heading-use-main-or-added-entry',
1140 'Heading-use-series-added-entry ',
1141 'Heading-use-subject-added-entry',
1142 'Host-item',
1143 'id-other',
1144 'ident',
1145 'Identifier-standard',
1146 'Illustration-code',
1147 'Index-term-genre',
1148 'Index-term-uncontrolled',
1149 'Interest-age-level',
1150 'Interest-grade-level',
1151 'ISBN',
1152 'isbn',
1153 'ISSN',
1154 'issn',
1155 'itemtype',
1156 'kw',
1157 'Koha-Auth-Number',
1158 'l-format',
1159 'language',
1160 'language-original',
1161 'lc-card',
1162 'LC-card-number',
1163 'lcn',
1164 'lex',
1165 'lexile-number',
1166 'llength',
1167 'ln',
1168 'ln-audio',
1169 'ln-subtitle',
1170 'Local-classification',
1171 'Local-number',
1172 'Match-heading',
1173 'Match-heading-see-from',
1174 'Material-type',
1175 'mc-itemtype',
1176 'mc-rtype',
1177 'mus',
1178 'name',
1179 'Music-number',
1180 'Name-geographic',
1181 'Name-geographic-heading',
1182 'Name-geographic-see',
1183 'Name-geographic-seealso',
1184 'nb',
1185 'Note',
1186 'notes',
1187 'ns',
1188 'nt',
1189 'Other-control-number',
1190 'pb',
1191 'Personal-name',
1192 'Personal-name-heading',
1193 'Personal-name-see',
1194 'Personal-name-seealso',
1195 'pl',
1196 'Place-publication',
1197 'pn',
1198 'popularity',
1199 'pubdate',
1200 'Publisher',
1201 'Provider',
1202 'pv',
1203 'Reading-grade-level',
1204 'Record-control-number',
1205 'rcn',
1206 'Record-type',
1207 'rtype',
1208 'se',
1209 'See',
1210 'See-also',
1211 'sn',
1212 'Stock-number',
1213 'su',
1214 'Subject',
1215 'Subject-heading-thesaurus',
1216 'Subject-name-personal',
1217 'Subject-subdivision',
1218 'Summary',
1219 'Suppress',
1220 'su-geo',
1221 'su-na',
1222 'su-to',
1223 'su-ut',
1224 'ut',
1225 'Term-genre-form',
1226 'Term-genre-form-heading',
1227 'Term-genre-form-see',
1228 'Term-genre-form-seealso',
1229 'ti',
1230 'Title',
1231 'Title-cover',
1232 'Title-series',
1233 'Title-uniform',
1234 'Title-uniform-heading',
1235 'Title-uniform-see',
1236 'Title-uniform-seealso',
1237 'totalissues',
1238 'yr',
1239
1240 # items indexes
1241 'acqsource',
1242 'barcode',
1243 'bc',
1244 'branch',
1245 'ccode',
1246 'classification-source',
1247 'cn-sort',
1248 'coded-location-qualifier',
1249 'copynumber',
1250 'damaged',
1251 'datelastborrowed',
1252 'datelastseen',
1253 'holdingbranch',
1254 'homebranch',
1255 'issues',
1256 'item',
1257 'itemnumber',
1258 'itype',
1259 'Local-classification',
1260 'location',
1261 'lost',
1262 'materials-specified',
1263 'mc-ccode',
1264 'mc-itype',
1265 'mc-loc',
1266 'notforloan',
1267 'Number-local-acquisition',
1268 'onloan',
1269 'price',
1270 'renewals',
1271 'replacementprice',
1272 'replacementpricedate',
1273 'reserves',
1274 'restricted',
1275 'stack',
1276 'stocknumber',
1277 'inv',
1278 'uri',
1279 'withdrawn',
1280
1281 # subject related
1282 );
1283
1284 return \@indexes;
1285 }
1286
1287 =head2 _handle_exploding_index
1288
1289 my $query = _handle_exploding_index($index, $term)
1290
1291 Callback routine to generate the search for "exploding" indexes (i.e.
1292 those indexes which are turned into multiple or-connected searches based
1293 on authority data).
1294
1295 =cut
1296
1297 sub _handle_exploding_index {
1298 my ($QParser, $filter, $params, $negate, $server) = @_;
1299 my $index = $filter;
1300 my $term = join(' ', @$params);
1301
1302 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1303
1304 my $marcflavour = C4::Context->preference('marcflavour');
1305
1306 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1307 my $wantedcodes = '';
1308 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1309 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1310 foreach my $auth (@$results) {
1311 my $record = MARC::Record->new_from_usmarc($auth);
1312 my @references = $record->field('5..');
1313 if (@references) {
1314 if ($index eq 'su-br') {
1315 $wantedcodes = 'g';
1316 } elsif ($index eq 'su-na') {
1317 $wantedcodes = 'h';
1318 } elsif ($index eq 'su-rl') {
1319 $wantedcodes = '';
1320 }
1321 foreach my $reference (@references) {
1322 my $codes = $reference->subfield($codesubfield);
1323 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1324 }
1325 }
1326 }
1327 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1328 return $query;
1329 }
1330
1331 =head2 parseQuery
1332
1333 ( $operators, $operands, $indexes, $limits,
1334 $sort_by, $scan, $lang ) =
1335 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1336
1337 Shim function to ease the transition from buildQuery to a new QueryParser.
1338 This function is called at the beginning of buildQuery, and modifies
1339 buildQuery's input. If it can handle the input, it returns a query that
1340 buildQuery will not try to parse.
