?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Bug 11593 - subtag_registry.sql files contain the same information for each language
[koha.git] / C4 / Search.pm
blob4de0931fe0f64861c8cd28f7738b8493cb450b4d
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 under the
6 # terms of the GNU General Public License as published by the Free Software
7 # Foundation; either version 2 of the License, or (at your option) any later
8 # version.
9 #
10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12 # A PARTICULAR PURPOSE. See the GNU General Public License for more details.
13 #
14 # You should have received a copy of the GNU General Public License along with
15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
16 # Suite 330, Boston, MA 02111-1307 USA
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 Lingua::Stem;
25 use C4::Search::PazPar2;
26 use XML::Simple;
27 use C4::Dates qw(format_date);
28 use C4::Members qw(GetHideLostItemsPreference);
29 use C4::XSLT;
30 use C4::Branch;
31 use C4::Reserves; # GetReserveStatus
32 use C4::Debug;
33 use C4::Charset;
34 use YAML;
35 use URI::Escape;
36 use Business::ISBN;
37 use MARC::Record;
38 use MARC::Field;
39 use utf8;
40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
41
42 # set the version for version checking
43 BEGIN {
44 $VERSION = 3.07.00.049;
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 &PurgeSearchHistory
74 );
75
76 # make all your functions, whether exported or not;
77
78 =head2 FindDuplicate
79
80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
81
82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
83
84 =cut
85
86 sub FindDuplicate {
87 my ($record) = @_;
88 my $dbh = C4::Context->dbh;
89 my $result = TransformMarcToKoha( $dbh, $record, '' );
90 my $sth;
91 my $query;
92 my $search;
93 my $type;
94 my ( $biblionumber, $title );
95
96 # search duplicate on ISBN, easy and fast..
97 # ... normalize first
98 if ( $result->{isbn} ) {
99 $result->{isbn} =~ s/\(.*$//;
100 $result->{isbn} =~ s/\s+$//;
101 $query = "isbn:$result->{isbn}";
102 }
103 else {
104 my $QParser;
105 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
106 my $titleindex;
107 my $authorindex;
108 my $op;
109
110 if ($QParser) {
111 $titleindex = 'title|exact';
112 $authorindex = 'author|exact';
113 $op = '&&';
114 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
115 } else {
116 $titleindex = 'ti,ext';
117 $authorindex = 'au,ext';
118 $op = 'and';
119 }
120
121 $result->{title} =~ s /\\//g;
122 $result->{title} =~ s /\"//g;
123 $result->{title} =~ s /\(//g;
124 $result->{title} =~ s /\)//g;
125
126 # FIXME: instead of removing operators, could just do
127 # quotes around the value
128 $result->{title} =~ s/(and|or|not)//g;
129 $query = "$titleindex:\"$result->{title}\"";
130 if ( $result->{author} ) {
131 $result->{author} =~ s /\\//g;
132 $result->{author} =~ s /\"//g;
133 $result->{author} =~ s /\(//g;
134 $result->{author} =~ s /\)//g;
135
136 # remove valid operators
137 $result->{author} =~ s/(and|or|not)//g;
138 $query .= " $op $authorindex:\"$result->{author}\"";
139 }
140 }
141
142 my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
143 my @results;
144 if (!defined $error) {
145 foreach my $possible_duplicate_record (@{$searchresults}) {
146 my $marcrecord = new_record_from_zebra(
147 'biblioserver',
148 $possible_duplicate_record
149 );
150
151 my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
152
153 # FIXME :: why 2 $biblionumber ?
154 if ($result) {
155 push @results, $result->{'biblionumber'};
156 push @results, $result->{'title'};
157 }
158 }
159 }
160 return @results;
161 }
162
163 =head2 SimpleSearch
164
165 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
166
167 This function provides a simple search API on the bibliographic catalog
168
169 =over 2
170
171 =item C<input arg:>
172
173 * $query can be a simple keyword or a complete CCL query
174 * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
175 * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
176 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
177
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 $biblio = TransformMarcToKoha(C4::Context->dbh,$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 ) = @_;
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;
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;
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
336 # Initialize variables for the ZOOM connection and results object
337 my $zconn;
338 my @zconns;
339 my @results;
340 my $results_hashref = ();
341
342 # Initialize variables for the faceted results objects
343 my $facets_counter = {};
344 my $facets_info = {};
345 my $facets = getFacets();
346
347 my @facets_loop; # stores the ref to array of hashes for template facets loop
348
349 ### LOOP THROUGH THE SERVERS
350 for ( my $i = 0 ; $i < @servers ; $i++ ) {
351 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
352
353 # perform the search, create the results objects
354 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
355 my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
356
357 #$query_to_use = $simple_query if $scan;
358 warn $simple_query if ( $scan and $DEBUG );
359
360 # Check if we've got a query_type defined, if so, use it
361 eval {
362 if ($query_type) {
363 if ($query_type =~ /^ccl/) {
364 $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME)
365 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
366 } elsif ($query_type =~ /^cql/) {
367 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
368 } elsif ($query_type =~ /^pqf/) {
369 $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
370 } else {
371 warn "Unknown query_type '$query_type'. Results undetermined.";
372 }
373 } elsif ($scan) {
374 $results[$i] = $zconns[$i]->scan( new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
375 } else {
376 $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
377 }
378 };
379 if ($@) {
380 warn "WARNING: query problem with $query_to_use " . $@;
381 }
382
383 # Concatenate the sort_by limits and pass them to the results object
384 # Note: sort will override rank
385 my $sort_by;
386 foreach my $sort (@sort_by) {
387 if ( $sort eq "author_az" || $sort eq "author_asc" ) {
388 $sort_by .= "1=1003 <i ";
389 }
390 elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
391 $sort_by .= "1=1003 >i ";
392 }
393 elsif ( $sort eq "popularity_asc" ) {
394 $sort_by .= "1=9003 <i ";
395 }
396 elsif ( $sort eq "popularity_dsc" ) {
397 $sort_by .= "1=9003 >i ";
398 }
399 elsif ( $sort eq "call_number_asc" ) {
400 $sort_by .= "1=8007 <i ";
401 }
402 elsif ( $sort eq "call_number_dsc" ) {
403 $sort_by .= "1=8007 >i ";
404 }
405 elsif ( $sort eq "pubdate_asc" ) {
406 $sort_by .= "1=31 <i ";
407 }
408 elsif ( $sort eq "pubdate_dsc" ) {
409 $sort_by .= "1=31 >i ";
410 }
411 elsif ( $sort eq "acqdate_asc" ) {
412 $sort_by .= "1=32 <i ";
413 }
414 elsif ( $sort eq "acqdate_dsc" ) {
415 $sort_by .= "1=32 >i ";
416 }
417 elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
418 $sort_by .= "1=4 <i ";
419 }
420 elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
421 $sort_by .= "1=4 >i ";
422 }
423 else {
424 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
425 }
426 }
427 if ( $sort_by && !$scan && $results[$i] ) {
428 if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
429 warn "WARNING sort $sort_by failed";
430 }
431 }
432 } # finished looping through servers
433
434 # The big moment: asynchronously retrieve results from all servers
435 _ZOOM_event_loop(
436 \@zconns,
437 \@results,
438 sub {
439 my ( $i, $size ) = @_;
440 my $results_hash;
441
442 # loop through the results
443 $results_hash->{'hits'} = $size;
444 my $times;
445 if ( $offset + $results_per_page <= $size ) {
446 $times = $offset + $results_per_page;
447 }
448 else {
449 $times = $size;
450 }
451
452 for ( my $j = $offset ; $j < $times ; $j++ ) {
453 my $records_hash;
454 my $record;
455
456 ## Check if it's an index scan
457 if ($scan) {
458 my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
459
460 # here we create a minimal MARC record and hand it off to the
461 # template just like a normal result ... perhaps not ideal, but
462 # it works for now
463 my $tmprecord = MARC::Record->new();
464 $tmprecord->encoding('UTF-8');
465 my $tmptitle;
466 my $tmpauthor;
467
468 # the minimal record in author/title (depending on MARC flavour)
469 if ( C4::Context->preference("marcflavour") eq
470 "UNIMARC" )
471 {
472 $tmptitle = MARC::Field->new(
473 '200', ' ', ' ',
474 a => $term,
475 f => $occ
476 );
477 $tmprecord->append_fields($tmptitle);
478 }
479 else {
480 $tmptitle =
481 MARC::Field->new( '245', ' ', ' ', a => $term, );
482 $tmpauthor =
483 MARC::Field->new( '100', ' ', ' ', a => $occ, );
484 $tmprecord->append_fields($tmptitle);
485 $tmprecord->append_fields($tmpauthor);
486 }
487 $results_hash->{'RECORDS'}[$j] =
488 $tmprecord->as_usmarc();
489 }
490
491 # not an index scan
492 else {
493 $record = $results[ $i - 1 ]->record($j)->raw();
494 # warn "RECORD $j:".$record;
495 $results_hash->{'RECORDS'}[$j] = $record;
496 }
497
498 }
499 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
500
501 # Fill the facets while we're looping, but only for the
502 # biblioserver and not for a scan
503 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
504 $facets_counter = GetFacets( $results[ $i - 1 ] );
505 $facets_info = _get_facets_info( $facets );
506 }
507
508 # BUILD FACETS
509 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
510 for my $link_value (
511 sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
512 keys %$facets_counter
513 )
514 {
515 my $expandable;
516 my $number_of_facets;
517 my @this_facets_array;
518 for my $one_facet (
519 sort {
520 $facets_counter->{$link_value}
521 ->{$b} <=> $facets_counter->{$link_value}
522 ->{$a}
523 } keys %{ $facets_counter->{$link_value} }
524 )
525 {
526 $number_of_facets++;
527 if ( ( $number_of_facets <= 5 )
528 || ( $expanded_facet eq $link_value )
529 || ( $facets_info->{$link_value}->{'expanded'} )
530 )
531 {
532
533 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
534 my $facet_link_value = $one_facet;
535 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
536
537 # fix the length that will display in the label,
538 my $facet_label_value = $one_facet;
539 my $facet_max_length = C4::Context->preference(
540 'FacetLabelTruncationLength')
541 || 20;
542 $facet_label_value =
543 substr( $one_facet, 0, $facet_max_length )
544 . "..."
