C4/Search.pm

   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 Lingua::Stem;
  26 use C4::Search::PazPar2;
  27 use XML::Simple;
  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 vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  40
  41 # set the version for version checking
  42 BEGIN {
  43     $VERSION = 3.07.00.049;
  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( $dbh, $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( $dbh, $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 $biblio = TransformMarcToKoha(C4::Context->dbh,$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     # Initialize variables for the faceted results objects
 341     my $facets_counter = {};
 342     my $facets_info    = {};
 343     my $facets         = getFacets();
 344
 345     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 346
 347     ### LOOP THROUGH THE SERVERS
 348     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 349         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 350
 351 # perform the search, create the results objects
 352 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 353         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 354
 355         #$query_to_use = $simple_query if $scan;
 356         warn $simple_query if ( $scan and $DEBUG );
 357
 358         # Check if we've got a query_type defined, if so, use it
 359         eval {
 360             if ($query_type) {
 361                 if ($query_type =~ /^ccl/) {
 362                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 363                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 364                 } elsif ($query_type =~ /^cql/) {
 365                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 366                 } elsif ($query_type =~ /^pqf/) {
 367                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 368                 } else {
 369                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 370                 }
 371             } elsif ($scan) {
 372                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 373             } else {
 374                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 375             }
 376         };
 377         if ($@) {
 378             warn "WARNING: query problem with $query_to_use " . $@;
 379         }
 380
 381         # Concatenate the sort_by limits and pass them to the results object
 382         # Note: sort will override rank
 383         my $sort_by;
 384         foreach my $sort (@sort_by) {
 385             if ( $sort eq "author_az" || $sort eq "author_asc" ) {
 386                 $sort_by .= "1=1003 <i ";
 387             }
 388             elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
 389                 $sort_by .= "1=1003 >i ";
 390             }
 391             elsif ( $sort eq "popularity_asc" ) {
 392                 $sort_by .= "1=9003 <i ";
 393             }
 394             elsif ( $sort eq "popularity_dsc" ) {
 395                 $sort_by .= "1=9003 >i ";
 396             }
 397             elsif ( $sort eq "call_number_asc" ) {
 398                 $sort_by .= "1=8007  <i ";
 399             }
 400             elsif ( $sort eq "call_number_dsc" ) {
 401                 $sort_by .= "1=8007 >i ";
 402             }
 403             elsif ( $sort eq "pubdate_asc" ) {
 404                 $sort_by .= "1=31 <i ";
 405             }
 406             elsif ( $sort eq "pubdate_dsc" ) {
 407                 $sort_by .= "1=31 >i ";
 408             }
 409             elsif ( $sort eq "acqdate_asc" ) {
 410                 $sort_by .= "1=32 <i ";
 411             }
 412             elsif ( $sort eq "acqdate_dsc" ) {
 413                 $sort_by .= "1=32 >i ";
 414             }
 415             elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
 416                 $sort_by .= "1=4 <i ";
 417             }
 418             elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
 419                 $sort_by .= "1=4 >i ";
 420             }
 421             else {
 422                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 423             }
 424         }
 425         if ( $sort_by && !$scan && $results[$i] ) {
 426             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 427                 warn "WARNING sort $sort_by failed";
 428             }
 429         }
 430     }    # finished looping through servers
 431
 432     # The big moment: asynchronously retrieve results from all servers
 433         _ZOOM_event_loop(
 434             \@zconns,
 435             \@results,
 436             sub {
 437                 my ( $i, $size ) = @_;
 438                 my $results_hash;
 439
 440                 # loop through the results
 441                 $results_hash->{'hits'} = $size;
 442                 my $times;
 443                 if ( $offset + $results_per_page <= $size ) {
 444                     $times = $offset + $results_per_page;
 445                 }
 446                 else {
 447                     $times = $size;
 448                 }
 449
 450                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 451                     my $records_hash;
 452                     my $record;
 453
 454                     ## Check if it's an index scan
 455                     if ($scan) {
 456                         my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
 457
 458                  # here we create a minimal MARC record and hand it off to the
 459                  # template just like a normal result ... perhaps not ideal, but
 460                  # it works for now
 461                         my $tmprecord = MARC::Record->new();
 462                         $tmprecord->encoding('UTF-8');
 463                         my $tmptitle;
 464                         my $tmpauthor;
 465
 466                 # the minimal record in author/title (depending on MARC flavour)
 467                         if ( C4::Context->preference("marcflavour") eq
 468                             "UNIMARC" )
 469                         {
 470                             $tmptitle = MARC::Field->new(
 471                                 '200', ' ', ' ',
 472                                 a => $term,
 473                                 f => $occ
 474                             );
 475                             $tmprecord->append_fields($tmptitle);
 476                         }
 477                         else {
 478                             $tmptitle =
 479                               MARC::Field->new( '245', ' ', ' ', a => $term, );
 480                             $tmpauthor =
 481                               MARC::Field->new( '100', ' ', ' ', a => $occ, );
 482                             $tmprecord->append_fields($tmptitle);
 483                             $tmprecord->append_fields($tmpauthor);
 484                         }
 485                         $results_hash->{'RECORDS'}[$j] =
 486                           $tmprecord->as_usmarc();
 487                     }
 488
 489                     # not an index scan
 490                     else {
 491                         $record = $results[ $i - 1 ]->record($j)->raw();
 492                         # warn "RECORD $j:".$record;
 493                         $results_hash->{'RECORDS'}[$j] = $record;
 494                     }
 495
 496                 }
 497                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 498
 499                 # Fill the facets while we're looping, but only for the
 500                 # biblioserver and not for a scan
 501                 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
 502                     $facets_counter = GetFacets( $results[ $i - 1 ] );
 503                     $facets_info    = _get_facets_info( $facets );
 504                 }
 505
 506                 # BUILD FACETS
 507                 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 508                     for my $link_value (
 509                         sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 510                         keys %$facets_counter
 511                       )
 512                     {
 513                         my $expandable;
 514                         my $number_of_facets;
 515                         my @this_facets_array;
 516                         for my $one_facet (
 517                             sort {
 518                                 $facets_counter->{$link_value}
 519                                   ->{$b} <=> $facets_counter->{$link_value}
 520                                   ->{$a}
 521                             } keys %{ $facets_counter->{$link_value} }
 522                           )
 523                         {
 524                             $number_of_facets++;
 525                             if (   ( $number_of_facets <= 5 )
 526                                 || ( $expanded_facet eq $link_value )
 527                                 || ( $facets_info->{$link_value}->{'expanded'} )
 528                               )
 529                             {
 530
 531 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
 532                                 my $facet_link_value = $one_facet;
 533                                 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
 534
 535                                 # fix the length that will display in the label,
 536                                 my $facet_label_value = $one_facet;
 537                                 my $facet_max_length  = C4::Context->preference(
 538                                     'FacetLabelTruncationLength')
 539                                   || 20;
 540                                 $facet_label_value =
 541                                   substr( $one_facet, 0, $facet_max_length )
 542                                   . "..."
 543                                   if length($facet_label_value) >
 544                                       $facet_max_length;
 545
 546                             # if it's a branch, label by the name, not the code,
 547                                 if ( $link_value =~ /branch/ ) {
 548                                     if (   defined $branches
 549                                         && ref($branches) eq "HASH"
 550                                         && defined $branches->{$one_facet}
 551                                         && ref( $branches->{$one_facet} ) eq
 552                                         "HASH" )
 553                                     {
 554                                         $facet_label_value =
 555                                           $branches->{$one_facet}
 556                                           ->{'branchname'};
 557                                     }
 558                                     else {
 559                                         $facet_label_value = "*";
 560                                     }
 561                                 }
 562
 563                           # if it's a itemtype, label by the name, not the code,
 564                                 if ( $link_value =~ /itype/ ) {
 565                                     if (   defined $itemtypes
 566                                         && ref($itemtypes) eq "HASH"
 567                                         && defined $itemtypes->{$one_facet}
 568                                         && ref( $itemtypes->{$one_facet} ) eq
 569                                         "HASH" )
 570                                     {
 571                                         $facet_label_value =
 572                                           $itemtypes->{$one_facet}
 573                                           ->{translated_description};
 574                                     }
 575                                 }
 576
 577                # also, if it's a location code, use the name instead of the code
 578                                 if ( $link_value =~ /location/ ) {
 579                                     $facet_label_value =
 580                                       GetKohaAuthorisedValueLib( 'LOC',
 581                                         $one_facet, $opac );
 582                                 }
 583
 584                 # but we're down with the whole label being in the link's title.