1341
1342 =cut
1343
1344 sub parseQuery {
1345 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1346
1347 my @operators = $operators ? @$operators : ();
1348 my @indexes = $indexes ? @$indexes : ();
1349 my @operands = $operands ? @$operands : ();
1350 my @limits = $limits ? @$limits : ();
1351 my @sort_by = $sort_by ? @$sort_by : ();
1352
1353 my $query = $operands[0];
1354 my $index;
1355 my $term;
1356 my $query_desc;
1357
1358 my $QParser;
1359 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1360 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1361 undef $QParser if (scalar @limits > 0);
1362
1363 if ($QParser)
1364 {
1365 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1366 $query = '';
1367 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1368 next unless $operands[$ii];
1369 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1370 if ($query);
1371 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1372 $query .= $operands[$ii];
1373 }
1374 elsif ( $indexes[$ii] =~ m/su-/ ) {
1375 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1376 }
1377 else {
1378 $query .=
1379 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1380 }
1381 }
1382 foreach my $limit (@limits) {
1383 }
1384 if ( scalar(@sort_by) > 0 ) {
1385 my $modifier_re =
1386 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1387 $query =~ s/$modifier_re//g;
1388 foreach my $modifier (@sort_by) {
1389 $query .= " #$modifier";
1390 }
1391 }
1392
1393 $query_desc = $query;
1394 $query_desc =~ s/\s+/ /g;
1395 if ( C4::Context->preference("QueryWeightFields") ) {
1396 }
1397 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1398 { 'target_syntax_callback' => \&_handle_exploding_index } );
1399 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1400 { 'target_syntax_callback' => \&_handle_exploding_index } );
1401 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1402 { 'target_syntax_callback' => \&_handle_exploding_index } );
1403 $QParser->parse($query);
1404 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1405 }
1406 else {
1407 require Koha::QueryParser::Driver::PQF;
1408 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1409 s/$modifier_re//g for @operands;
1410 }
1411
1412 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1413 }
1414
1415 =head2 buildQuery
1416
1417 ( $error, $query,
1418 $simple_query, $query_cgi,
1419 $query_desc, $limit,
1420 $limit_cgi, $limit_desc,
1421 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1422
1423 Build queries and limits in CCL, CGI, Human,
1424 handle truncation, stemming, field weighting, fuzziness, etc.
1425
1426 See verbose embedded documentation.
1427
1428
1429 =cut
1430
1431 sub buildQuery {
1432 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1433
1434 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1435
1436 my $query_desc;
1437 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1438
1439 # dereference
1440 my @operators = $operators ? @$operators : ();
1441 my @indexes = $indexes ? @$indexes : ();
1442 my @operands = $operands ? @$operands : ();
1443 my @limits = $limits ? @$limits : ();
1444 my @sort_by = $sort_by ? @$sort_by : ();
1445
1446 my $stemming = C4::Context->preference("QueryStemming") || 0;
1447 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1448 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1449 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1450
1451 my $query = $operands[0];
1452 my $simple_query = $operands[0];
1453
1454 # initialize the variables we're passing back
1455 my $query_cgi;
1456 my $query_type;
1457
1458 my $limit;
1459 my $limit_cgi;
1460 my $limit_desc;
1461
1462 my $cclq = 0;
1463 my $cclindexes = getIndexes();
1464 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1465 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1466 my $dx = lc($1);
1467 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1468 }
1469 $query = "ccl=$query" if $cclq;
1470 }
1471
1472 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1473 # DIAGNOSTIC ONLY!!
1474 if ( $query =~ /^ccl=/ ) {
1475 my $q=$';
1476 # This is needed otherwise ccl= and &limit won't work together, and
1477 # this happens when selecting a subject on the opac-detail page
1478 @limits = grep {!/^$/} @limits;
1479 my $original_q = $q; # without available part
1480 unless ( grep { /^available$/ } @limits ) {
1481 $q =~ s| and \( \( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:''\) and \(lost,st-numeric=0\) \)||;
1482 $original_q = $q;
1483 }
1484 if ( @limits ) {
1485 if ( grep { /^available$/ } @limits ) {
1486 $q .= q| and ( ( allrecords,AlwaysMatches:'' not onloan,AlwaysMatches:'') and (lost,st-numeric=0) )|;
1487 delete $limits['available'];
1488 }
1489 $q .= ' and '.join(' and ', @limits) if @limits;
1490 }
1491 return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' );
1492 }
1493 if ( $query =~ /^cql=/ ) {
1494 return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1495 }
1496 if ( $query =~ /^pqf=/ ) {
1497 if ($query_desc) {
1498 $query_cgi = "q=".uri_escape_utf8($query_desc);
1499 } else {
1500 $query_desc = $';
1501 $query_cgi = "q=pqf=".uri_escape_utf8($');
1502 }
1503 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1504 }
1505
1506 # pass nested queries directly
1507 # FIXME: need better handling of some of these variables in this case
1508 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1509 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1510 # if ( $query =~ /(\(|\))/ ) {
1511 # return (
1512 # undef, $query, $simple_query, $query_cgi,
1513 # $query, $limit, $limit_cgi, $limit_desc,
1514 # 'ccl'
1515 # );
1516 # }
1517
1518 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1519 # query operands and indexes and add stemming, truncation, field weighting, etc.