545 if length($facet_label_value) >
546 $facet_max_length;
547
548 # if it's a branch, label by the name, not the code,
549 if ( $link_value =~ /branch/ ) {
550 if ( defined $branches
551 && ref($branches) eq "HASH"
552 && defined $branches->{$one_facet}
553 && ref( $branches->{$one_facet} ) eq
554 "HASH" )
555 {
556 $facet_label_value =
557 $branches->{$one_facet}
558 ->{'branchname'};
559 }
560 else {
561 $facet_label_value = "*";
562 }
563 }
564
565 # if it's a itemtype, label by the name, not the code,
566 if ( $link_value =~ /itype/ ) {
567 if ( defined $itemtypes
568 && ref($itemtypes) eq "HASH"
569 && defined $itemtypes->{$one_facet}
570 && ref( $itemtypes->{$one_facet} ) eq
571 "HASH" )
572 {
573 $facet_label_value =
574 $itemtypes->{$one_facet}
575 ->{'description'};
576 }
577 }
578
579 # also, if it's a location code, use the name instead of the code
580 if ( $link_value =~ /location/ ) {
581 $facet_label_value =
582 GetKohaAuthorisedValueLib( 'LOC',
583 $one_facet, $opac );
584 }
585
586 # but we're down with the whole label being in the link's title.
587 push @this_facets_array,
588 {
589 facet_count =>
590 $facets_counter->{$link_value}
591 ->{$one_facet},
592 facet_label_value => $facet_label_value,
593 facet_title_value => $one_facet,
594 facet_link_value => $facet_link_value,
595 type_link_value => $link_value,
596 }
597 if ($facet_label_value);
598 }
599 }
600
601 # handle expanded option
602 unless ( $facets_info->{$link_value}->{'expanded'} ) {
603 $expandable = 1
604 if ( ( $number_of_facets > 5 )
605 && ( $expanded_facet ne $link_value ) );
606 }
607 push @facets_loop,
608 {
609 type_link_value => $link_value,
610 type_id => $link_value . "_id",
611 "type_label_"
612 . $facets_info->{$link_value}->{'label_value'} =>
613 1,
614 facets => \@this_facets_array,
615 expandable => $expandable,
616 expand => $link_value,
617 }
618 unless (
619 (
620 $facets_info->{$link_value}->{'label_value'} =~
621 /Libraries/
622 )
623 and ( C4::Context->preference('singleBranchMode') )
624 );
625 }
626 }
627 }
628 );
629 return ( undef, $results_hashref, \@facets_loop );
630 }
631
632 sub GetFacets {
633
634 my $rs = shift;
635 my $facets;
636
637 my $indexing_mode = C4::Context->config('zebra_bib_index_mode') // 'dom';
638 my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
639
640 if ( $indexing_mode eq 'dom' &&
641 $use_zebra_facets ) {
642 $facets = _get_facets_from_zebra( $rs );
643 } else {
644 $facets = _get_facets_from_records( $rs );
645 }
646
647 return $facets;
648 }
649
650 sub _get_facets_from_records {
651
652 my $rs = shift;
653
654 my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
655 my $facets_config = getFacets();
656 my $facets = {};
657 my $size = $rs->size();
658 my $jmax = $size > $facets_maxrecs
659 ? $facets_maxrecs
660 : $size;
661
662 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
663
664 my $marc_record = new_record_from_zebra (
665 'biblioserver',
666 $rs->record( $j )->raw()
667 );
668
669 if ( ! defined $marc_record ) {
670 warn "ERROR DECODING RECORD - $@: " .
671 $rs->record( $j )->raw();
672 next;
673 }
674
675 _get_facets_data_from_record( $marc_record, $facets_config, $facets );
676 }
677
678 return $facets;
679 }
680
681 =head2 _get_facets_data_from_record
682
683 C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
684
685 Internal function that extracts facets information from a MARC::Record object
686 and populates $facets_counter for using in getRecords.
687
688 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
689 facets for Zebra).
690
691 =cut
692
693 sub _get_facets_data_from_record {
694
695 my ( $marc_record, $facets, $facets_counter ) = @_;
696
697 for my $facet (@$facets) {
698
699 my @used_datas = ();
700
701 foreach my $tag ( @{ $facet->{ tags } } ) {
702
703 # tag number is the first three digits
704 my $tag_num = substr( $tag, 0, 3 );
705 # subfields are the remainder
706 my $subfield_letters = substr( $tag, 3 );
707
708 my @fields = $marc_record->field( $tag_num );
709 foreach my $field (@fields) {
710 # If $field->indicator(1) eq 'z', it means it is a 'see from'
711 # field introduced because of IncludeSeeFromInSearches, so skip it
712 next if $field->indicator(1) eq 'z';
713
714 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
715
716 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
717 push @used_datas, $data;
718 $facets_counter->{ $facet->{ idx } }->{ $data }++;
719 }
720 }
721 }
722 }
723 }
724
725 =head2 _get_facets_from_zebra
726
727 my $facets = _get_facets_from_zebra( $result_set )
728
729 Retrieves facets for a specified result set. It loops through the facets defined
730 in C4::Koha::getFacets and returns a hash with the following structure:
731
732 { facet_idx => {
733 facet_value => count
734 },
735 ...
736 }
737
738 =cut
739
740 sub _get_facets_from_zebra {
741
742 my $rs = shift;
743
744 # save current elementSetName
745 my $elementSetName = $rs->option( 'elementSetName' );
746
747 my $facets_loop = getFacets();
748 my $facets_data = {};
749 # loop through defined facets and fill the facets hashref
750 foreach my $facet ( @$facets_loop ) {
751
752 my $idx = $facet->{ idx };
753 my $sep = $facet->{ sep };
754 my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
755 if ( $facet_values ) {
756 # we've actually got a result
757 $facets_data->{ $idx } = $facet_values;
758 }
759 }
760 # set elementSetName to its previous value to avoid side effects
761 $rs->option( elementSetName => $elementSetName );
762
763 return $facets_data;
764 }
765
766 =head2 _get_facet_from_result_set
767
768 my $facet_values =
769 C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
770
771 Internal function that extracts facet information for a specific index ($facet_idx) and
772 returns a hash containing facet values and count:
773
774 {
775 $facet_value => $count ,
776 ...
777 }
778
779 Warning: this function has the side effect of changing the elementSetName for the result
780 set. It is a helper function for the main loop, which takes care of backing it up for
781 restoring.
782
783 =cut
784
785 sub _get_facet_from_result_set {
786
787 my $facet_idx = shift;
788 my $rs = shift;
789 my $sep = shift;
790
791 my $internal_sep = '<*>';
792 my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
793
794 return if ( ! defined $facet_idx || ! defined $rs );
795 # zebra's facet element, untokenized index
796 my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
797 # configure zebra results for retrieving the desired facet
798 $rs->option( elementSetName => $facet_element );
799 # get the facet record from result set
800 my $facet = $rs->record( 0 )->raw;
801 # if the facet has no restuls...
802 return if !defined $facet;
803 # TODO: benchmark DOM vs. SAX performance
804 my $facet_dom = XML::LibXML->load_xml(
805 string => ($facet)
806 );
807 my @terms = $facet_dom->getElementsByTagName('term');
808 return if ! @terms;
809
810 my $facets = {};
811 foreach my $term ( @terms ) {
812 my $facet_value = $term->textContent;
813 $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
814 $facets->{ $facet_value } = $term->getAttribute( 'occur' );
815 }
816
817 return $facets;
818 }
819
820 =head2 _get_facets_info
821
822 my $facets_info = C4::Search::_get_facets_info( $facets )
823
824 Internal function that extracts facets information and properly builds
825 the data structure needed to render facet labels.