 585                                 push @this_facets_array,
 586                                   {
 587                                     facet_count =>
 588                                       $facets_counter->{$link_value}
 589                                       ->{$one_facet},
 590                                     facet_label_value => $facet_label_value,
 591                                     facet_title_value => $one_facet,
 592                                     facet_link_value  => $facet_link_value,
 593                                     type_link_value   => $link_value,
 594                                   }
 595                                   if ($facet_label_value);
 596                             }
 597                         }
 598
 599                         # handle expanded option
 600                         unless ( $facets_info->{$link_value}->{'expanded'} ) {
 601                             $expandable = 1
 602                               if ( ( $number_of_facets > 5 )
 603                                 && ( $expanded_facet ne $link_value ) );
 604                         }
 605                         push @facets_loop,
 606                           {
 607                             type_link_value => $link_value,
 608                             type_id         => $link_value . "_id",
 609                             "type_label_"
 610                               . $facets_info->{$link_value}->{'label_value'} =>
 611                               1,
 612                             facets     => \@this_facets_array,
 613                             expandable => $expandable,
 614                             expand     => $link_value,
 615                           }
 616                           unless (
 617                             (
 618                                 $facets_info->{$link_value}->{'label_value'} =~
 619                                 /Libraries/
 620                             )
 621                             and ( C4::Context->preference('singleBranchMode') )
 622                           );
 623                     }
 624                 }
 625             }
 626         );
 627     return ( undef, $results_hashref, \@facets_loop );
 628 }
 629
 630 sub GetFacets {
 631
 632     my $rs = shift;
 633     my $facets;
 634
 635     my $indexing_mode    = C4::Context->config('zebra_bib_index_mode') // 'dom';
 636     my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
 637
 638     if ( $indexing_mode eq 'dom' &&
 639          $use_zebra_facets ) {
 640         $facets = _get_facets_from_zebra( $rs );
 641     } else {
 642         $facets = _get_facets_from_records( $rs );
 643     }
 644
 645     return $facets;
 646 }
 647
 648 sub _get_facets_from_records {
 649
 650     my $rs = shift;
 651
 652     my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
 653     my $facets_config  = getFacets();
 654     my $facets         = {};
 655     my $size           = $rs->size();
 656     my $jmax           = $size > $facets_maxrecs
 657                             ? $facets_maxrecs
 658                             : $size;
 659
 660     for ( my $j = 0 ; $j < $jmax ; $j++ ) {
 661
 662         my $marc_record = new_record_from_zebra (
 663                 'biblioserver',
 664                 $rs->record( $j )->raw()
 665         );
 666
 667         if ( ! defined $marc_record ) {
 668             warn "ERROR DECODING RECORD - $@: " .
 669                 $rs->record( $j )->raw();
 670             next;
 671         }
 672
 673         _get_facets_data_from_record( $marc_record, $facets_config, $facets );
 674     }
 675
 676     return $facets;
 677 }
 678
 679 =head2 _get_facets_data_from_record
 680
 681     C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
 682
 683 Internal function that extracts facets information from a MARC::Record object
 684 and populates $facets_counter for using in getRecords.
 685
 686 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
 687 facets for Zebra).
 688
 689 =cut
 690
 691 sub _get_facets_data_from_record {
 692
 693     my ( $marc_record, $facets, $facets_counter ) = @_;
 694
 695     for my $facet (@$facets) {
 696
 697         my @used_datas = ();
 698
 699         foreach my $tag ( @{ $facet->{ tags } } ) {
 700
 701             # tag number is the first three digits
 702             my $tag_num          = substr( $tag, 0, 3 );
 703             # subfields are the remainder
 704             my $subfield_letters = substr( $tag, 3 );
 705
 706             my @fields = $marc_record->field( $tag_num );
 707             foreach my $field (@fields) {
 708                 # If $field->indicator(1) eq 'z', it means it is a 'see from'
 709                 # field introduced because of IncludeSeeFromInSearches, so skip it
 710                 next if $field->indicator(1) eq 'z';
 711
 712                 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
 713
 714                 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
 715                     push @used_datas, $data;
 716                     $facets_counter->{ $facet->{ idx } }->{ $data }++;
 717                 }
 718             }
 719         }
 720     }
 721 }
 722
 723 =head2 _get_facets_from_zebra
 724
 725     my $facets = _get_facets_from_zebra( $result_set )
 726
 727 Retrieves facets for a specified result set. It loops through the facets defined
 728 in C4::Koha::getFacets and returns a hash with the following structure:
 729
 730    {  facet_idx => {
 731             facet_value => count
 732       },
 733       ...
 734    }
 735
 736 =cut
 737
 738 sub _get_facets_from_zebra {
 739
 740     my $rs = shift;
 741
 742     # save current elementSetName
 743     my $elementSetName = $rs->option( 'elementSetName' );
 744
 745     my $facets_loop = getFacets();
 746     my $facets_data  = {};
 747     # loop through defined facets and fill the facets hashref
 748     foreach my $facet ( @$facets_loop ) {
 749
 750         my $idx = $facet->{ idx };
 751         my $sep = $facet->{ sep };
 752         my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
 753         if ( $facet_values ) {
 754             # we've actually got a result
 755             $facets_data->{ $idx } = $facet_values;
 756         }
 757     }
 758     # set elementSetName to its previous value to avoid side effects
 759     $rs->option( elementSetName => $elementSetName );
 760
 761     return $facets_data;
 762 }
 763
 764 =head2 _get_facet_from_result_set
 765
 766     my $facet_values =
 767         C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
 768
 769 Internal function that extracts facet information for a specific index ($facet_idx) and
 770 returns a hash containing facet values and count:
 771
 772     {
 773         $facet_value => $count ,
 774         ...
 775     }
 776
 777 Warning: this function has the side effect of changing the elementSetName for the result
 778 set. It is a helper function for the main loop, which takes care of backing it up for
 779 restoring.
 780
 781 =cut
 782
 783 sub _get_facet_from_result_set {
 784
 785     my $facet_idx = shift;
 786     my $rs        = shift;
 787     my $sep       = shift;
 788
 789     my $internal_sep  = '<*>';
 790     my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
 791
 792     return if ( ! defined $facet_idx || ! defined $rs );
 793     # zebra's facet element, untokenized index
 794     my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount;
 795     # configure zebra results for retrieving the desired facet
 796     $rs->option( elementSetName => $facet_element );
 797     # get the facet record from result set
 798     my $facet = $rs->record( 0 )->raw;
 799     # if the facet has no restuls...
 800     return if !defined $facet;
 801     # TODO: benchmark DOM vs. SAX performance
 802     my $facet_dom = XML::LibXML->load_xml(
 803       string => ($facet)
 804     );
 805     my @terms = $facet_dom->getElementsByTagName('term');
 806     return if ! @terms;
 807
 808     my $facets = {};
 809     foreach my $term ( @terms ) {
 810         my $facet_value = $term->textContent;
 811         $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
 812         $facets->{ $facet_value } = $term->getAttribute( 'occur' );
 813     }
 814
 815     return $facets;
 816 }
 817
 818 =head2 _get_facets_info
 819
 820     my $facets_info = C4::Search::_get_facets_info( $facets )
 821
 822 Internal function that extracts facets information and properly builds
 823 the data structure needed to render facet labels.