1520 # Once we do so, we'll end up with a value in $query, just like if we had an
1521 # incoming $query from the user
1522 else {
1523 $query = ""
1524 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1525 my $previous_operand
1526 ; # a flag used to keep track if there was a previous query
1527 # if there was, we can apply the current operator
1528 # for every operand
1529 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1530
1531 # COMBINE OPERANDS, INDEXES AND OPERATORS
1532 if ( $operands[$i] ) {
1533 $operands[$i]=~s/^\s+//;
1534
1535 # A flag to determine whether or not to add the index to the query
1536 my $indexes_set;
1537
1538 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1539 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1540 $weight_fields = 0;
1541 $stemming = 0;
1542 } else {
1543 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1544 }
1545 my $operand = $operands[$i];
1546 my $index = $indexes[$i];
1547
1548 # Add index-specific attributes
1549
1550 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1551 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1552 #which is processed higher up in this sub. Other than that, year searches are typically
1553 #handled as limits which are not processed her either.
1554
1555 # Search ranges: Date of Publication, st-numeric
1556 if ( $index =~ /(yr|st-numeric)/ ) {
1557 #weight_fields/relevance search causes errors with date ranges
1558 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1559 #In the case of YYYY-YYYY, it will return no results
1560 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1561 }
1562
1563 # Date of Acquisition
1564 elsif ( $index =~ /acqdate/ ) {
1565 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1566 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1567 #top of the results just because they have lots of item records matching that date.
1568 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1569 #irrelevant here
1570 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1571 }
1572 # ISBN,ISSN,Standard Number, don't need special treatment
1573 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1574 (
1575 $stemming, $auto_truncation,
1576 $weight_fields, $fuzzy_enabled
1577 ) = ( 0, 0, 0, 0 );
1578
1579 if ( $index eq 'nb' ) {
1580 if ( C4::Context->preference("SearchWithISBNVariations") ) {
1581 my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1582 $operands[$i] = $operand = '(nb=' . join(' OR nb=', @isbns) . ')';
1583 $indexes[$i] = $index = '';
1584 }
1585 }
1586 }
1587
1588 if(not $index){
1589 $index = 'kw';
1590 }
1591
1592 # Set default structure attribute (word list)
1593 my $struct_attr = q{};
1594 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1595 $struct_attr = ",wrdl";
1596 }
1597
1598 # Some helpful index variants
1599 my $index_plus = $index . $struct_attr . ':';
1600 my $index_plus_comma = $index . $struct_attr . ',';
1601
1602 if ($auto_truncation){
1603 unless ( $index =~ /,(st-|phr|ext)/ ) {
1604 #FIXME only valid with LTR scripts
1605 $operand=join(" ",map{
1606 (index($_,"*")>0?"$_":"$_*")
1607 }split (/\s+/,$operand));
1608 warn $operand if $DEBUG;
1609 }
1610 }
1611
1612 # Detect Truncation
1613 my $truncated_operand;
1614 my( $nontruncated, $righttruncated, $lefttruncated,
1615 $rightlefttruncated, $regexpr
1616 ) = _detect_truncation( $operand, $index );
1617 warn
1618 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1619 if $DEBUG;
1620
1621 # Apply Truncation
1622 if (
1623 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1624 scalar(@$rightlefttruncated) > 0 )
1625 {
1626
1627 # Don't field weight or add the index to the query, we do it here
1628 $indexes_set = 1;
1629 undef $weight_fields;
1630 my $previous_truncation_operand;
1631 if (scalar @$nontruncated) {
1632 $truncated_operand .= "$index_plus @$nontruncated ";
1633 $previous_truncation_operand = 1;
1634 }
1635 if (scalar @$righttruncated) {
1636 $truncated_operand .= "and " if $previous_truncation_operand;
1637 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1638 $previous_truncation_operand = 1;
1639 }
1640 if (scalar @$lefttruncated) {
1641 $truncated_operand .= "and " if $previous_truncation_operand;
1642 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1643 $previous_truncation_operand = 1;
1644 }
1645 if (scalar @$rightlefttruncated) {
1646 $truncated_operand .= "and " if $previous_truncation_operand;
1647 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1648 $previous_truncation_operand = 1;
1649 }
1650 }
1651 $operand = $truncated_operand if $truncated_operand;
1652 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1653
1654 # Handle Stemming
1655 my $stemmed_operand;
1656 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1657 if $stemming;
1658
1659 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1660
1661 # Handle Field Weighting
1662 my $weighted_operand;
1663 if ($weight_fields) {
1664 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1665 $operand = $weighted_operand;
1666 $indexes_set = 1;
1667 }
1668
1669 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1670
1671 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1672 query => $query,
1673 query_cgi => $query_cgi,
1674 query_desc => $query_desc,
1675 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1676 parsed_operand => $operand,
1677 original_operand => ($operands[$i]) ? $operands[$i] : '',
1678 index => $index,
1679 index_plus => $index_plus,
1680 indexes_set => $indexes_set,
1681 previous_operand => $previous_operand,
1682 });
1683