826
827 =cut
828
829 sub _get_facets_info {
830
831 my $facets = shift;
832
833 my $facets_info = {};
834
835 for my $facet ( @$facets ) {
836 $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
837 $facets_info->{ $facet->{ idx } }->{ expanded } = $facet->{ expanded };
838 }
839
840 return $facets_info;
841 }
842
843 sub pazGetRecords {
844 my (
845 $koha_query, $simple_query, $sort_by_ref, $servers_ref,
846 $results_per_page, $offset, $expanded_facet, $branches,
847 $query_type, $scan
848 ) = @_;
849
850 my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
851 $paz->init();
852 $paz->search($simple_query);
853 sleep 1; # FIXME: WHY?
854
855 # do results
856 my $results_hashref = {};
857 my $stats = XMLin($paz->stat);
858 my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
859
860 # for a grouped search result, the number of hits
861 # is the number of groups returned; 'bib_hits' will have
862 # the total number of bibs.
863 $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
864 $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
865
866 HIT: foreach my $hit (@{ $results->{'hit'} }) {
867 my $recid = $hit->{recid}->[0];
868
869 my $work_title = $hit->{'md-work-title'}->[0];
870 my $work_author;
871 if (exists $hit->{'md-work-author'}) {
872 $work_author = $hit->{'md-work-author'}->[0];
873 }
874 my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
875
876 my $result_group = {};
877 $result_group->{'group_label'} = $group_label;
878 $result_group->{'group_merge_key'} = $recid;
879
880 my $count = 1;
881 if (exists $hit->{count}) {
882 $count = $hit->{count}->[0];
883 }
884 $result_group->{'group_count'} = $count;
885
886 for (my $i = 0; $i < $count; $i++) {
887 # FIXME -- may need to worry about diacritics here
888 my $rec = $paz->record($recid, $i);
889 push @{ $result_group->{'RECORDS'} }, $rec;
890 }
891
892 push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
893 }
894
895 # pass through facets
896 my $termlist_xml = $paz->termlist('author,subject');
897 my $terms = XMLin($termlist_xml, forcearray => 1);
898 my @facets_loop = ();
899 #die Dumper($results);
900 # foreach my $list (sort keys %{ $terms->{'list'} }) {
901 # my @facets = ();
902 # foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
903 # push @facets, {
904 # facet_label_value => $facet->{'name'}->[0],
905 # };
906 # }
907 # push @facets_loop, ( {
908 # type_label => $list,
909 # facets => \@facets,
910 # } );
911 # }
912
913 return ( undef, $results_hashref, \@facets_loop );
914 }
915
916 # STOPWORDS
917 sub _remove_stopwords {
918 my ( $operand, $index ) = @_;
919 my @stopwords_removed;
920
921 # phrase and exact-qualified indexes shouldn't have stopwords removed
922 if ( $index !~ m/,(phr|ext)/ ) {
923
924 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
925 # we use IsAlpha unicode definition, to deal correctly with diacritics.
926 # otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
927 # is a stopword, we'd get "çon" and wouldn't find anything...
928 #
929 foreach ( keys %{ C4::Context->stopwords } ) {
930 next if ( $_ =~ /(and|or|not)/ ); # don't remove operators
931 if ( my ($matched) = ($operand =~
932 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
933 {
934 $operand =~ s/\Q$matched\E/ /gi;
935 push @stopwords_removed, $_;
936 }
937 }
938 }
939 return ( $operand, \@stopwords_removed );
940 }
941
942 # TRUNCATION
943 sub _detect_truncation {
944 my ( $operand, $index ) = @_;
945 my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
946 @regexpr );
947 $operand =~ s/^ //g;
948 my @wordlist = split( /\s/, $operand );
949 foreach my $word (@wordlist) {
950 if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
951 push @rightlefttruncated, $word;
952 }
953 elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
954 push @lefttruncated, $word;
955 }
956 elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
957 push @righttruncated, $word;
958 }
959 elsif ( index( $word, "*" ) < 0 ) {
960 push @nontruncated, $word;
961 }
962 else {
963 push @regexpr, $word;
964 }
965 }
966 return (
967 \@nontruncated, \@righttruncated, \@lefttruncated,
968 \@rightlefttruncated, \@regexpr
969 );
970 }
971
972 # STEMMING
973 sub _build_stemmed_operand {
974 my ($operand,$lang) = @_;
975 require Lingua::Stem::Snowball ;
976 my $stemmed_operand=q{};
977
978 # If operand contains a digit, it is almost certainly an identifier, and should
979 # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which
980 # can contain the letter "X" - for example, _build_stemmend_operand would reduce
981 # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
982 # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098.
983 return $operand if $operand =~ /\d/;
984
985 # FIXME: the locale should be set based on the user's language and/or search choice
986 #warn "$lang";
987 # Make sure we only use the first two letters from the language code
988 $lang = lc(substr($lang, 0, 2));
989 # The language codes for the two variants of Norwegian will now be "nb" and "nn",
990 # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
991 if ($lang eq 'nb' || $lang eq 'nn') {
992 $lang = 'no';
993 }
994 my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
995 encoding => "UTF-8" );
996
997 my @words = split( / /, $operand );
998 my @stems = $stemmer->stem(\@words);
999 for my $stem (@stems) {
1000 $stemmed_operand .= "$stem";
1001 $stemmed_operand .= "?"
1002 unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
1003 $stemmed_operand .= " ";
1004 }
1005 warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1006 return $stemmed_operand;
1007 }
1008
1009 # FIELD WEIGHTING
1010 sub _build_weighted_query {
1011
1012 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1013 # pretty well but could work much better if we had a smarter query parser
1014 my ( $operand, $stemmed_operand, $index ) = @_;
1015 my $stemming = C4::Context->preference("QueryStemming") || 0;
1016 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1017 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1018 $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work
1019
1020 my $weighted_query .= "(rk=("; # Specifies that we're applying rank
1021
1022 # Keyword, or, no index specified
1023 if ( ( $index eq 'kw' ) || ( !$index ) ) {
1024 $weighted_query .=
1025 "Title-cover,ext,r1=\"$operand\""; # exact title-cover
1026 $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title
1027 $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title
1028 $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title
1029 #$weighted_query .= " or any,ext,r4=$operand"; # exact any
1030 #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any
1031 $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1032 if $fuzzy_enabled; # add fuzzy, word list
1033 $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1034 if ( $stemming and $stemmed_operand )
1035 ; # add stemming, right truncation
1036 $weighted_query .= " or wrdl,r9=\"$operand\"";
1037
1038 # embedded sorting: 0 a-z; 1 z-a
1039 # $weighted_query .= ") or (sort1,aut=1";
1040 }
1041
1042 # Barcode searches should skip this process
1043 elsif ( $index eq 'bc' ) {
1044 $weighted_query .= "bc=\"$operand\"";
1045 }
1046
1047 # Authority-number searches should skip this process
1048 elsif ( $index eq 'an' ) {
1049 $weighted_query .= "an=\"$operand\"";
1050 }
1051
1052 # If the index already has more than one qualifier, wrap the operand
1053 # in quotes and pass it back (assumption is that the user knows what they
1054 # are doing and won't appreciate us mucking up their query
1055 elsif ( $index =~ ',' ) {
1056 $weighted_query .= " $index=\"$operand\"";
1057 }
1058
1059 #TODO: build better cases based on specific search indexes
1060 else {
1061 $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index
1062 #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1063 $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index
1064 $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index
1065 $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1066 if $fuzzy_enabled; # add fuzzy, word list
1067 $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1068 if ( $stemming and $stemmed_operand ); # add stemming, right truncation
1069 }
1070
1071 $weighted_query .= "))"; # close rank specification
1072 return $weighted_query;