 824
 825 =cut
 826
 827 sub _get_facets_info {
 828
 829     my $facets = shift;
 830
 831     my $facets_info = {};
 832
 833     for my $facet ( @$facets ) {
 834         $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
 835         $facets_info->{ $facet->{ idx } }->{ expanded }    = $facet->{ expanded };
 836     }
 837
 838     return $facets_info;
 839 }
 840
 841 sub pazGetRecords {
 842     my (
 843         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 844         $results_per_page, $offset,       $expanded_facet, $branches,
 845         $query_type,       $scan
 846     ) = @_;
 847
 848     my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
 849     $paz->init();
 850     $paz->search($simple_query);
 851     sleep 1;   # FIXME: WHY?
 852
 853     # do results
 854     my $results_hashref = {};
 855     my $stats = XMLin($paz->stat);
 856     my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
 857
 858     # for a grouped search result, the number of hits
 859     # is the number of groups returned; 'bib_hits' will have
 860     # the total number of bibs.
 861     $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
 862     $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
 863
 864     HIT: foreach my $hit (@{ $results->{'hit'} }) {
 865         my $recid = $hit->{recid}->[0];
 866
 867         my $work_title = $hit->{'md-work-title'}->[0];
 868         my $work_author;
 869         if (exists $hit->{'md-work-author'}) {
 870             $work_author = $hit->{'md-work-author'}->[0];
 871         }
 872         my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
 873
 874         my $result_group = {};
 875         $result_group->{'group_label'} = $group_label;
 876         $result_group->{'group_merge_key'} = $recid;
 877
 878         my $count = 1;
 879         if (exists $hit->{count}) {
 880             $count = $hit->{count}->[0];
 881         }
 882         $result_group->{'group_count'} = $count;
 883
 884         for (my $i = 0; $i < $count; $i++) {
 885             # FIXME -- may need to worry about diacritics here
 886             my $rec = $paz->record($recid, $i);
 887             push @{ $result_group->{'RECORDS'} }, $rec;
 888         }
 889
 890         push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
 891     }
 892
 893     # pass through facets
 894     my $termlist_xml = $paz->termlist('author,subject');
 895     my $terms = XMLin($termlist_xml, forcearray => 1);
 896     my @facets_loop = ();
 897     #die Dumper($results);
 898 #    foreach my $list (sort keys %{ $terms->{'list'} }) {
 899 #        my @facets = ();
 900 #        foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
 901 #            push @facets, {
 902 #                facet_label_value => $facet->{'name'}->[0],
 903 #            };
 904 #        }
 905 #        push @facets_loop, ( {
 906 #            type_label => $list,
 907 #            facets => \@facets,
 908 #        } );
 909 #    }
 910
 911     return ( undef, $results_hashref, \@facets_loop );
 912 }
 913
 914 # TRUNCATION
 915 sub _detect_truncation {
 916     my ( $operand, $index ) = @_;
 917     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 918         @regexpr );
 919     $operand =~ s/^ //g;
 920     my @wordlist = split( /\s/, $operand );
 921     foreach my $word (@wordlist) {
 922         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 923             push @rightlefttruncated, $word;
 924         }
 925         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 926             push @lefttruncated, $word;
 927         }
 928         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 929             push @righttruncated, $word;
 930         }
 931         elsif ( index( $word, "*" ) < 0 ) {
 932             push @nontruncated, $word;
 933         }
 934         else {
 935             push @regexpr, $word;
 936         }
 937     }
 938     return (
 939         \@nontruncated,       \@righttruncated, \@lefttruncated,
 940         \@rightlefttruncated, \@regexpr
 941     );
 942 }
 943
 944 # STEMMING
 945 sub _build_stemmed_operand {
 946     my ($operand,$lang) = @_;
 947     require Lingua::Stem::Snowball ;
 948     my $stemmed_operand=q{};
 949
 950     # If operand contains a digit, it is almost certainly an identifier, and should
 951     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 952     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 953     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 954     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 955     return $operand if $operand =~ /\d/;
 956
 957 # FIXME: the locale should be set based on the user's language and/or search choice
 958     #warn "$lang";
 959     # Make sure we only use the first two letters from the language code
 960     $lang = lc(substr($lang, 0, 2));
 961     # The language codes for the two variants of Norwegian will now be "nb" and "nn",
 962     # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
 963     if ($lang eq 'nb' || $lang eq 'nn') {
 964       $lang = 'no';
 965     }
 966     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 967                                                encoding => "UTF-8" );
 968
 969     my @words = split( / /, $operand );
 970     my @stems = $stemmer->stem(\@words);
 971     for my $stem (@stems) {
 972         $stemmed_operand .= "$stem";
 973         $stemmed_operand .= "?"
 974           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
 975         $stemmed_operand .= " ";
 976     }
 977     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
 978     return $stemmed_operand;
 979 }
 980
 981 # FIELD WEIGHTING
 982 sub _build_weighted_query {
 983
 984 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
 985 # pretty well but could work much better if we had a smarter query parser
 986     my ( $operand, $stemmed_operand, $index ) = @_;
 987     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
 988     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
 989     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
 990     $operand =~ s/"/ /g;    # Bug 7518: searches with quotation marks don't work
 991
 992     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
 993
 994     # Keyword, or, no index specified
 995     if ( ( $index eq 'kw' ) || ( !$index ) ) {
 996         $weighted_query .=
 997           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
 998         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
 999         $weighted_query .= " or Title-cover,phr,r3=\"$operand\"";    # phrase title
1000         $weighted_query .= " or ti,wrdl,r4=\"$operand\"";    # words in title
1001           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
1002           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
1003         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1004           if $fuzzy_enabled;    # add fuzzy, word list
1005         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1006           if ( $stemming and $stemmed_operand )
1007           ;                     # add stemming, right truncation
1008         $weighted_query .= " or wrdl,r9=\"$operand\"";
1009
1010         # embedded sorting: 0 a-z; 1 z-a
1011         # $weighted_query .= ") or (sort1,aut=1";
1012     }
1013
1014     # Barcode searches should skip this process
1015     elsif ( $index eq 'bc' ) {
1016         $weighted_query .= "bc=\"$operand\"";
1017     }
1018
1019     # Authority-number searches should skip this process
1020     elsif ( $index eq 'an' ) {
1021         $weighted_query .= "an=\"$operand\"";
1022     }
1023
1024     # If the index already has more than one qualifier, wrap the operand
1025     # in quotes and pass it back (assumption is that the user knows what they
1026     # are doing and won't appreciate us mucking up their query
1027     elsif ( $index =~ ',' ) {
1028         $weighted_query .= " $index=\"$operand\"";
1029     }
1030
1031     #TODO: build better cases based on specific search indexes
1032     else {
1033         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
1034           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1035         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
1036         $weighted_query .= " or $index,wrdl,r6=\"$operand\"";    # word list index
1037         $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1038           if $fuzzy_enabled;    # add fuzzy, word list
1039         $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1040           if ( $stemming and $stemmed_operand );    # add stemming, right truncation
1041     }
1042
1043     $weighted_query .= "))";                       # close rank specification
1044     return $weighted_query;