1684 } #/if $operands
1685 } # /for
1686 }
1687 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1688
1689 # add limits
1690 my %group_OR_limits;
1691 my $availability_limit;
1692 foreach my $this_limit (@limits) {
1693 next unless $this_limit;
1694 if ( $this_limit =~ /available/ ) {
1695 #
1696 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1697 ## In English:
1698 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1699 $availability_limit .=
1700 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1701 $limit_cgi .= "&limit=available";
1702 $limit_desc .= "";
1703 }
1704
1705 # group_OR_limits, prefixed by mc-
1706 # OR every member of the group
1707 elsif ( $this_limit =~ /mc/ ) {
1708 my ($k,$v) = split(/:/, $this_limit,2);
1709 if ( $k !~ /mc-i(tem)?type/ ) {
1710 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1711 $this_limit =~ tr/"//d;
1712 $this_limit = $k.":'".$v."'";
1713 }
1714
1715 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1716 $limit_desc .= " or " if $group_OR_limits{$k};
1717 $group_OR_limits{$k} .= "$this_limit";
1718 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1719 $limit_desc .= " $this_limit";
1720 }
1721
1722 # Regular old limits
1723 else {
1724 $limit .= " and " if $limit || $query;
1725 $limit .= "$this_limit";
1726 $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit);
1727 if ($this_limit =~ /^branch:(.+)/) {
1728 my $branchcode = $1;
1729 my $library = Koha::Libraries->find( $branchcode );
1730 if (defined $library) {
1731 $limit_desc .= " branch:" . $library->branchname;
1732 } else {
1733 $limit_desc .= " $this_limit";
1734 }
1735 } else {
1736 $limit_desc .= " $this_limit";
1737 }
1738 }
1739 }
1740 foreach my $k (keys (%group_OR_limits)) {
1741 $limit .= " and " if ( $query || $limit );
1742 $limit .= "($group_OR_limits{$k})";
1743 }
1744 if ($availability_limit) {
1745 $limit .= " and " if ( $query || $limit );
1746 $limit .= "($availability_limit)";
1747 }
1748
1749 # Normalize the query and limit strings
1750 # This is flawed , means we can't search anything with : in it
1751 # if user wants to do ccl or cql, start the query with that
1752 # $query =~ s/:/=/g;
1753 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1754 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1755 $query =~ s/(?<=(wrdl)):/=/g;
1756 $query =~ s/(?<=(trn|phr)):/=/g;
1757 $query =~ s/(?<=(st-numeric)):/=/g;
1758 $query =~ s/(?<=(st-year)):/=/g;
1759 $query =~ s/(?<=(st-date-normalized)):/=/g;
1760 $limit =~ s/:/=/g;
1761 for ( $query, $query_desc, $limit, $limit_desc ) {
1762 s/ +/ /g; # remove extra spaces
1763 s/^ //g; # remove any beginning spaces
1764 s/ $//g; # remove any ending spaces
1765 s/==/=/g; # remove double == from query
1766 }
1767 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1768
1769 for ($query_cgi,$simple_query) {
1770 s/"//g;
1771 }
1772 # append the limit to the query
1773 $query .= " " . $limit;
1774
1775 # Warnings if DEBUG
1776 if ($DEBUG) {
1777 warn "QUERY:" . $query;
1778 warn "QUERY CGI:" . $query_cgi;
1779 warn "QUERY DESC:" . $query_desc;
1780 warn "LIMIT:" . $limit;
1781 warn "LIMIT CGI:" . $limit_cgi;
1782 warn "LIMIT DESC:" . $limit_desc;
1783 warn "---------\nLeave buildQuery\n---------";
1784 }
1785
1786 return (
1787 undef, $query, $simple_query, $query_cgi,
1788 $query_desc, $limit, $limit_cgi, $limit_desc,
1789 $query_type
1790 );
1791 }
1792
1793 =head2 _build_initial_query
1794
1795 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1796
1797 Build a section of the initial query containing indexes, operators, and operands.
1798
1799 =cut
1800
1801 sub _build_initial_query {
1802 my ($params) = @_;
1803
1804 my $operator = "";
1805 if ($params->{previous_operand}){
1806 #If there is a previous operand, add a supplied operator or the default 'and'
1807 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1808 }
1809
1810 #NOTE: indexes_set is typically set when doing truncation or field weighting
1811 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1812
1813 #e.g. "kw,wrdl:test"
1814 #e.g. " and kw,wrdl:test"
1815 $params->{query} .= $operator . $operand;
1816
1817 $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1818 $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1819 $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1820
1821 #e.g. " and kw,wrdl: test"
1822 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1823
1824 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1825
1826 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1827 }
1828
1829 =head2 searchResults
1830
1831 my @search_results = searchResults($search_context, $searchdesc, $hits,
1832 $results_per_page, $offset, $scan,
1833 @marcresults);
1834
1835 Format results in a form suitable for passing to the template
1836
1837 =cut
1838
1839 # IMO this subroutine is pretty messy still -- it's responsible for
1840 # building the HTML output for the template
1841 sub searchResults {
1842 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1843 my $dbh = C4::Context->dbh;
1844 my @newresults;
1845
1846 require C4::Items;
1847
1848 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1849 my ($is_opac, $hidelostitems);
1850 if ($search_context eq 'opac') {
1851 $hidelostitems = C4::Context->preference('hidelostitems');
1852 $is_opac = 1;
1853 }
1854
1855 #Build branchnames hash
1856 my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' });
1857
1858 # FIXME - We build an authorised values hash here, using the default framework
1859 # though it is possible to have different authvals for different fws.