1073 }
1074
1075 =head2 getIndexes
1076
1077 Return an array with available indexes.
1078
1079 =cut
1080
1081 sub getIndexes{
1082 my @indexes = (
1083 # biblio indexes
1084 'ab',
1085 'Abstract',
1086 'acqdate',
1087 'allrecords',
1088 'an',
1089 'Any',
1090 'at',
1091 'au',
1092 'aub',
1093 'aud',
1094 'audience',
1095 'auo',
1096 'aut',
1097 'Author',
1098 'Author-in-order ',
1099 'Author-personal-bibliography',
1100 'Authority-Number',
1101 'authtype',
1102 'bc',
1103 'Bib-level',
1104 'biblionumber',
1105 'bio',
1106 'biography',
1107 'callnum',
1108 'cfn',
1109 'Chronological-subdivision',
1110 'cn-bib-source',
1111 'cn-bib-sort',
1112 'cn-class',
1113 'cn-item',
1114 'cn-prefix',
1115 'cn-suffix',
1116 'cpn',
1117 'Code-institution',
1118 'Conference-name',
1119 'Conference-name-heading',
1120 'Conference-name-see',
1121 'Conference-name-seealso',
1122 'Content-type',
1123 'Control-number',
1124 'copydate',
1125 'Corporate-name',
1126 'Corporate-name-heading',
1127 'Corporate-name-see',
1128 'Corporate-name-seealso',
1129 'Country-publication',
1130 'ctype',
1131 'curriculum',
1132 'date-entered-on-file',
1133 'Date-of-acquisition',
1134 'Date-of-publication',
1135 'Dewey-classification',
1136 'Dissertation-information',
1137 'EAN',
1138 'extent',
1139 'fic',
1140 'fiction',
1141 'Form-subdivision',
1142 'format',
1143 'Geographic-subdivision',
1144 'he',
1145 'Heading',
1146 'Heading-use-main-or-added-entry',
1147 'Heading-use-series-added-entry ',
1148 'Heading-use-subject-added-entry',
1149 'Host-item',
1150 'id-other',
1151 'Illustration-code',
1152 'Index-term-genre',
1153 'Index-term-uncontrolled',
1154 'ISBN',
1155 'isbn',
1156 'ISSN',
1157 'issn',
1158 'itemtype',
1159 'kw',
1160 'Koha-Auth-Number',
1161 'l-format',
1162 'language',
1163 'language-original',
1164 'lc-card',
1165 'LC-card-number',
1166 'lcn',
1167 'lex',
1168 'llength',
1169 'ln',
1170 'ln-audio',
1171 'ln-subtitle',
1172 'Local-classification',
1173 'Local-number',
1174 'Match-heading',
1175 'Match-heading-see-from',
1176 'Material-type',
1177 'mc-itemtype',
1178 'mc-rtype',
1179 'mus',
1180 'name',
1181 'Music-number',
1182 'Name-geographic',
1183 'Name-geographic-heading',
1184 'Name-geographic-see',
1185 'Name-geographic-seealso',
1186 'nb',
1187 'Note',
1188 'notes',
1189 'ns',
1190 'nt',
1191 'pb',
1192 'Personal-name',
1193 'Personal-name-heading',
1194 'Personal-name-see',
1195 'Personal-name-seealso',
1196 'pl',
1197 'Place-publication',
1198 'pn',
1199 'popularity',
1200 'pubdate',
1201 'Publisher',
1202 'Record-control-number',
1203 'rcn',
1204 'Record-type',
1205 'rtype',
1206 'se',
1207 'See',
1208 'See-also',
1209 'sn',
1210 'Stock-number',
1211 'su',
1212 'Subject',
1213 'Subject-heading-thesaurus',
1214 'Subject-name-personal',
1215 'Subject-subdivision',
1216 'Summary',
1217 'Suppress',
1218 'su-chrono',
1219 'su-corp',
1220 'su-conf',
1221 'su-geo',
1222 'su-form',
1223 'su-genre',
1224 'su-na',
1225 'su-to',
1226 'su-ti',
1227 'su-ut',
1228 'ut',
1229 'Term-genre-form',
1230 'Term-genre-form-heading',
1231 'Term-genre-form-see',
1232 'Term-genre-form-seealso',
1233 'ti',
1234 'Title',
1235 'Title-cover',
1236 'Title-series',
1237 'Title-uniform',
1238 'Title-uniform-heading',
1239 'Title-uniform-see',
1240 'Title-uniform-seealso',
1241 'totalissues',
1242 'yr',
1243
1244 # items indexes
1245 'acqsource',
1246 'barcode',
1247 'bc',
1248 'branch',
1249 'ccode',
1250 'classification-source',
1251 'cn-sort',
1252 'coded-location-qualifier',
1253 'copynumber',
1254 'damaged',
1255 'datelastborrowed',
1256 'datelastseen',
1257 'holdingbranch',
1258 'homebranch',
1259 'issues',
1260 'item',
1261 'itemnumber',
1262 'itype',
1263 'Local-classification',
1264 'location',
1265 'lost',
1266 'materials-specified',
1267 'mc-ccode',
1268 'mc-itype',
1269 'mc-loc',
1270 'notforloan',
1271 'Number-local-acquisition',
1272 'onloan',
1273 'price',
1274 'renewals',
1275 'replacementprice',
1276 'replacementpricedate',
1277 'reserves',
1278 'restricted',
1279 'stack',
1280 'stocknumber',
1281 'inv',
1282 'uri',
1283 'withdrawn',
1284
1285 # subject related
1286 );
1287
1288 return \@indexes;
1289 }
1290
1291 =head2 _handle_exploding_index
1292
1293 my $query = _handle_exploding_index($index, $term)
1294
1295 Callback routine to generate the search for "exploding" indexes (i.e.
1296 those indexes which are turned into multiple or-connected searches based
1297 on authority data).
1298
1299 =cut
1300
1301 sub _handle_exploding_index {
1302 my ($QParser, $filter, $params, $negate, $server) = @_;
1303 my $index = $filter;
1304 my $term = join(' ', @$params);
1305
1306 return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1307
1308 my $marcflavour = C4::Context->preference('marcflavour');
1309
1310 my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1311 my $wantedcodes = '';
1312 my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1313 my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1314 foreach my $auth (@$results) {
1315 my $record = MARC::Record->new_from_usmarc($auth);
1316 my @references = $record->field('5..');
1317 if (@references) {
1318 if ($index eq 'su-br') {
1319 $wantedcodes = 'g';
1320 } elsif ($index eq 'su-na') {
1321 $wantedcodes = 'h';
1322 } elsif ($index eq 'su-rl') {
1323 $wantedcodes = '';
1324 }
1325 foreach my $reference (@references) {
1326 my $codes = $reference->subfield($codesubfield);
1327 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1328 }
1329 }
1330 }
1331 my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1332 return $query;
1333 }
1334
1335 =head2 parseQuery
1336
1337 ( $operators, $operands, $indexes, $limits,
1338 $sort_by, $scan, $lang ) =
1339 buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1340
1341 Shim function to ease the transition from buildQuery to a new QueryParser.
1342 This function is called at the beginning of buildQuery, and modifies
1343 buildQuery's input. If it can handle the input, it returns a query that
1344 buildQuery will not try to parse.
1345 =cut
1346
1347 sub parseQuery {
1348 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1349
1350 my @operators = $operators ? @$operators : ();
1351 my @indexes = $indexes ? @$indexes : ();
1352 my @operands = $operands ? @$operands : ();
1353 my @limits = $limits ? @$limits : ();
1354 my @sort_by = $sort_by ? @$sort_by : ();
1355
1356 my $query = $operands[0];
1357 my $index;
1358 my $term;
1359 my $query_desc;
1360
1361 my $QParser;
1362 $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1363 undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1364 undef $QParser if (scalar @limits > 0);
1365
1366 if ($QParser)
1367 {
1368 $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1369 $query = '';
1370 for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1371 next unless $operands[$ii];
1372 $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1373 if ($query);
1374 if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1375 $query .= $operands[$ii];
1376 }
1377 elsif ( $indexes[$ii] =~ m/su-/ ) {
1378 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1379 }
1380 else {
1381 $query .=
1382 ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1383 }
1384 }
1385 foreach my $limit (@limits) {
1386 }
1387 if ( scalar(@sort_by) > 0 ) {
1388 my $modifier_re =
1389 '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1390 $query =~ s/$modifier_re//g;
1391 foreach my $modifier (@sort_by) {
1392 $query .= " #$modifier";
1393 }
1394 }
1395
1396 $query_desc = $query;
1397 $query_desc =~ s/\s+/ /g;
1398 if ( C4::Context->preference("QueryWeightFields") ) {
1399 }
1400 $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1401 { 'target_syntax_callback' => \&_handle_exploding_index } );
1402 $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1403 { 'target_syntax_callback' => \&_handle_exploding_index } );
1404 $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1405 { 'target_syntax_callback' => \&_handle_exploding_index } );
1406 $QParser->parse($query);
1407 $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1408 }
1409 else {
1410 require Koha::QueryParser::Driver::PQF;
1411 my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1412 s/$modifier_re//g for @operands;
1413 }
1414
1415 return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1416 }
1417
1418 =head2 buildQuery
1419
1420 ( $error, $query,
1421 $simple_query, $query_cgi,
1422 $query_desc, $limit,
1423 $limit_cgi, $limit_desc,
1424 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1425
1426 Build queries and limits in CCL, CGI, Human,
1427 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1428
1429 See verbose embedded documentation.
1430
1431
1432 =cut
1433
1434 sub buildQuery {
1435 my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1436
1437 warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1438
1439 my $query_desc;
1440 ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1441
1442 # dereference
1443 my @operators = $operators ? @$operators : ();
1444 my @indexes = $indexes ? @$indexes : ();
1445 my @operands = $operands ? @$operands : ();
1446 my @limits = $limits ? @$limits : ();
1447 my @sort_by = $sort_by ? @$sort_by : ();
1448
1449 my $stemming = C4::Context->preference("QueryStemming") || 0;
1450 my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0;
1451 my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1452 my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0;
1453 my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1454
1455 my $query = $operands[0];
1456 my $simple_query = $operands[0];
1457
1458 # initialize the variables we're passing back
1459 my $query_cgi;
1460 my $query_type;
1461
1462 my $limit;
1463 my $limit_cgi;
1464 my $limit_desc;
1465
1466 my $stopwords_removed; # flag to determine if stopwords have been removed
1467
1468 my $cclq = 0;
1469 my $cclindexes = getIndexes();
1470 if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1471 while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1472 my $dx = lc($1);
1473 $cclq = grep { lc($_) eq $dx } @$cclindexes;
1474 }
1475 $query = "ccl=$query" if $cclq;
1476 }
1477
1478 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1479 # DIAGNOSTIC ONLY!!