1045 }
1046
1047 =head2 getIndexes
1048
1049 Return an array with available indexes.
1050
1051 =cut
1052
1053 sub getIndexes{
1054     my @indexes = (
1055                     # biblio indexes
1056                     'ab',
1057                     'Abstract',
1058                     'acqdate',
1059                     'allrecords',
1060                     'an',
1061                     'Any',
1062                     'at',
1063                     'au',
1064                     'aub',
1065                     'aud',
1066                     'audience',
1067                     'auo',
1068                     'aut',
1069                     'Author',
1070                     'Author-in-order ',
1071                     'Author-personal-bibliography',
1072                     'Authority-Number',
1073                     'authtype',
1074                     'bc',
1075                     'Bib-level',
1076                     'biblionumber',
1077                     'bio',
1078                     'biography',
1079                     'callnum',
1080                     'cfn',
1081                     'Chronological-subdivision',
1082                     'cn-bib-source',
1083                     'cn-bib-sort',
1084                     'cn-class',
1085                     'cn-item',
1086                     'cn-prefix',
1087                     'cn-suffix',
1088                     'cpn',
1089                     'Code-institution',
1090                     'Conference-name',
1091                     'Conference-name-heading',
1092                     'Conference-name-see',
1093                     'Conference-name-seealso',
1094                     'Content-type',
1095                     'Control-number',
1096                     'copydate',
1097                     'Corporate-name',
1098                     'Corporate-name-heading',
1099                     'Corporate-name-see',
1100                     'Corporate-name-seealso',
1101                     'Country-publication',
1102                     'ctype',
1103                     'curriculum',
1104                     'date-entered-on-file',
1105                     'Date-of-acquisition',
1106                     'Date-of-publication',
1107                     'Dewey-classification',
1108                     'Dissertation-information',
1109                     'diss',
1110                     'EAN',
1111                     'extent',
1112                     'fic',
1113                     'fiction',
1114                     'Form-subdivision',
1115                     'format',
1116                     'Geographic-subdivision',
1117                     'he',
1118                     'Heading',
1119                     'Heading-use-main-or-added-entry',
1120                     'Heading-use-series-added-entry ',
1121                     'Heading-use-subject-added-entry',
1122                     'Host-item',
1123                     'id-other',
1124                     'Illustration-code',
1125                     'Index-term-genre',
1126                     'Index-term-uncontrolled',
1127                     'ISBN',
1128                     'isbn',
1129                     'ISSN',
1130                     'issn',
1131                     'itemtype',
1132                     'kw',
1133                     'Koha-Auth-Number',
1134                     'l-format',
1135                     'language',
1136                     'language-original',
1137                     'lc-card',
1138                     'LC-card-number',
1139                     'lcn',
1140                     'lex',
1141                     'llength',
1142                     'ln',
1143                     'ln-audio',
1144                     'ln-subtitle',
1145                     'Local-classification',
1146                     'Local-number',
1147                     'Match-heading',
1148                     'Match-heading-see-from',
1149                     'Material-type',
1150                     'mc-itemtype',
1151                     'mc-rtype',
1152                     'mus',
1153                     'name',
1154                     'Music-number',
1155                     'Name-geographic',
1156                     'Name-geographic-heading',
1157                     'Name-geographic-see',
1158                     'Name-geographic-seealso',
1159                     'nb',
1160                     'Note',
1161                     'notes',
1162                     'ns',
1163                     'nt',
1164                     'pb',
1165                     'Personal-name',
1166                     'Personal-name-heading',
1167                     'Personal-name-see',
1168                     'Personal-name-seealso',
1169                     'pl',
1170                     'Place-publication',
1171                     'pn',
1172                     'popularity',
1173                     'pubdate',
1174                     'Publisher',
1175                     'Provider',
1176                     'pv',
1177                     'Record-control-number',
1178                     'rcn',
1179                     'Record-type',
1180                     'rtype',
1181                     'se',
1182                     'See',
1183                     'See-also',
1184                     'sn',
1185                     'Stock-number',
1186                     'su',
1187                     'Subject',
1188                     'Subject-heading-thesaurus',
1189                     'Subject-name-personal',
1190                     'Subject-subdivision',
1191                     'Summary',
1192                     'Suppress',
1193                     'su-geo',
1194                     'su-na',
1195                     'su-to',
1196                     'su-ut',
1197                     'ut',
1198                     'Term-genre-form',
1199                     'Term-genre-form-heading',
1200                     'Term-genre-form-see',
1201                     'Term-genre-form-seealso',
1202                     'ti',
1203                     'Title',
1204                     'Title-cover',
1205                     'Title-series',
1206                     'Title-uniform',
1207                     'Title-uniform-heading',
1208                     'Title-uniform-see',
1209                     'Title-uniform-seealso',
1210                     'totalissues',
1211                     'yr',
1212
1213                     # items indexes
1214                     'acqsource',
1215                     'barcode',
1216                     'bc',
1217                     'branch',
1218                     'ccode',
1219                     'classification-source',
1220                     'cn-sort',
1221                     'coded-location-qualifier',
1222                     'copynumber',
1223                     'damaged',
1224                     'datelastborrowed',
1225                     'datelastseen',
1226                     'holdingbranch',
1227                     'homebranch',
1228                     'issues',
1229                     'item',
1230                     'itemnumber',
1231                     'itype',
1232                     'Local-classification',
1233                     'location',
1234                     'lost',
1235                     'materials-specified',
1236                     'mc-ccode',
1237                     'mc-itype',
1238                     'mc-loc',
1239                     'notforloan',
1240                     'Number-local-acquisition',
1241                     'onloan',
1242                     'price',
1243                     'renewals',
1244                     'replacementprice',
1245                     'replacementpricedate',
1246                     'reserves',
1247                     'restricted',
1248                     'stack',
1249                     'stocknumber',
1250                     'inv',
1251                     'uri',
1252                     'withdrawn',
1253
1254                     # subject related
1255                   );
1256
1257     return \@indexes;
1258 }
1259
1260 =head2 _handle_exploding_index
1261
1262     my $query = _handle_exploding_index($index, $term)
1263
1264 Callback routine to generate the search for "exploding" indexes (i.e.
1265 those indexes which are turned into multiple or-connected searches based
1266 on authority data).
1267
1268 =cut
1269
1270 sub _handle_exploding_index {
1271     my ($QParser, $filter, $params, $negate, $server) = @_;
1272     my $index = $filter;
1273     my $term = join(' ', @$params);
1274
1275     return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1276
1277     my $marcflavour = C4::Context->preference('marcflavour');
1278
1279     my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1280     my $wantedcodes = '';
1281     my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1282     my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1283     foreach my $auth (@$results) {
1284         my $record = MARC::Record->new_from_usmarc($auth);
1285         my @references = $record->field('5..');
1286         if (@references) {
1287             if ($index eq 'su-br') {
1288                 $wantedcodes = 'g';
1289             } elsif ($index eq 'su-na') {
1290                 $wantedcodes = 'h';
1291             } elsif ($index eq 'su-rl') {
1292                 $wantedcodes = '';
1293             }
1294             foreach my $reference (@references) {
1295                 my $codes = $reference->subfield($codesubfield);
1296                 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1297             }
1298         }
1299     }
1300     my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1301     return $query;
1302 }
1303
1304 =head2 parseQuery
1305
1306     ( $operators, $operands, $indexes, $limits,
1307       $sort_by, $scan, $lang ) =
1308             buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1309
1310 Shim function to ease the transition from buildQuery to a new QueryParser.
1311 This function is called at the beginning of buildQuery, and modifies
1312 buildQuery's input. If it can handle the input, it returns a query that
1313 buildQuery will not try to parse.