1860
1861 my $shelflocations =
1862 { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) };
1863
1864 # get notforloan authorised value list (see $shelflocations FIXME)
1865 my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => { not => undef } });
1866 my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef;
1867
1868 #Get itemtype hash
1869 my %itemtypes = %{ GetItemTypes() };
1870
1871 #search item field code
1872 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1873
1874 ## find column names of items related to MARC
1875 my %subfieldstosearch;
1876 my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1877 for my $column ( @columns ) {
1878 my ( $tagfield, $tagsubfield ) =
1879 &GetMarcFromKohaField( "items." . $column, "" );
1880 if ( defined $tagsubfield ) {
1881 $subfieldstosearch{$column} = $tagsubfield;
1882 }
1883 }
1884
1885 # handle which records to actually retrieve
1886 my $times;
1887 if ( $hits && $offset + $results_per_page <= $hits ) {
1888 $times = $offset + $results_per_page;
1889 }
1890 else {
1891 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1892 }
1893
1894 my $marcflavour = C4::Context->preference("marcflavour");
1895 # We get the biblionumber position in MARC
1896 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1897
1898 # set stuff for XSLT processing here once, not later again for every record we retrieved
1899 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1900 my $xslsyspref = $interface . "XSLTResultsDisplay";
1901 my $xslfile = C4::Context->preference($xslsyspref);
1902 my $lang = $xslfile ? C4::Languages::getlanguage() : undef;
1903 my $sysxml = $xslfile ? C4::XSLT::get_xslt_sysprefs() : undef;
1904
1905 # loop through all of the records we've retrieved
1906 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1907
1908 my $marcrecord;
1909 if ($scan) {
1910 # For Scan searches we built USMARC data
1911 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1912 } else {
1913 # Normal search, render from Zebra's output
1914 $marcrecord = new_record_from_zebra(
1915 'biblioserver',
1916 $marcresults->[$i]
1917 );
1918
1919 if ( ! defined $marcrecord ) {
1920 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1921 next;
1922 }
1923 }
1924
1925 my $fw = $scan
1926 ? undef
1927 : $bibliotag < 10
1928 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1929 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1930
1931 SetUTF8Flag($marcrecord);
1932 my $oldbiblio = TransformMarcToKoha( $marcrecord, $fw );
1933 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1934 $oldbiblio->{result_number} = $i + 1;
1935
1936 # add imageurl to itemtype if there is one
1937 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1938
1939 $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour);
1940 $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour);
1941 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1942 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1943 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1944
1945 # edition information, if any
1946 $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1947 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1948 # Build summary if there is one (the summary is defined in the itemtypes table)
1949 # FIXME: is this used anywhere, I think it can be commented out? -- JF
1950 if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1951 my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1952 my @fields = $marcrecord->fields();
1953
1954 my $newsummary;
1955 foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1956 my $tags = {};
1957 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1958 $tag =~ /(.{3})(.)/;
1959 if($marcrecord->field($1)){
1960 my @abc = $marcrecord->field($1)->subfield($2);
1961 $tags->{$tag} = $#abc + 1 ;
1962 }
1963 }
1964
1965 # We catch how many times to repeat this line
1966 my $max = 0;
1967 foreach my $tag (keys(%$tags)){
1968 $max = $tags->{$tag} if($tags->{$tag} > $max);
1969 }
1970
1971 # we replace, and repeat each line
1972 for (my $i = 0 ; $i < $max ; $i++){
1973 my $newline = $line;
1974
1975 foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1976 $tag =~ /(.{3})(.)/;
1977
1978 if($marcrecord->field($1)){
1979 my @repl = $marcrecord->field($1)->subfield($2);
1980 my $subfieldvalue = $repl[$i];
1981 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1982 }
1983 }
1984 $newsummary .= "$newline\n";
1985 }
1986 }
1987
1988 $newsummary =~ s/\[(.*?)]//g;
1989 $newsummary =~ s/\n/<br\/>/g;
1990 $oldbiblio->{summary} = $newsummary;
1991 }
1992
1993 # Pull out the items fields
1994 my @fields = $marcrecord->field($itemtag);
1995 my $marcflavor = C4::Context->preference("marcflavour");
1996 # adding linked items that belong to host records
1997 my $analyticsfield = '773';
1998 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1999 $analyticsfield = '773';
2000 } elsif ($marcflavor eq 'UNIMARC') {
2001 $analyticsfield = '461';
2002 }
2003 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2004 my $hostbiblionumber = $hostfield->subfield("0");