1480 if ( $query =~ /^ccl=/ ) {
1481 my $q=$';
1482 # This is needed otherwise ccl= and &limit won't work together, and
1483 # this happens when selecting a subject on the opac-detail page
1484 @limits = grep {!/^$/} @limits;
1485 if ( @limits ) {
1486 $q .= ' and '.join(' and ', @limits);
1487 }
1488 return ( undef, $q, $q, "q=ccl=".uri_escape($q), $q, '', '', '', '', 'ccl' );
1489 }
1490 if ( $query =~ /^cql=/ ) {
1491 return ( undef, $', $', "q=cql=".uri_escape($'), $', '', '', '', '', 'cql' );
1492 }
1493 if ( $query =~ /^pqf=/ ) {
1494 if ($query_desc) {
1495 $query_cgi = "q=".uri_escape($query_desc);
1496 } else {
1497 $query_desc = $';
1498 $query_cgi = "q=pqf=".uri_escape($');
1499 }
1500 return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1501 }
1502
1503 # pass nested queries directly
1504 # FIXME: need better handling of some of these variables in this case
1505 # Nested queries aren't handled well and this implementation is flawed and causes users to be
1506 # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1507 # if ( $query =~ /(\(|\))/ ) {
1508 # return (
1509 # undef, $query, $simple_query, $query_cgi,
1510 # $query, $limit, $limit_cgi, $limit_desc,
1511 # $stopwords_removed, 'ccl'
1512 # );
1513 # }
1514
1515 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1516 # query operands and indexes and add stemming, truncation, field weighting, etc.
1517 # Once we do so, we'll end up with a value in $query, just like if we had an
1518 # incoming $query from the user
1519 else {
1520 $query = ""
1521 ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1522 my $previous_operand
1523 ; # a flag used to keep track if there was a previous query
1524 # if there was, we can apply the current operator
1525 # for every operand
1526 for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1527
1528 # COMBINE OPERANDS, INDEXES AND OPERATORS
1529 if ( $operands[$i] ) {
1530 $operands[$i]=~s/^\s+//;
1531
1532 # A flag to determine whether or not to add the index to the query
1533 my $indexes_set;
1534
1535 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1536 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1537 $weight_fields = 0;
1538 $stemming = 0;
1539 $remove_stopwords = 0;
1540 } else {
1541 $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1542 }
1543 my $operand = $operands[$i];
1544 my $index = $indexes[$i];
1545
1546 # Add index-specific attributes
1547
1548 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1549 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1550 #which is processed higher up in this sub. Other than that, year searches are typically
1551 #handled as limits which are not processed her either.
1552
1553 # Date of Publication
1554 if ( $index =~ /yr/ ) {
1555 #weight_fields/relevance search causes errors with date ranges
1556 #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1557 #In the case of YYYY-YYYY, it will return no results
1558 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1559 }
1560
1561 # Date of Acquisition
1562 elsif ( $index =~ /acqdate/ ) {
1563 #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1564 #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1565 #top of the results just because they have lots of item records matching that date.
1566 #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1567 #irrelevant here
1568 #remove_stopwords doesn't function anymore so is irrelevant
1569 $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1570 }
1571 # ISBN,ISSN,Standard Number, don't need special treatment
1572 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1573 (
1574 $stemming, $auto_truncation,
1575 $weight_fields, $fuzzy_enabled,
1576 $remove_stopwords
1577 ) = ( 0, 0, 0, 0, 0 );
1578
1579 }
1580
1581 if(not $index){
1582 $index = 'kw';
1583 }
1584
1585 # Set default structure attribute (word list)
1586 my $struct_attr = q{};
1587 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1588 $struct_attr = ",wrdl";
1589 }
1590
1591 # Some helpful index variants
1592 my $index_plus = $index . $struct_attr . ':';
1593 my $index_plus_comma = $index . $struct_attr . ',';
1594
1595 # Remove Stopwords
1596 if ($remove_stopwords) {
1597 ( $operand, $stopwords_removed ) =
1598 _remove_stopwords( $operand, $index );
1599 warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1600 warn "REMOVED STOPWORDS: @$stopwords_removed"
1601 if ( $stopwords_removed && $DEBUG );
1602 }
1603
1604 if ($auto_truncation){
1605 unless ( $index =~ /,(st-|phr|ext)/ ) {
1606 #FIXME only valid with LTR scripts
1607 $operand=join(" ",map{
1608 (index($_,"*")>0?"$_":"$_*")
1609 }split (/\s+/,$operand));
1610 warn $operand if $DEBUG;
1611 }
1612 }
1613
1614 # Detect Truncation
1615 my $truncated_operand;
1616 my( $nontruncated, $righttruncated, $lefttruncated,
1617 $rightlefttruncated, $regexpr
1618 ) = _detect_truncation( $operand, $index );
1619 warn
1620 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1621 if $DEBUG;
1622
1623 # Apply Truncation
1624 if (
1625 scalar(@$righttruncated) + scalar(@$lefttruncated) +
1626 scalar(@$rightlefttruncated) > 0 )
1627 {
1628
1629 # Don't field weight or add the index to the query, we do it here
1630 $indexes_set = 1;
1631 undef $weight_fields;
1632 my $previous_truncation_operand;
1633 if (scalar @$nontruncated) {
1634 $truncated_operand .= "$index_plus @$nontruncated ";
1635 $previous_truncation_operand = 1;
1636 }
1637 if (scalar @$righttruncated) {
1638 $truncated_operand .= "and " if $previous_truncation_operand;
1639 $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1640 $previous_truncation_operand = 1;
1641 }
1642 if (scalar @$lefttruncated) {
1643 $truncated_operand .= "and " if $previous_truncation_operand;
1644 $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1645 $previous_truncation_operand = 1;
1646 }
1647 if (scalar @$rightlefttruncated) {
1648 $truncated_operand .= "and " if $previous_truncation_operand;
1649 $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1650 $previous_truncation_operand = 1;
1651 }
1652 }
1653 $operand = $truncated_operand if $truncated_operand;
1654 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1655
1656 # Handle Stemming
1657 my $stemmed_operand;
1658 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1659 if $stemming;
1660
1661 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1662
1663 # Handle Field Weighting
1664 my $weighted_operand;
1665 if ($weight_fields) {
1666 $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1667 $operand = $weighted_operand;
1668 $indexes_set = 1;
1669 }
1670
1671 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1672
1673 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1674 query => $query,
1675 query_cgi => $query_cgi,
1676 query_desc => $query_desc,
1677 operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1678 parsed_operand => $operand,
1679 original_operand => ($operands[$i]) ? $operands[$i] : '',
1680 index => $index,
1681 index_plus => $index_plus,
1682 indexes_set => $indexes_set,
1683 previous_operand => $previous_operand,
1684 });
1685
1686 } #/if $operands
1687 } # /for
1688 }
1689 warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1690
1691 # add limits
1692 my %group_OR_limits;
1693 my $availability_limit;
1694 foreach my $this_limit (@limits) {
1695 next unless $this_limit;
1696 if ( $this_limit =~ /available/ ) {
1697 #
1698 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1699 ## In English:
1700 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1701 $availability_limit .=
1702 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1703 $limit_cgi .= "&limit=available";
1704 $limit_desc .= "";
1705 }
1706
1707 # group_OR_limits, prefixed by mc-
1708 # OR every member of the group
1709 elsif ( $this_limit =~ /mc/ ) {
1710 my ($k,$v) = split(/:/, $this_limit,2);
1711 if ( $k !~ /mc-i(tem)?type/ ) {
1712 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1713 $this_limit =~ tr/"//d;
1714 $this_limit = $k.":\"".$v."\"";
1715 }
1716
1717 $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1718 $limit_desc .= " or " if $group_OR_limits{$k};
1719 $group_OR_limits{$k} .= "$this_limit";
1720 $limit_cgi .= "&limit=" . uri_escape($this_limit);
1721 $limit_desc .= " $this_limit";
1722 }
1723
1724 # Regular old limits
1725 else {
1726 $limit .= " and " if $limit || $query;
1727 $limit .= "$this_limit";
1728 $limit_cgi .= "&limit=" . uri_escape($this_limit);
1729 if ($this_limit =~ /^branch:(.+)/) {
1730 my $branchcode = $1;
1731 my $branchname = GetBranchName($branchcode);
1732 if (defined $branchname) {
1733 $limit_desc .= " branch:$branchname";
1734 } else {
1735 $limit_desc .= " $this_limit";
1736 }
1737 } else {
1738 $limit_desc .= " $this_limit";
1739 }
1740 }
1741 }
1742 foreach my $k (keys (%group_OR_limits)) {
1743 $limit .= " and " if ( $query || $limit );
1744 $limit .= "($group_OR_limits{$k})";
1745 }
1746 if ($availability_limit) {
1747 $limit .= " and " if ( $query || $limit );
1748 $limit .= "($availability_limit)";
1749 }
1750
1751 # Normalize the query and limit strings
1752 # This is flawed , means we can't search anything with : in it
1753 # if user wants to do ccl or cql, start the query with that
1754 # $query =~ s/:/=/g;
1755 #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1756 $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1757 $query =~ s/(?<=(wrdl)):/=/g;
1758 $query =~ s/(?<=(trn|phr)):/=/g;
1759 $query =~ s/(?<=(st-numeric)):/=/g;
1760 $query =~ s/(?<=(st-year)):/=/g;
1761 $query =~ s/(?<=(st-date-normalized)):/=/g;
1762 $limit =~ s/:/=/g;
1763 for ( $query, $query_desc, $limit, $limit_desc ) {
1764 s/ +/ /g; # remove extra spaces
1765 s/^ //g; # remove any beginning spaces
1766 s/ $//g; # remove any ending spaces
1767 s/==/=/g; # remove double == from query
1768 }
1769 $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1770
1771 for ($query_cgi,$simple_query) {
1772 s/"//g;
1773 }
1774 # append the limit to the query
1775 $query .= " " . $limit;
1776
1777 # Warnings if DEBUG
1778 if ($DEBUG) {
1779 warn "QUERY:" . $query;
1780 warn "QUERY CGI:" . $query_cgi;
1781 warn "QUERY DESC:" . $query_desc;
1782 warn "LIMIT:" . $limit;
1783 warn "LIMIT CGI:" . $limit_cgi;
1784 warn "LIMIT DESC:" . $limit_desc;
1785 warn "---------\nLeave buildQuery\n---------";
1786 }
1787 return (
1788 undef, $query, $simple_query, $query_cgi,
1789 $query_desc, $limit, $limit_cgi, $limit_desc,
1790 $stopwords_removed, $query_type
1791 );