1314 =cut
1315
1316 sub parseQuery {
1317     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1318
1319     my @operators = $operators ? @$operators : ();
1320     my @indexes   = $indexes   ? @$indexes   : ();
1321     my @operands  = $operands  ? @$operands  : ();
1322     my @limits    = $limits    ? @$limits    : ();
1323     my @sort_by   = $sort_by   ? @$sort_by   : ();
1324
1325     my $query = $operands[0];
1326     my $index;
1327     my $term;
1328     my $query_desc;
1329
1330     my $QParser;
1331     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1332     undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1333     undef $QParser if (scalar @limits > 0);
1334
1335     if ($QParser)
1336     {
1337         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1338         $query = '';
1339         for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1340             next unless $operands[$ii];
1341             $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1342               if ($query);
1343             if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1344                 $query .= $operands[$ii];
1345             }
1346             elsif ( $indexes[$ii] =~ m/su-/ ) {
1347                 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1348             }
1349             else {
1350                 $query .=
1351                   ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1352             }
1353         }
1354         foreach my $limit (@limits) {
1355         }
1356         if ( scalar(@sort_by) > 0 ) {
1357             my $modifier_re =
1358               '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1359             $query =~ s/$modifier_re//g;
1360             foreach my $modifier (@sort_by) {
1361                 $query .= " #$modifier";
1362             }
1363         }
1364
1365         $query_desc = $query;
1366         $query_desc =~ s/\s+/ /g;
1367         if ( C4::Context->preference("QueryWeightFields") ) {
1368         }
1369         $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1370               { 'target_syntax_callback' => \&_handle_exploding_index } );
1371         $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1372               { 'target_syntax_callback' => \&_handle_exploding_index } );
1373         $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1374               { 'target_syntax_callback' => \&_handle_exploding_index } );
1375         $QParser->parse($query);
1376         $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1377     }
1378     else {
1379         require Koha::QueryParser::Driver::PQF;
1380         my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1381         s/$modifier_re//g for @operands;
1382     }
1383
1384     return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1385 }
1386
1387 =head2 buildQuery
1388
1389 ( $error, $query,
1390 $simple_query, $query_cgi,
1391 $query_desc, $limit,
1392 $limit_cgi, $limit_desc,
1393 $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1394
1395 Build queries and limits in CCL, CGI, Human,
1396 handle truncation, stemming, field weighting, fuzziness, etc.
1397
1398 See verbose embedded documentation.
1399
1400
1401 =cut
1402
1403 sub buildQuery {
1404     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1405
1406     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1407
1408     my $query_desc;
1409     ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1410
1411     # dereference
1412     my @operators = $operators ? @$operators : ();
1413     my @indexes   = $indexes   ? @$indexes   : ();
1414     my @operands  = $operands  ? @$operands  : ();
1415     my @limits    = $limits    ? @$limits    : ();
1416     my @sort_by   = $sort_by   ? @$sort_by   : ();
1417
1418     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1419     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1420     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1421     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1422
1423     my $query        = $operands[0];
1424     my $simple_query = $operands[0];
1425
1426     # initialize the variables we're passing back
1427     my $query_cgi;
1428     my $query_type;
1429
1430     my $limit;
1431     my $limit_cgi;
1432     my $limit_desc;
1433
1434     my $cclq       = 0;
1435     my $cclindexes = getIndexes();
1436     if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1437         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1438             my $dx = lc($1);
1439             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1440         }
1441         $query = "ccl=$query" if $cclq;
1442     }
1443
1444 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1445 # DIAGNOSTIC ONLY!!
1446     if ( $query =~ /^ccl=/ ) {
1447         my $q=$';
1448         # This is needed otherwise ccl= and &limit won't work together, and
1449         # this happens when selecting a subject on the opac-detail page
1450         @limits = grep {!/^$/} @limits;
1451         if ( @limits ) {
1452             $q .= ' and '.join(' and ', @limits);
1453         }
1454         return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $q, '', '', '', 'ccl' );
1455     }
1456     if ( $query =~ /^cql=/ ) {
1457         return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' );
1458     }
1459     if ( $query =~ /^pqf=/ ) {
1460         if ($query_desc) {
1461             $query_cgi = "q=".uri_escape_utf8($query_desc);
1462         } else {
1463             $query_desc = $';
1464             $query_cgi = "q=pqf=".uri_escape_utf8($');
1465         }
1466         return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' );
1467     }
1468
1469     # pass nested queries directly
1470     # FIXME: need better handling of some of these variables in this case
1471     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1472     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1473 #    if ( $query =~ /(\(|\))/ ) {
1474 #        return (
1475 #            undef,              $query, $simple_query, $query_cgi,
1476 #            $query,             $limit, $limit_cgi,    $limit_desc,
1477 #            'ccl'
1478 #        );
1479 #    }
1480
1481 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1482 # query operands and indexes and add stemming, truncation, field weighting, etc.
1483 # Once we do so, we'll end up with a value in $query, just like if we had an
1484 # incoming $query from the user
1485     else {
1486         $query = ""
1487           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1488         my $previous_operand
1489           ;    # a flag used to keep track if there was a previous query
1490                # if there was, we can apply the current operator
1491                # for every operand
1492         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1493
1494             # COMBINE OPERANDS, INDEXES AND OPERATORS
1495             if ( $operands[$i] ) {
1496                 $operands[$i]=~s/^\s+//;
1497
1498               # A flag to determine whether or not to add the index to the query
1499                 my $indexes_set;
1500
1501 # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling
1502                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1503                     $weight_fields    = 0;
1504                     $stemming         = 0;
1505                 } else {
1506                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1507                 }
1508                 my $operand = $operands[$i];
1509                 my $index   = $indexes[$i];
1510
1511                 # Add index-specific attributes
1512
1513                 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1514                 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1515                 #which is processed higher up in this sub. Other than that, year searches are typically
1516                 #handled as limits which are not processed her either.
1517
1518                 # Date of Publication
1519                 if ( $index =~ /yr/ ) {
1520                     #weight_fields/relevance search causes errors with date ranges
1521                     #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1522                     #In the case of YYYY-YYYY, it will return no results
1523                     $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1524                 }
1525
1526                 # Date of Acquisition
1527                 elsif ( $index =~ /acqdate/ ) {
1528                     #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1529                     #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1530                       #top of the results just because they have lots of item records matching that date.