2005 my $linkeditemnumber = $hostfield->subfield("9");
2006 if(!$hostbiblionumber eq undef){
2007 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2008 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2009 if(!$hostbiblio eq undef){
2010 my @hostitems = $hostbiblio->field($itemfield);
2011 foreach my $hostitem (@hostitems){
2012 if ($hostitem->subfield("9") eq $linkeditemnumber){
2013 my $linkeditem =$hostitem;
2014 # append linked items if they exist
2015 if (!$linkeditem eq undef){
2016 push (@fields, $linkeditem);}
2017 }
2018 }
2019 }
2020 }
2021 }
2022
2023 # Setting item statuses for display
2024 my @available_items_loop;
2025 my @onloan_items_loop;
2026 my @other_items_loop;
2027
2028 my $available_items;
2029 my $onloan_items;
2030 my $other_items;
2031
2032 my $ordered_count = 0;
2033 my $available_count = 0;
2034 my $onloan_count = 0;
2035 my $longoverdue_count = 0;
2036 my $other_count = 0;
2037 my $withdrawn_count = 0;
2038 my $itemlost_count = 0;
2039 my $hideatopac_count = 0;
2040 my $itembinding_count = 0;
2041 my $itemdamaged_count = 0;
2042 my $item_in_transit_count = 0;
2043 my $can_place_holds = 0;
2044 my $item_onhold_count = 0;
2045 my $notforloan_count = 0;
2046 my $items_count = scalar(@fields);
2047 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2048 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2049 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2050
2051 # loop through every item
2052 foreach my $field (@fields) {
2053 my $item;
2054
2055 # populate the items hash
2056 foreach my $code ( keys %subfieldstosearch ) {
2057 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2058 }
2059 $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2060
2061 # OPAC hidden items
2062 if ($is_opac) {
2063 # hidden because lost
2064 if ($hidelostitems && $item->{itemlost}) {
2065 $hideatopac_count++;
2066 next;
2067 }
2068 # hidden based on OpacHiddenItems syspref
2069 my @hi = C4::Items::GetHiddenItemnumbers($item);
2070 if (scalar @hi) {
2071 push @hiddenitems, @hi;
2072 $hideatopac_count++;
2073 next;
2074 }
2075 }
2076
2077 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2078 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2079
2080 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2081 if ($item->{$hbranch}) {
2082 $item->{'branchname'} = $branches{$item->{$hbranch}};
2083 }
2084 elsif ($item->{$otherbranch}) { # Last resort
2085 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2086 }
2087
2088 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2089 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2090 my $userenv = C4::Context->userenv;
2091 if ( $item->{onloan}
2092 && $userenv
2093 && $userenv->{number}
2094 && !( Koha::Patrons->find($userenv->{number})->category->hidelostitems && $item->{itemlost} ) )
2095 {
2096 $onloan_count++;
2097 my $key = $prefix . $item->{onloan} . $item->{barcode};
2098 $onloan_items->{$key}->{due_date} = output_pref( { str => $item->{onloan}, dateonly => 1 } );
2099 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2100 $onloan_items->{$key}->{branchname} = $item->{branchname};
2101 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2102 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2103 $onloan_items->{$key}->{description} = $item->{description};
2104 $onloan_items->{$key}->{imageurl} =
2105 getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2106
2107 # if something's checked out and lost, mark it as 'long overdue'
2108 if ( $item->{itemlost} ) {
2109 $onloan_items->{$key}->{longoverdue}++;
2110 $longoverdue_count++;
2111 }
2112 else { # can place holds as long as item isn't lost
2113 $can_place_holds = 1;
2114 }
2115 }
2116
2117 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2118 else {
2119
2120 $item->{notforloan}=1 if !$item->{notforloan} && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2121
2122 # item is on order
2123 if ( $item->{notforloan} < 0 ) {
2124 $ordered_count++;
2125 } elsif ( $item->{notforloan} > 0 ) {
2126 $notforloan_count++;
2127 }
2128
2129 # is item in transit?
2130 my $transfertwhen = '';
2131 my ($transfertfrom, $transfertto);
2132
2133 # is item on the reserve shelf?
2134 my $reservestatus = '';
2135
2136 unless ($item->{withdrawn}
2137 || $item->{itemlost}
2138 || $item->{damaged}
2139 || $item->{notforloan}
2140 || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2141 && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2142
2143 # A couple heuristics to limit how many times
2144 # we query the database for item transfer information, sacrificing
2145 # accuracy in some cases for speed;
2146 #
2147 # 1. don't query if item has one of the other statuses
2148 # 2. don't check transit status if the bib has
2149 # more than 20 items
2150 #
2151 # FIXME: to avoid having the query the database like this, and to make
2152 # the in transit status count as unavailable for search limiting,
2153 # should map transit status to record indexed in Zebra.