1792 }
1793
1794 =head2 _build_initial_query
1795
1796 ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1797
1798 Build a section of the initial query containing indexes, operators, and operands.
1799
1800 =cut
1801
1802 sub _build_initial_query {
1803 my ($params) = @_;
1804
1805 my $operator = "";
1806 if ($params->{previous_operand}){
1807 #If there is a previous operand, add a supplied operator or the default 'and'
1808 $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1809 }
1810
1811 #NOTE: indexes_set is typically set when doing truncation or field weighting
1812 my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1813
1814 #e.g. "kw,wrdl:test"
1815 #e.g. " and kw,wrdl:test"
1816 $params->{query} .= $operator . $operand;
1817
1818 $params->{query_cgi} .= "&op=".uri_escape($operator) if $operator;
1819 $params->{query_cgi} .= "&idx=".uri_escape($params->{index}) if $params->{index};
1820 $params->{query_cgi} .= "&q=".uri_escape($params->{original_operand}) if $params->{original_operand};
1821
1822 #e.g. " and kw,wrdl: test"
1823 $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1824
1825 $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1826
1827 return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1828 }
1829
1830 =head2 searchResults
1831
1832 my @search_results = searchResults($search_context, $searchdesc, $hits,
1833 $results_per_page, $offset, $scan,
1834 @marcresults);
1835
1836 Format results in a form suitable for passing to the template
1837
1838 =cut
1839
1840 # IMO this subroutine is pretty messy still -- it's responsible for
1841 # building the HTML output for the template
1842 sub searchResults {
1843 my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1844 my $dbh = C4::Context->dbh;
1845 my @newresults;
1846
1847 require C4::Items;
1848
1849 $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1850 my ($is_opac, $hidelostitems);
1851 if ($search_context eq 'opac') {
1852 $hidelostitems = C4::Context->preference('hidelostitems');
1853 $is_opac = 1;
1854 }
1855
1856 #Build branchnames hash
1857 #find branchname
1858 #get branch information.....
1859 my %branches;
1860 my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1861 $bsth->execute();
1862 while ( my $bdata = $bsth->fetchrow_hashref ) {
1863 $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1864 }
1865 # FIXME - We build an authorised values hash here, using the default framework
1866 # though it is possible to have different authvals for different fws.
1867
1868 my $shelflocations =GetKohaAuthorisedValues('items.location','');
1869
1870 # get notforloan authorised value list (see $shelflocations FIXME)
1871 my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1872
1873 #Build itemtype hash
1874 #find itemtype & itemtype image
1875 my %itemtypes;
1876 $bsth =
1877 $dbh->prepare(
1878 "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1879 );
1880 $bsth->execute();
1881 while ( my $bdata = $bsth->fetchrow_hashref ) {
1882 foreach (qw(description imageurl summary notforloan)) {
1883 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1884 }
1885 }
1886
1887 #search item field code
1888 my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1889
1890 ## find column names of items related to MARC
1891 my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1892 $sth2->execute;
1893 my %subfieldstosearch;
1894 while ( ( my $column ) = $sth2->fetchrow ) {
1895 my ( $tagfield, $tagsubfield ) =
1896 &GetMarcFromKohaField( "items." . $column, "" );
1897 if ( defined $tagsubfield ) {
1898 $subfieldstosearch{$column} = $tagsubfield;
1899 }
1900 }
1901
1902 # handle which records to actually retrieve
1903 my $times;
1904 if ( $hits && $offset + $results_per_page <= $hits ) {
1905 $times = $offset + $results_per_page;
1906 }
1907 else {
1908 $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it?
1909 }
1910
1911 my $marcflavour = C4::Context->preference("marcflavour");
1912 # We get the biblionumber position in MARC
1913 my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1914
1915 # loop through all of the records we've retrieved
1916 for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1917
1918 my $marcrecord;
1919 if ($scan) {
1920 # For Scan searches we built USMARC data
1921 $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1922 } else {
1923 # Normal search, render from Zebra's output
1924 $marcrecord = new_record_from_zebra(
1925 'biblioserver',
1926 $marcresults->[$i]
1927 );
1928
1929 if ( ! defined $marcrecord ) {
1930 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1931 next;
1932 }
1933 }
1934
1935 my $fw = $scan
1936 ? undef
1937 : $bibliotag < 10
1938 ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1939 : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1940 my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1941 $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1942 $oldbiblio->{result_number} = $i + 1;
1943
1944 # add imageurl to itemtype if there is one
1945 $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1946
1947 $oldbiblio->{'authorised_value_images'} = ($search_context eq 'opac' && C4::Context->preference('AuthorisedValueImages')) || ($search_context eq 'intranet' && C4::Context->preference('StaffAuthorisedValueImages')) ? C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) ) : [];
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} }->{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
1991 if (! utf8::is_utf8($subfieldvalue)) {
1992 utf8::decode($subfieldvalue);
1993 }
1994
1995 $newline =~ s/\[$tag\]/$subfieldvalue/g;
1996 }
1997 }
1998 $newsummary .= "$newline\n";
1999 }
2000 }
2001
2002 $newsummary =~ s/\[(.*?)]//g;
2003 $newsummary =~ s/\n/<br\/>/g;
2004 $oldbiblio->{summary} = $newsummary;
2005 }
2006
2007 # Pull out the items fields
2008 my @fields = $marcrecord->field($itemtag);
2009 my $marcflavor = C4::Context->preference("marcflavour");
2010 # adding linked items that belong to host records
2011 my $analyticsfield = '773';
2012 if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2013 $analyticsfield = '773';
2014 } elsif ($marcflavor eq 'UNIMARC') {
2015 $analyticsfield = '461';
2016 }
2017 foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2018 my $hostbiblionumber = $hostfield->subfield("0");
2019 my $linkeditemnumber = $hostfield->subfield("9");
2020 if(!$hostbiblionumber eq undef){
2021 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2022 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2023 if(!$hostbiblio eq undef){
2024 my @hostitems = $hostbiblio->field($itemfield);
2025 foreach my $hostitem (@hostitems){
2026 if ($hostitem->subfield("9") eq $linkeditemnumber){
2027 my $linkeditem =$hostitem;
2028 # append linked items if they exist
2029 if (!$linkeditem eq undef){
2030 push (@fields, $linkeditem);}
2031 }
2032 }
2033 }
2034 }
2035 }
2036
2037 # Setting item statuses for display
2038 my @available_items_loop;
2039 my @onloan_items_loop;
2040 my @other_items_loop;
2041
2042 my $available_items;
2043 my $onloan_items;
2044 my $other_items;
2045
2046 my $ordered_count = 0;
2047 my $available_count = 0;
2048 my $onloan_count = 0;
2049 my $longoverdue_count = 0;
2050 my $other_count = 0;
2051 my $withdrawn_count = 0;
2052 my $itemlost_count = 0;
2053 my $hideatopac_count = 0;
2054 my $itembinding_count = 0;
2055 my $itemdamaged_count = 0;
2056 my $item_in_transit_count = 0;
2057 my $can_place_holds = 0;
2058 my $item_onhold_count = 0;
2059 my $notforloan_count = 0;
2060 my $items_count = scalar(@fields);
2061 my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2062 my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2063 my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2064
2065 # loop through every item
2066 foreach my $field (@fields) {
2067 my $item;
2068
2069 # populate the items hash
2070 foreach my $code ( keys %subfieldstosearch ) {
2071 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2072 }
2073 $item->{description} = $itemtypes{ $item->{itype} }{description};
2074
2075 # OPAC hidden items
2076 if ($is_opac) {
2077 # hidden because lost
2078 if ($hidelostitems && $item->{itemlost}) {
2079 $hideatopac_count++;
2080 next;
2081 }
2082 # hidden based on OpacHiddenItems syspref
2083 my @hi = C4::Items::GetHiddenItemnumbers($item);
2084 if (scalar @hi) {
2085 push @hiddenitems, @hi;
2086 $hideatopac_count++;
2087 next;
2088 }
2089 }
2090
2091 my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch');
2092 my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2093
2094 # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2095 if ($item->{$hbranch}) {
2096 $item->{'branchname'} = $branches{$item->{$hbranch}};
2097 }
2098 elsif ($item->{$otherbranch}) { # Last resort
2099 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2100 }
2101
2102 my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2103 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2104 my $userenv = C4::Context->userenv;
2105 if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
2106 $onloan_count++;
2107 my $key = $prefix . $item->{onloan} . $item->{barcode};
2108 $onloan_items->{$key}->{due_date} = format_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} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2115 # if something's checked out and lost, mark it as 'long overdue'
2116 if ( $item->{itemlost} ) {
2117 $onloan_items->{$prefix}->{longoverdue}++;
2118 $longoverdue_count++;
2119 } else { # can place holds as long as item isn't lost
2120 $can_place_holds = 1;
2121 }
2122 }
2123
2124 # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2125 else {
2126
2127 # item is on order
2128 if ( $item->{notforloan} < 0 ) {
2129 $ordered_count++;
2130 } elsif ( $item->{notforloan} > 0 ) {
2131 $notforloan_count++;
2132 }
2133
2134 # is item in transit?