1531                     #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1532                       #irrelevant here
1533                     $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0;
1534                 }
1535                 # ISBN,ISSN,Standard Number, don't need special treatment
1536                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1537                     (
1538                         $stemming,      $auto_truncation,
1539                         $weight_fields, $fuzzy_enabled
1540                     ) = ( 0, 0, 0, 0 );
1541
1542                     if ( $index eq 'nb' ) {
1543                         if ( C4::Context->preference("SearchWithISBNVariations") ) {
1544                             my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1545                             $operands[$i] = $operand =  '(nb=' . join(' OR nb=', @isbns) . ')';
1546                             $indexes[$i] = $index = '';
1547                         }
1548                     }
1549                 }
1550
1551                 if(not $index){
1552                     $index = 'kw';
1553                 }
1554
1555                 # Set default structure attribute (word list)
1556                 my $struct_attr = q{};
1557                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1558                     $struct_attr = ",wrdl";
1559                 }
1560
1561                 # Some helpful index variants
1562                 my $index_plus       = $index . $struct_attr . ':';
1563                 my $index_plus_comma = $index . $struct_attr . ',';
1564
1565                 if ($auto_truncation){
1566                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1567                                                 #FIXME only valid with LTR scripts
1568                                                 $operand=join(" ",map{
1569                                                                                         (index($_,"*")>0?"$_":"$_*")
1570                                                                                          }split (/\s+/,$operand));
1571                                                 warn $operand if $DEBUG;
1572                                         }
1573                                 }
1574
1575                 # Detect Truncation
1576                 my $truncated_operand;
1577                 my( $nontruncated, $righttruncated, $lefttruncated,
1578                     $rightlefttruncated, $regexpr
1579                 ) = _detect_truncation( $operand, $index );
1580                 warn
1581 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1582                   if $DEBUG;
1583
1584                 # Apply Truncation
1585                 if (
1586                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1587                     scalar(@$rightlefttruncated) > 0 )
1588                 {
1589
1590                # Don't field weight or add the index to the query, we do it here
1591                     $indexes_set = 1;
1592                     undef $weight_fields;
1593                     my $previous_truncation_operand;
1594                     if (scalar @$nontruncated) {
1595                         $truncated_operand .= "$index_plus @$nontruncated ";
1596                         $previous_truncation_operand = 1;
1597                     }
1598                     if (scalar @$righttruncated) {
1599                         $truncated_operand .= "and " if $previous_truncation_operand;
1600                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1601                         $previous_truncation_operand = 1;
1602                     }
1603                     if (scalar @$lefttruncated) {
1604                         $truncated_operand .= "and " if $previous_truncation_operand;
1605                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1606                         $previous_truncation_operand = 1;
1607                     }
1608                     if (scalar @$rightlefttruncated) {
1609                         $truncated_operand .= "and " if $previous_truncation_operand;
1610                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1611                         $previous_truncation_operand = 1;
1612                     }
1613                 }
1614                 $operand = $truncated_operand if $truncated_operand;
1615                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1616
1617                 # Handle Stemming
1618                 my $stemmed_operand;
1619                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1620                                                                                 if $stemming;
1621
1622                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1623
1624                 # Handle Field Weighting
1625                 my $weighted_operand;
1626                 if ($weight_fields) {
1627                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1628                     $operand = $weighted_operand;
1629                     $indexes_set = 1;
1630                 }
1631
1632                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1633
1634                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1635                     query => $query,
1636                     query_cgi => $query_cgi,
1637                     query_desc => $query_desc,
1638                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1639                     parsed_operand => $operand,
1640                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1641                     index => $index,
1642                     index_plus => $index_plus,
1643                     indexes_set => $indexes_set,
1644                     previous_operand => $previous_operand,
1645                 });
1646
1647             }    #/if $operands
1648         }    # /for
1649     }
1650     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1651
1652     # add limits
1653     my %group_OR_limits;
1654     my $availability_limit;
1655     foreach my $this_limit (@limits) {
1656         next unless $this_limit;
1657         if ( $this_limit =~ /available/ ) {
1658 #
1659 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1660 ## In English:
1661 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1662             $availability_limit .=
1663 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1664             $limit_cgi  .= "&limit=available";
1665             $limit_desc .= "";
1666         }
1667
1668         # group_OR_limits, prefixed by mc-
1669         # OR every member of the group
1670         elsif ( $this_limit =~ /mc/ ) {
1671             my ($k,$v) = split(/:/, $this_limit,2);
1672             if ( $k !~ /mc-i(tem)?type/ ) {
1673                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1674                 $this_limit =~ tr/"//d;
1675                 $this_limit = $k.":'".$v."'";
1676             }
1677
1678             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1679             $limit_desc      .= " or " if $group_OR_limits{$k};
1680             $group_OR_limits{$k} .= "$this_limit";
1681             $limit_cgi       .= "&limit=" . uri_escape_utf8($this_limit);
1682             $limit_desc      .= " $this_limit";
1683         }
1684
1685         # Regular old limits
1686         else {
1687             $limit .= " and " if $limit || $query;
1688             $limit      .= "$this_limit";
1689             $limit_cgi  .= "&limit=" . uri_escape_utf8($this_limit);
1690             if ($this_limit =~ /^branch:(.+)/) {
1691                 my $branchcode = $1;
1692                 my $branchname = GetBranchName($branchcode);
1693                 if (defined $branchname) {
1694                     $limit_desc .= " branch:$branchname";
1695                 } else {
1696                     $limit_desc .= " $this_limit";
1697                 }
1698             } else {
1699                 $limit_desc .= " $this_limit";
1700             }
1701         }
1702     }
1703     foreach my $k (keys (%group_OR_limits)) {
1704         $limit .= " and " if ( $query || $limit );
1705         $limit .= "($group_OR_limits{$k})";
1706     }
1707     if ($availability_limit) {
1708         $limit .= " and " if ( $query || $limit );
1709         $limit .= "($availability_limit)";
1710     }
1711
1712     # Normalize the query and limit strings
1713     # This is flawed , means we can't search anything with : in it
1714     # if user wants to do ccl or cql, start the query with that
1715 #    $query =~ s/:/=/g;
1716     #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1717     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1718     $query =~ s/(?<=(wrdl)):/=/g;
1719     $query =~ s/(?<=(trn|phr)):/=/g;
1720     $query =~ s/(?<=(st-numeric)):/=/g;
1721     $query =~ s/(?<=(st-year)):/=/g;
1722     $query =~ s/(?<=(st-date-normalized)):/=/g;
1723     $limit =~ s/:/=/g;
1724     for ( $query, $query_desc, $limit, $limit_desc ) {
1725         s/  +/ /g;    # remove extra spaces
1726         s/^ //g;     # remove any beginning spaces
1727         s/ $//g;     # remove any ending spaces
1728         s/==/=/g;    # remove double == from query
1729     }
1730     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1731
1732     for ($query_cgi,$simple_query) {
1733         s/"//g;
1734     }
1735     # append the limit to the query
1736     $query .= " " . $limit;
1737
1738     # Warnings if DEBUG
1739     if ($DEBUG) {
1740         warn "QUERY:" . $query;
1741         warn "QUERY CGI:" . $query_cgi;
1742         warn "QUERY DESC:" . $query_desc;
1743         warn "LIMIT:" . $limit;
1744         warn "LIMIT CGI:" . $limit_cgi;
1745         warn "LIMIT DESC:" . $limit_desc;
1746         warn "---------\nLeave buildQuery\n---------";
1747     }
1748     return (
1749         undef,              $query, $simple_query, $query_cgi,
1750         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1751         $query_type
1752     );
1753 }
1754
1755 =head2 _build_initial_query
1756
1757   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1758
1759   Build a section of the initial query containing indexes, operators, and operands.
1760
1761 =cut
1762
1763 sub _build_initial_query {
1764     my ($params) = @_;
1765
1766     my $operator = "";
1767     if ($params->{previous_operand}){
1768         #If there is a previous operand, add a supplied operator or the default 'and'
1769         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1770     }
1771
1772     #NOTE: indexes_set is typically set when doing truncation or field weighting
1773     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1774
1775     #e.g. "kw,wrdl:test"
1776     #e.g. " and kw,wrdl:test"
1777     $params->{query} .= $operator . $operand;
1778
1779     $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1780     $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1781     $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1782
1783     #e.g. " and kw,wrdl: test"
1784     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1785
1786     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1787
1788     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1789 }
1790
1791 =head2 searchResults
1792
1793   my @search_results = searchResults($search_context, $searchdesc, $hits,
1794                                      $results_per_page, $offset, $scan,
1795                                      @marcresults);
1796
1797 Format results in a form suitable for passing to the template
1798
1799 =cut
1800
1801 # IMO this subroutine is pretty messy still -- it's responsible for
1802 # building the HTML output for the template
1803 sub searchResults {
1804     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1805     my $dbh = C4::Context->dbh;
1806     my @newresults;
1807
1808     require C4::Items;
1809
1810     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1811     my ($is_opac, $hidelostitems);
1812     if ($search_context eq 'opac') {
1813         $hidelostitems = C4::Context->preference('hidelostitems');
1814         $is_opac       = 1;
1815     }
1816
1817     #Build branchnames hash
1818     #find branchname
1819     #get branch information.....
1820     my %branches;
1821     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1822     $bsth->execute();
1823     while ( my $bdata = $bsth->fetchrow_hashref ) {
1824         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1825     }
1826 # FIXME - We build an authorised values hash here, using the default framework
1827 # though it is possible to have different authvals for different fws.