2154 #
2155 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2156 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2157 }
2158
2159 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2160 if ( $item->{withdrawn}
2161 || $item->{itemlost}
2162 || $item->{damaged}
2163 || $item->{notforloan}
2164 || $reservestatus eq 'Waiting'
2165 || ($transfertwhen ne ''))
2166 {
2167 $withdrawn_count++ if $item->{withdrawn};
2168 $itemlost_count++ if $item->{itemlost};
2169 $itemdamaged_count++ if $item->{damaged};
2170 $item_in_transit_count++ if $transfertwhen ne '';
2171 $item_onhold_count++ if $reservestatus eq 'Waiting';
2172 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2173
2174 # can place a hold on a item if
2175 # not lost nor withdrawn
2176 # not damaged unless AllowHoldsOnDamagedItems is true
2177 # item is either for loan or on order (notforloan < 0)
2178 $can_place_holds = 1
2179 if (
2180 !$item->{itemlost}
2181 && !$item->{withdrawn}
2182 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2183 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2184 );
2185
2186 $other_count++;
2187
2188 my $key = $prefix . $item->{status};
2189 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2190 $other_items->{$key}->{$_} = $item->{$_};
2191 }
2192 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2193 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2194 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2195 $other_items->{$key}->{count}++ if $item->{$hbranch};
2196 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2197 $other_items->{$key}->{description} = $item->{description};
2198 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2199 }
2200 # item is available
2201 else {
2202 $can_place_holds = 1;
2203 $available_count++;
2204 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2205 foreach (qw(branchname itemcallnumber description)) {
2206 $available_items->{$prefix}->{$_} = $item->{$_};
2207 }
2208 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2209 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2210 }
2211 }
2212 } # notforloan, item level and biblioitem level
2213
2214 # if all items are hidden, do not show the record
2215 if ($items_count > 0 && $hideatopac_count == $items_count) {
2216 next;
2217 }
2218
2219 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2220 for my $key ( sort keys %$onloan_items ) {
2221 (++$onloanitemscount > $maxitems) and last;
2222 push @onloan_items_loop, $onloan_items->{$key};
2223 }
2224 for my $key ( sort keys %$other_items ) {
2225 (++$otheritemscount > $maxitems) and last;
2226 push @other_items_loop, $other_items->{$key};
2227 }
2228 for my $key ( sort keys %$available_items ) {
2229 (++$availableitemscount > $maxitems) and last;
2230 push @available_items_loop, $available_items->{$key}
2231 }
2232
2233 # XSLT processing of some stuff
2234 # we fetched the sysprefs already before the loop through all retrieved record!
2235 if (!$scan && $xslfile) {
2236 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $xslsyspref, 1, \@hiddenitems, $sysxml, $xslfile, $lang);
2237 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2238 }
2239
2240 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2241 if (!C4::Context->preference("item-level_itypes")) {
2242 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2243 $can_place_holds = 0;
2244 }
2245 }
2246 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2247 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2248 $oldbiblio->{items_count} = $items_count;
2249 $oldbiblio->{available_items_loop} = \@available_items_loop;
2250 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2251 $oldbiblio->{other_items_loop} = \@other_items_loop;
2252 $oldbiblio->{availablecount} = $available_count;
2253 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2254 $oldbiblio->{onloancount} = $onloan_count;
2255 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2256 $oldbiblio->{othercount} = $other_count;
2257 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2258 $oldbiblio->{withdrawncount} = $withdrawn_count;
2259 $oldbiblio->{itemlostcount} = $itemlost_count;
2260 $oldbiblio->{damagedcount} = $itemdamaged_count;
2261 $oldbiblio->{intransitcount} = $item_in_transit_count;
2262 $oldbiblio->{onholdcount} = $item_onhold_count;
2263 $oldbiblio->{orderedcount} = $ordered_count;
2264 $oldbiblio->{notforloancount} = $notforloan_count;
2265
2266 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2267 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2268 my $subfields = substr $fieldspec, 3;
2269 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2270 my @alternateholdingsinfo = ();
2271 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2272 my $alternateholdingscount = 0;
2273
2274 for my $field (@holdingsfields) {
2275 my %holding = ( holding => '' );
2276 my $havesubfield = 0;
2277 for my $subfield ($field->subfields()) {
2278 if ((index $subfields, $$subfield[0]) >= 0) {
2279 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2280 $holding{'holding'} .= $$subfield[1];
2281 $havesubfield++;
2282 }
2283 }
2284 if ($havesubfield) {
2285 push(@alternateholdingsinfo, \%holding);
2286 $alternateholdingscount++;
2287 }
2288 }
2289
2290 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2291 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2292 }
2293
2294 push( @newresults, $oldbiblio );
2295 }
2296
2297 return @newresults;
2298 }
2299
2300 =head2 enabled_staff_search_views
2301
2302 %hash = enabled_staff_search_views()
2303
2304 This function returns a hash that contains three flags obtained from the system
2305 preferences, used to determine whether a particular staff search results view
2306 is enabled.
2307
2308 =over 2
2309
2310 =item C<Output arg:>
2311
2312 * $hash{can_view_MARC} is true only if the MARC view is enabled
2313 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2314 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2315
2316 =item C<usage in the script:>
2317
2318 =back
2319
2320 $template->param ( C4::Search::enabled_staff_search_views );
2321
2322 =cut
2323
2324 sub enabled_staff_search_views
2325 {
2326 return (
2327 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2328 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2329 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2330 );
2331 }
2332
2333 =head2 z3950_search_args
2334
2335 $arrayref = z3950_search_args($matchpoints)
2336
2337 This function returns an array reference that contains the search parameters to be
2338 passed to the Z39.50 search script (z3950_search.pl). The array elements
2339 are hash refs whose keys are name and value, and whose values are the
2340 name of a search parameter, the value of that search parameter and the URL encoded
2341 value of that parameter.
2342
2343 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2344
2345 The search parameter values are obtained from the bibliographic record whose
2346 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2347
2348 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2349 a general purpose search argument. In this case, the returned array contains only
2350 entry: the key is 'title' and the value is derived from $matchpoints.