2135 my $transfertwhen = '';
2136 my ($transfertfrom, $transfertto);
2137
2138 # is item on the reserve shelf?
2139 my $reservestatus = '';
2140
2141 unless ($item->{withdrawn}
2142 || $item->{itemlost}
2143 || $item->{damaged}
2144 || $item->{notforloan}
2145 || $items_count > 20) {
2146
2147 # A couple heuristics to limit how many times
2148 # we query the database for item transfer information, sacrificing
2149 # accuracy in some cases for speed;
2150 #
2151 # 1. don't query if item has one of the other statuses
2152 # 2. don't check transit status if the bib has
2153 # more than 20 items
2154 #
2155 # FIXME: to avoid having the query the database like this, and to make
2156 # the in transit status count as unavailable for search limiting,
2157 # should map transit status to record indexed in Zebra.
2158 #
2159 ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2160 $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber}, $oldbiblio->{biblionumber} );
2161 }
2162
2163 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2164 if ( $item->{withdrawn}
2165 || $item->{itemlost}
2166 || $item->{damaged}
2167 || $item->{notforloan}
2168 || $reservestatus eq 'Waiting'
2169 || ($transfertwhen ne ''))
2170 {
2171 $withdrawn_count++ if $item->{withdrawn};
2172 $itemlost_count++ if $item->{itemlost};
2173 $itemdamaged_count++ if $item->{damaged};
2174 $item_in_transit_count++ if $transfertwhen ne '';
2175 $item_onhold_count++ if $reservestatus eq 'Waiting';
2176 $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2177
2178 # can place a hold on a item if
2179 # not lost nor withdrawn
2180 # not damaged unless AllowHoldsOnDamagedItems is true
2181 # item is either for loan or on order (notforloan < 0)
2182 $can_place_holds = 1
2183 if (
2184 !$item->{itemlost}
2185 && !$item->{withdrawn}
2186 && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2187 && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2188 );
2189
2190 $other_count++;
2191
2192 my $key = $prefix . $item->{status};
2193 foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2194 $other_items->{$key}->{$_} = $item->{$_};
2195 }
2196 $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2197 $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2198 $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2199 $other_items->{$key}->{count}++ if $item->{$hbranch};
2200 $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2201 $other_items->{$key}->{description} = $item->{description};
2202 $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2203 }
2204 # item is available
2205 else {
2206 $can_place_holds = 1;
2207 $available_count++;
2208 $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2209 foreach (qw(branchname itemcallnumber description)) {
2210 $available_items->{$prefix}->{$_} = $item->{$_};
2211 }
2212 $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2213 $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2214 }
2215 }
2216 } # notforloan, item level and biblioitem level
2217
2218 # if all items are hidden, do not show the record
2219 if ($items_count > 0 && $hideatopac_count == $items_count) {
2220 next;
2221 }
2222
2223 my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2224 for my $key ( sort keys %$onloan_items ) {
2225 (++$onloanitemscount > $maxitems) and last;
2226 push @onloan_items_loop, $onloan_items->{$key};
2227 }
2228 for my $key ( sort keys %$other_items ) {
2229 (++$otheritemscount > $maxitems) and last;
2230 push @other_items_loop, $other_items->{$key};
2231 }
2232 for my $key ( sort keys %$available_items ) {
2233 (++$availableitemscount > $maxitems) and last;
2234 push @available_items_loop, $available_items->{$key}
2235 }
2236
2237 # XSLT processing of some stuff
2238 SetUTF8Flag($marcrecord);
2239 warn $marcrecord->as_formatted if $DEBUG;
2240 my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2241 if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2242 $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2243 # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2244 }
2245
2246 # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2247 if (!C4::Context->preference("item-level_itypes")) {
2248 if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2249 $can_place_holds = 0;
2250 }
2251 }
2252 $oldbiblio->{norequests} = 1 unless $can_place_holds;
2253 $oldbiblio->{itemsplural} = 1 if $items_count > 1;
2254 $oldbiblio->{items_count} = $items_count;
2255 $oldbiblio->{available_items_loop} = \@available_items_loop;
2256 $oldbiblio->{onloan_items_loop} = \@onloan_items_loop;
2257 $oldbiblio->{other_items_loop} = \@other_items_loop;
2258 $oldbiblio->{availablecount} = $available_count;
2259 $oldbiblio->{availableplural} = 1 if $available_count > 1;
2260 $oldbiblio->{onloancount} = $onloan_count;
2261 $oldbiblio->{onloanplural} = 1 if $onloan_count > 1;
2262 $oldbiblio->{othercount} = $other_count;
2263 $oldbiblio->{otherplural} = 1 if $other_count > 1;
2264 $oldbiblio->{withdrawncount} = $withdrawn_count;
2265 $oldbiblio->{itemlostcount} = $itemlost_count;
2266 $oldbiblio->{damagedcount} = $itemdamaged_count;
2267 $oldbiblio->{intransitcount} = $item_in_transit_count;
2268 $oldbiblio->{onholdcount} = $item_onhold_count;
2269 $oldbiblio->{orderedcount} = $ordered_count;
2270 $oldbiblio->{notforloancount} = $notforloan_count;
2271
2272 if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2273 my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2274 my $subfields = substr $fieldspec, 3;
2275 my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2276 my @alternateholdingsinfo = ();
2277 my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2278 my $alternateholdingscount = 0;
2279
2280 for my $field (@holdingsfields) {
2281 my %holding = ( holding => '' );
2282 my $havesubfield = 0;
2283 for my $subfield ($field->subfields()) {
2284 if ((index $subfields, $$subfield[0]) >= 0) {
2285 $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2286 $holding{'holding'} .= $$subfield[1];
2287 $havesubfield++;
2288 }
2289 }
2290 if ($havesubfield) {
2291 push(@alternateholdingsinfo, \%holding);
2292 $alternateholdingscount++;
2293 }
2294 }
2295
2296 $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2297 $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2298 }
2299
2300 push( @newresults, $oldbiblio );
2301 }
2302
2303 return @newresults;
2304 }
2305
2306 =head2 SearchAcquisitions
2307 Search for acquisitions
2308 =cut
2309
2310 sub SearchAcquisitions{
2311 my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2312
2313 my $dbh=C4::Context->dbh;
2314 # Variable initialization
2315 my $str=qq|
2316 SELECT marcxml
2317 FROM biblio
2318 LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2319 LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2320 WHERE dateaccessioned BETWEEN ? AND ?