1828
1829     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1830
1831     # get notforloan authorised value list (see $shelflocations  FIXME)
1832     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1833
1834     #Get itemtype hash
1835     my %itemtypes = %{ GetItemTypes() };
1836
1837     #search item field code
1838     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1839
1840     ## find column names of items related to MARC
1841     my %subfieldstosearch;
1842     my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1843     for my $column ( @columns ) {
1844         my ( $tagfield, $tagsubfield ) =
1845           &GetMarcFromKohaField( "items." . $column, "" );
1846         if ( defined $tagsubfield ) {
1847             $subfieldstosearch{$column} = $tagsubfield;
1848         }
1849     }
1850
1851     # handle which records to actually retrieve
1852     my $times;
1853     if ( $hits && $offset + $results_per_page <= $hits ) {
1854         $times = $offset + $results_per_page;
1855     }
1856     else {
1857         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1858     }
1859
1860     my $marcflavour = C4::Context->preference("marcflavour");
1861     # We get the biblionumber position in MARC
1862     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1863
1864     # loop through all of the records we've retrieved
1865     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1866
1867         my $marcrecord;
1868         if ($scan) {
1869             # For Scan searches we built USMARC data
1870             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1871         } else {
1872             # Normal search, render from Zebra's output
1873             $marcrecord = new_record_from_zebra(
1874                 'biblioserver',
1875                 $marcresults->[$i]
1876             );
1877
1878             if ( ! defined $marcrecord ) {
1879                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1880                 next;
1881             }
1882         }
1883
1884         my $fw = $scan
1885              ? undef
1886              : $bibliotag < 10
1887                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1888                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1889
1890         SetUTF8Flag($marcrecord);
1891         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1892         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1893         $oldbiblio->{result_number} = $i + 1;
1894
1895         # add imageurl to itemtype if there is one
1896         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1897
1898         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1899                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1900                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1901                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1902                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1903                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1904
1905                 # edition information, if any
1906         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1907         $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1908  # Build summary if there is one (the summary is defined in the itemtypes table)
1909  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1910         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1911             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1912             my @fields  = $marcrecord->fields();
1913
1914             my $newsummary;
1915             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1916                 my $tags = {};
1917                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1918                     $tag =~ /(.{3})(.)/;
1919                     if($marcrecord->field($1)){
1920                         my @abc = $marcrecord->field($1)->subfield($2);
1921                         $tags->{$tag} = $#abc + 1 ;
1922                     }
1923                 }
1924
1925                 # We catch how many times to repeat this line
1926                 my $max = 0;
1927                 foreach my $tag (keys(%$tags)){
1928                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1929                  }
1930
1931                 # we replace, and repeat each line
1932                 for (my $i = 0 ; $i < $max ; $i++){
1933                     my $newline = $line;
1934
1935                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1936                         $tag =~ /(.{3})(.)/;
1937
1938                         if($marcrecord->field($1)){
1939                             my @repl = $marcrecord->field($1)->subfield($2);
1940                             my $subfieldvalue = $repl[$i];
1941                             $newline =~ s/\[$tag\]/$subfieldvalue/g;
1942                         }
1943                     }
1944                     $newsummary .= "$newline\n";
1945                 }
1946             }
1947
1948             $newsummary =~ s/\[(.*?)]//g;
1949             $newsummary =~ s/\n/<br\/>/g;
1950             $oldbiblio->{summary} = $newsummary;
1951         }
1952
1953         # Pull out the items fields
1954         my @fields = $marcrecord->field($itemtag);
1955         my $marcflavor = C4::Context->preference("marcflavour");
1956         # adding linked items that belong to host records
1957         my $analyticsfield = '773';
1958         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1959             $analyticsfield = '773';
1960         } elsif ($marcflavor eq 'UNIMARC') {
1961             $analyticsfield = '461';
1962         }
1963         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1964             my $hostbiblionumber = $hostfield->subfield("0");
1965             my $linkeditemnumber = $hostfield->subfield("9");
1966             if(!$hostbiblionumber eq undef){
1967                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1968                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1969                 if(!$hostbiblio eq undef){
1970                     my @hostitems = $hostbiblio->field($itemfield);
1971                     foreach my $hostitem (@hostitems){
1972                         if ($hostitem->subfield("9") eq $linkeditemnumber){
1973                             my $linkeditem =$hostitem;
1974                             # append linked items if they exist
1975                             if (!$linkeditem eq undef){
1976                                 push (@fields, $linkeditem);}
1977                         }
1978                     }
1979                 }
1980             }
1981         }
1982
1983         # Setting item statuses for display
1984         my @available_items_loop;
1985         my @onloan_items_loop;
1986         my @other_items_loop;
1987
1988         my $available_items;
1989         my $onloan_items;
1990         my $other_items;
1991
1992         my $ordered_count         = 0;
1993         my $available_count       = 0;
1994         my $onloan_count          = 0;
1995         my $longoverdue_count     = 0;
1996         my $other_count           = 0;
1997         my $withdrawn_count        = 0;
1998         my $itemlost_count        = 0;
1999         my $hideatopac_count      = 0;
2000         my $itembinding_count     = 0;
2001         my $itemdamaged_count     = 0;
2002         my $item_in_transit_count = 0;
2003         my $can_place_holds       = 0;
2004         my $item_onhold_count     = 0;
2005         my $notforloan_count      = 0;
2006         my $items_count           = scalar(@fields);
2007         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2008         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2009         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2010
2011         # loop through every item
2012         foreach my $field (@fields) {
2013             my $item;
2014
2015             # populate the items hash
2016             foreach my $code ( keys %subfieldstosearch ) {
2017                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2018             }
2019             $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2020
2021                 # OPAC hidden items
2022             if ($is_opac) {
2023                 # hidden because lost
2024                 if ($hidelostitems && $item->{itemlost}) {
2025                     $hideatopac_count++;
2026                     next;
2027                 }
2028                 # hidden based on OpacHiddenItems syspref
2029                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2030                 if (scalar @hi) {
2031                     push @hiddenitems, @hi;
2032                     $hideatopac_count++;
2033                     next;
2034                 }
2035             }
2036
2037             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2038             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2039
2040             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2041             if ($item->{$hbranch}) {
2042                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2043             }
2044             elsif ($item->{$otherbranch}) {     # Last resort
2045                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2046             }
2047
2048                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2049 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2050             my $userenv = C4::Context->userenv;
2051             if ( $item->{onloan}
2052                 && !( C4::Members::GetHideLostItemsPreference( $userenv->{'number'} ) && $item->{itemlost} ) )
2053             {
2054                 $onloan_count++;
2055                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2056                 $onloan_items->{$key}->{due_date} = output_pref( { dt => dt_from_string( $item->{onloan} ), dateonly => 1 } );
2057                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2058                 $onloan_items->{$key}->{branchname}     = $item->{branchname};
2059                 $onloan_items->{$key}->{location}       = $shelflocations->{ $item->{location} };
2060                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2061                 $onloan_items->{$key}->{description}    = $item->{description};
2062                 $onloan_items->{$key}->{imageurl} =
2063                   getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2064
2065                 # if something's checked out and lost, mark it as 'long overdue'
2066                 if ( $item->{itemlost} ) {
2067                     $onloan_items->{$key}->{longoverdue}++;
2068                     $longoverdue_count++;
2069                 }
2070                 else {    # can place holds as long as item isn't lost
2071                     $can_place_holds = 1;
2072                 }
2073             }
2074
2075          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2076             else {
2077
2078                 $item->{notforloan}=1 if !$item->{notforloan}  && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2079
2080                 # item is on order
2081                 if ( $item->{notforloan} < 0 ) {
2082                     $ordered_count++;
2083                 } elsif ( $item->{notforloan} > 0 ) {
2084                     $notforloan_count++;
2085                 }
2086
2087                 # is item in transit?
2088                 my $transfertwhen = '';
2089                 my ($transfertfrom, $transfertto);
2090
2091                 # is item on the reserve shelf?