2351
2352 If a search parameter value is undefined or empty, it is not included in the returned
2353 array.
2354
2355 The returned array reference may be passed directly to the template parameters.
2356
2357 =over 2
2358
2359 =item C<Output arg:>
2360
2361 * $array containing hash refs as described above
2362
2363 =item C<usage in the script:>
2364
2365 =back
2366
2367 $data = Biblio::GetBiblioData($bibno);
2368 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2369
2370 *OR*
2371
2372 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2373
2374 =cut
2375
2376 sub z3950_search_args {
2377 my $bibrec = shift;
2378
2379 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2380 my $isbn = Business::ISBN->new( $isbn_string );
2381
2382 if (defined $isbn && $isbn->is_valid)
2383 {
2384 if ( ref($bibrec) ) {
2385 $bibrec->{isbn} = $isbn_string;
2386 $bibrec->{title} = undef;
2387 } else {
2388 $bibrec = { isbn => $isbn_string };
2389 }
2390 }
2391 else {
2392 $bibrec = { title => $bibrec } if !ref $bibrec;
2393 }
2394 my $array = [];
2395 for my $field (qw/ lccn isbn issn title author dewey subject /)
2396 {
2397 push @$array, { name => $field, value => $bibrec->{$field} }
2398 if defined $bibrec->{$field};
2399 }
2400 return $array;
2401 }
2402
2403 =head2 GetDistinctValues($field);
2404
2405 C<$field> is a reference to the fields array
2406
2407 =cut
2408
2409 sub GetDistinctValues {
2410 my ($fieldname,$string)=@_;
2411 # returns a reference to a hash of references to branches...
2412 if ($fieldname=~/\./){
2413 my ($table,$column)=split /\./, $fieldname;
2414 my $dbh = C4::Context->dbh;
2415 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2416 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 ");
2417 $sth->execute;
2418 my $elements=$sth->fetchall_arrayref({});
2419 return $elements;
2420 }
2421 else {
2422 $string||= qq("");
2423 my @servers=qw<biblioserver authorityserver>;
2424 my (@zconns,@results);
2425 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2426 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2427 $results[$i] =
2428 $zconns[$i]->scan(
2429 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2430 );
2431 }
2432 # The big moment: asynchronously retrieve results from all servers
2433 my @elements;
2434 _ZOOM_event_loop(
2435 \@zconns,
2436 \@results,
2437 sub {
2438 my ( $i, $size ) = @_;
2439 for ( my $j = 0 ; $j < $size ; $j++ ) {
2440 my %hashscan;
2441 @hashscan{qw(value cnt)} =
2442 $results[ $i - 1 ]->display_term($j);
2443 push @elements, \%hashscan;
2444 }
2445 }
2446 );
2447 return \@elements;
2448 }
2449 }
2450
2451 =head2 _ZOOM_event_loop
2452
2453 _ZOOM_event_loop(\@zconns, \@results, sub {
2454 my ( $i, $size ) = @_;
2455 ....
2456 } );
2457
2458 Processes a ZOOM event loop and passes control to a closure for
2459 processing the results, and destroying the resultsets.
2460
2461 =cut
2462
2463 sub _ZOOM_event_loop {
2464 my ($zconns, $results, $callback) = @_;
2465 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2466 my $ev = $zconns->[ $i - 1 ]->last_event();
2467 if ( $ev == ZOOM::Event::ZEND ) {
2468 next unless $results->[ $i - 1 ];
2469 my $size = $results->[ $i - 1 ]->size();
2470 if ( $size > 0 ) {
2471 $callback->($i, $size);
2472 }
2473 }
2474 }
2475
2476 foreach my $result (@$results) {
2477 $result->destroy();
2478 }
2479 }
2480
2481 =head2 new_record_from_zebra
2482
2483 Given raw data from a searchengine result set, return a MARC::Record object
2484
2485 This helper function is needed to take into account all the involved
2486 system preferences and configuration variables to properly create the
2487 MARC::Record object.
2488
2489 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2490 data. If we are using DOM, then it has to be MARCXML.
2491
2492 If we are using elasticsearch, it'll already be a MARC::Record and this
2493 function needs a new name.
2494
2495 =cut
2496
2497 sub new_record_from_zebra {
2498
2499 my $server = shift;
2500 my $raw_data = shift;
2501 # Set the default indexing modes
2502 my $search_engine = C4::Context->preference("SearchEngine");
2503 if ($search_engine eq 'Elasticsearch') {
2504 return $raw_data;
2505 }
2506 my $index_mode = ( $server eq 'biblioserver' )
2507 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2508 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2509
2510 my $marc_record = eval {
2511 if ( $index_mode eq 'dom' ) {
2512 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2513 } else {
2514 MARC::Record->new_from_usmarc( $raw_data );
2515 }
2516 };
2517
2518 if ($@) {
2519 return;
2520 } else {
2521 return $marc_record;
2522 }
2523
2524 }
2525
2526 END { } # module clean-up code here (global destructor)
2527
2528 1;
2529 __END__
2530
2531 =head1 AUTHOR
2532
2533 Koha Development Team <http://koha-community.org/>
2534
2535 =cut
Koha ILS