2321 |;
2322
2323 my (@params,@loopcriteria);
2324
2325 push @params, $datebegin->output("iso");
2326 push @params, $dateend->output("iso");
2327
2328 if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2329 if(C4::Context->preference("item-level_itypes")){
2330 $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2331 }else{
2332 $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2333 }
2334 push @params, @$itemtypes;
2335 }
2336
2337 if ($criteria =~/itemtype/){
2338 if(C4::Context->preference("item-level_itypes")){
2339 $str .= "AND items.itype=? ";
2340 }else{
2341 $str .= "AND biblioitems.itemtype=? ";
2342 }
2343
2344 if(scalar(@$itemtypes) == 0){
2345 my $itypes = GetItemTypes();
2346 for my $key (keys %$itypes){
2347 push @$itemtypes, $key;
2348 }
2349 }
2350
2351 @loopcriteria= @$itemtypes;
2352 }elsif ($criteria=~/itemcallnumber/){
2353 $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2354 OR items.itemcallnumber is NULL
2355 OR items.itemcallnumber = '')";
2356
2357 @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2358 }else {
2359 $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2360 @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2361 }
2362
2363 if ($orderby =~ /date_desc/){
2364 $str.=" ORDER BY dateaccessioned DESC";
2365 } else {
2366 $str.=" ORDER BY title";
2367 }
2368
2369 my $qdataacquisitions=$dbh->prepare($str);
2370
2371 my @loopacquisitions;
2372 foreach my $value(@loopcriteria){
2373 push @params,$value;
2374 my %cell;
2375 $cell{"title"}=$value;
2376 $cell{"titlecode"}=$value;
2377
2378 eval{$qdataacquisitions->execute(@params);};
2379
2380 if ($@){ warn "recentacquisitions Error :$@";}
2381 else {
2382 my @loopdata;
2383 while (my $data=$qdataacquisitions->fetchrow_hashref){
2384 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2385 }
2386 $cell{"loopdata"}=\@loopdata;
2387 }
2388 push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2389 pop @params;
2390 }
2391 $qdataacquisitions->finish;
2392 return \@loopacquisitions;
2393 }
2394
2395 =head2 enabled_staff_search_views
2396
2397 %hash = enabled_staff_search_views()
2398
2399 This function returns a hash that contains three flags obtained from the system
2400 preferences, used to determine whether a particular staff search results view
2401 is enabled.
2402
2403 =over 2
2404
2405 =item C<Output arg:>
2406
2407 * $hash{can_view_MARC} is true only if the MARC view is enabled
2408 * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2409 * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2410
2411 =item C<usage in the script:>
2412
2413 =back
2414
2415 $template->param ( C4::Search::enabled_staff_search_views );
2416
2417 =cut
2418
2419 sub enabled_staff_search_views
2420 {
2421 return (
2422 can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view
2423 can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view
2424 can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view
2425 );
2426 }
2427
2428 sub PurgeSearchHistory{
2429 my ($pSearchhistory)=@_;
2430 my $dbh = C4::Context->dbh;
2431 my $sth = $dbh->prepare("DELETE FROM search_history WHERE time < DATE_SUB( NOW(), INTERVAL ? DAY )");
2432 $sth->execute($pSearchhistory) or die $dbh->errstr;
2433 }
2434
2435 =head2 z3950_search_args
2436
2437 $arrayref = z3950_search_args($matchpoints)
2438
2439 This function returns an array reference that contains the search parameters to be
2440 passed to the Z39.50 search script (z3950_search.pl). The array elements
2441 are hash refs whose keys are name and value, and whose values are the
2442 name of a search parameter, the value of that search parameter and the URL encoded
2443 value of that parameter.
2444
2445 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2446
2447 The search parameter values are obtained from the bibliographic record whose
2448 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2449
2450 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2451 a general purpose search argument. In this case, the returned array contains only
2452 entry: the key is 'title' and the value is derived from $matchpoints.
2453
2454 If a search parameter value is undefined or empty, it is not included in the returned
2455 array.
2456
2457 The returned array reference may be passed directly to the template parameters.
2458
2459 =over 2
2460
2461 =item C<Output arg:>
2462
2463 * $array containing hash refs as described above
2464
2465 =item C<usage in the script:>
2466
2467 =back
2468
2469 $data = Biblio::GetBiblioData($bibno);
2470 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2471
2472 *OR*
2473
2474 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2475
2476 =cut
2477
2478 sub z3950_search_args {
2479 my $bibrec = shift;
2480
2481 my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2482 my $isbn = Business::ISBN->new( $isbn_string );
2483
2484 if (defined $isbn && $isbn->is_valid)
2485 {
2486 if ( ref($bibrec) ) {
2487 $bibrec->{isbn} = $isbn_string;
2488 $bibrec->{title} = undef;
2489 } else {
2490 $bibrec = { isbn => $isbn_string };
2491 }
2492 }
2493 else {
2494 $bibrec = { title => $bibrec } if !ref $bibrec;
2495 }
2496 my $array = [];
2497 for my $field (qw/ lccn isbn issn title author dewey subject /)
2498 {
2499 push @$array, { name => $field, value => $bibrec->{$field} }
2500 if defined $bibrec->{$field};
2501 }
2502 return $array;
2503 }
2504
2505 =head2 GetDistinctValues($field);
2506
2507 C<$field> is a reference to the fields array
2508
2509 =cut
2510
2511 sub GetDistinctValues {
2512 my ($fieldname,$string)=@_;
2513 # returns a reference to a hash of references to branches...
2514 if ($fieldname=~/\./){
2515 my ($table,$column)=split /\./, $fieldname;
2516 my $dbh = C4::Context->dbh;
2517 warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2518 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 ");
2519 $sth->execute;
2520 my $elements=$sth->fetchall_arrayref({});
2521 return $elements;
2522 }
2523 else {
2524 $string||= qq("");
2525 my @servers=qw<biblioserver authorityserver>;
2526 my (@zconns,@results);
2527 for ( my $i = 0 ; $i < @servers ; $i++ ) {
2528 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2529 $results[$i] =
2530 $zconns[$i]->scan(
2531 ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2532 );
2533 }
2534 # The big moment: asynchronously retrieve results from all servers
2535 my @elements;
2536 _ZOOM_event_loop(
2537 \@zconns,
2538 \@results,
2539 sub {
2540 my ( $i, $size ) = @_;
2541 for ( my $j = 0 ; $j < $size ; $j++ ) {
2542 my %hashscan;
2543 @hashscan{qw(value cnt)} =
2544 $results[ $i - 1 ]->display_term($j);
2545 push @elements, \%hashscan;
2546 }
2547 }
2548 );
2549 return \@elements;
2550 }
2551 }
2552
2553 =head2 _ZOOM_event_loop
2554
2555 _ZOOM_event_loop(\@zconns, \@results, sub {
2556 my ( $i, $size ) = @_;
2557 ....
2558 } );
2559
2560 Processes a ZOOM event loop and passes control to a closure for
2561 processing the results, and destroying the resultsets.
2562
2563 =cut
2564
2565 sub _ZOOM_event_loop {
2566 my ($zconns, $results, $callback) = @_;
2567 while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2568 my $ev = $zconns->[ $i - 1 ]->last_event();
2569 if ( $ev == ZOOM::Event::ZEND ) {
2570 next unless $results->[ $i - 1 ];
2571 my $size = $results->[ $i - 1 ]->size();
2572 if ( $size > 0 ) {
2573 $callback->($i, $size);
2574 }
2575 }
2576 }
2577
2578 foreach my $result (@$results) {
2579 $result->destroy();
2580 }
2581 }
2582
2583 =head2 new_record_from_zebra
2584
2585 Given raw data from a Zebra result set, return a MARC::Record object
2586
2587 This helper function is needed to take into account all the involved
2588 system preferences and configuration variables to properly create the
2589 MARC::Record object.
2590
2591 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2592 data. If we are using DOM, then it has to be MARCXML.
2593
2594 =cut
2595
2596 sub new_record_from_zebra {
2597
2598 my $server = shift;
2599 my $raw_data = shift;
2600 # Set the default indexing modes
2601 my $index_mode = ( $server eq 'biblioserver' )
2602 ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2603 : C4::Context->config('zebra_auth_index_mode') // 'dom';
2604
2605 my $marc_record = eval {
2606 if ( $index_mode eq 'dom' ) {
2607 MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2608 } else {
2609 MARC::Record->new_from_usmarc( $raw_data );
2610 }
2611 };
2612
2613 if ($@) {
2614 return;
2615 } else {
2616 return $marc_record;
2617 }
2618
2619 }
2620
2621 END { } # module clean-up code here (global destructor)
2622
2623 1;
2624 __END__
2625
2626 =head1 AUTHOR
2627
2628 Koha Development Team <http://koha-community.org/>
2629
2630 =cut
Koha ILS