2092                 my $reservestatus = '';
2093
2094                 unless ($item->{withdrawn}
2095                         || $item->{itemlost}
2096                         || $item->{damaged}
2097                         || $item->{notforloan}
2098                         || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2099                         && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2100
2101                     # A couple heuristics to limit how many times
2102                     # we query the database for item transfer information, sacrificing
2103                     # accuracy in some cases for speed;
2104                     #
2105                     # 1. don't query if item has one of the other statuses
2106                     # 2. don't check transit status if the bib has
2107                     #    more than 20 items
2108                     #
2109                     # FIXME: to avoid having the query the database like this, and to make
2110                     #        the in transit status count as unavailable for search limiting,
2111                     #        should map transit status to record indexed in Zebra.
2112                     #
2113                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2114                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2115                 }
2116
2117                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2118                 if (   $item->{withdrawn}
2119                     || $item->{itemlost}
2120                     || $item->{damaged}
2121                     || $item->{notforloan}
2122                     || $reservestatus eq 'Waiting'
2123                     || ($transfertwhen ne ''))
2124                 {
2125                     $withdrawn_count++        if $item->{withdrawn};
2126                     $itemlost_count++        if $item->{itemlost};
2127                     $itemdamaged_count++     if $item->{damaged};
2128                     $item_in_transit_count++ if $transfertwhen ne '';
2129                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2130                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2131
2132                     # can place a hold on a item if
2133                     # not lost nor withdrawn
2134                     # not damaged unless AllowHoldsOnDamagedItems is true
2135                     # item is either for loan or on order (notforloan < 0)
2136                     $can_place_holds = 1
2137                       if (
2138                            !$item->{itemlost}
2139                         && !$item->{withdrawn}
2140                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2141                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2142                       );
2143
2144                     $other_count++;
2145
2146                     my $key = $prefix . $item->{status};
2147                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2148                         $other_items->{$key}->{$_} = $item->{$_};
2149                     }
2150                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2151                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2152                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2153                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2154                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2155                                         $other_items->{$key}->{description} = $item->{description};
2156                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2157                 }
2158                 # item is available
2159                 else {
2160                     $can_place_holds = 1;
2161                     $available_count++;
2162                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2163                                         foreach (qw(branchname itemcallnumber description)) {
2164                         $available_items->{$prefix}->{$_} = $item->{$_};
2165                                         }
2166                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2167                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2168                 }
2169             }
2170         }    # notforloan, item level and biblioitem level
2171
2172         # if all items are hidden, do not show the record
2173         if ($items_count > 0 && $hideatopac_count == $items_count) {
2174             next;
2175         }
2176
2177         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2178         for my $key ( sort keys %$onloan_items ) {
2179             (++$onloanitemscount > $maxitems) and last;
2180             push @onloan_items_loop, $onloan_items->{$key};
2181         }
2182         for my $key ( sort keys %$other_items ) {
2183             (++$otheritemscount > $maxitems) and last;
2184             push @other_items_loop, $other_items->{$key};
2185         }
2186         for my $key ( sort keys %$available_items ) {
2187             (++$availableitemscount > $maxitems) and last;
2188             push @available_items_loop, $available_items->{$key}
2189         }
2190
2191         # XSLT processing of some stuff
2192         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2193         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2194             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2195         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2196         }
2197
2198         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2199         if (!C4::Context->preference("item-level_itypes")) {
2200             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2201                 $can_place_holds = 0;
2202             }
2203         }
2204         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2205         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2206         $oldbiblio->{items_count}          = $items_count;
2207         $oldbiblio->{available_items_loop} = \@available_items_loop;
2208         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2209         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2210         $oldbiblio->{availablecount}       = $available_count;
2211         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2212         $oldbiblio->{onloancount}          = $onloan_count;
2213         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2214         $oldbiblio->{othercount}           = $other_count;
2215         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2216         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2217         $oldbiblio->{itemlostcount}        = $itemlost_count;
2218         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2219         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2220         $oldbiblio->{onholdcount}          = $item_onhold_count;
2221         $oldbiblio->{orderedcount}         = $ordered_count;
2222         $oldbiblio->{notforloancount}      = $notforloan_count;
2223
2224         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2225             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2226             my $subfields = substr $fieldspec, 3;
2227             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2228             my @alternateholdingsinfo = ();
2229             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2230             my $alternateholdingscount = 0;
2231
2232             for my $field (@holdingsfields) {
2233                 my %holding = ( holding => '' );
2234                 my $havesubfield = 0;
2235                 for my $subfield ($field->subfields()) {
2236                     if ((index $subfields, $$subfield[0]) >= 0) {
2237                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2238                         $holding{'holding'} .= $$subfield[1];
2239                         $havesubfield++;
2240                     }
2241                 }
2242                 if ($havesubfield) {
2243                     push(@alternateholdingsinfo, \%holding);
2244                     $alternateholdingscount++;
2245                 }
2246             }
2247
2248             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2249             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2250         }
2251
2252         push( @newresults, $oldbiblio );
2253     }
2254
2255     return @newresults;
2256 }
2257
2258 =head2 SearchAcquisitions
2259     Search for acquisitions
2260 =cut
2261
2262 sub SearchAcquisitions{
2263     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2264
2265     my $dbh=C4::Context->dbh;
2266     # Variable initialization
2267     my $str=qq|
2268     SELECT marcxml
2269     FROM biblio
2270     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2271     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2272     WHERE dateaccessioned BETWEEN ? AND ?
2273     |;
2274
2275     my (@params,@loopcriteria);
2276
2277     push @params, $datebegin->output("iso");
2278     push @params, $dateend->output("iso");
2279
2280     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2281         if(C4::Context->preference("item-level_itypes")){
2282             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2283         }else{
2284             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2285         }
2286         push @params, @$itemtypes;
2287     }
2288
2289     if ($criteria =~/itemtype/){
2290         if(C4::Context->preference("item-level_itypes")){
2291             $str .= "AND items.itype=? ";
2292         }else{
2293             $str .= "AND biblioitems.itemtype=? ";
2294         }
2295
2296         if(scalar(@$itemtypes) == 0){
2297             my $itypes = GetItemTypes();
2298             for my $key (keys %$itypes){
2299                 push @$itemtypes, $key;
2300             }
2301         }
2302
2303         @loopcriteria= @$itemtypes;
2304     }elsif ($criteria=~/itemcallnumber/){
2305         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2306                  OR items.itemcallnumber is NULL
2307                  OR items.itemcallnumber = '')";
2308
2309         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2310     }else {
2311         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2312         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2313     }
2314
2315     if ($orderby =~ /date_desc/){
2316         $str.=" ORDER BY dateaccessioned DESC";
2317     } else {
2318         $str.=" ORDER BY title";
2319     }
2320
2321     my $qdataacquisitions=$dbh->prepare($str);
2322
2323     my @loopacquisitions;
2324     foreach my $value(@loopcriteria){
2325         push @params,$value;
2326         my %cell;
2327         $cell{"title"}=$value;
2328         $cell{"titlecode"}=$value;
2329
2330         eval{$qdataacquisitions->execute(@params);};
2331
2332         if ($@){ warn "recentacquisitions Error :$@";}
2333         else {
2334             my @loopdata;
2335             while (my $data=$qdataacquisitions->fetchrow_hashref){
2336                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2337             }
2338             $cell{"loopdata"}=\@loopdata;
2339         }
2340         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2341         pop @params;
2342     }
2343     $qdataacquisitions->finish;
2344     return \@loopacquisitions;
2345 }
2346
2347 =head2 enabled_staff_search_views
2348
2349 %hash = enabled_staff_search_views()
2350
2351 This function returns a hash that contains three flags obtained from the system
2352 preferences, used to determine whether a particular staff search results view
2353 is enabled.
2354
2355 =over 2
2356
2357 =item C<Output arg:>
2358
2359     * $hash{can_view_MARC} is true only if the MARC view is enabled
2360     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2361     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2362
2363 =item C<usage in the script:>
2364
2365 =back
2366
2367 $template->param ( C4::Search::enabled_staff_search_views );
2368
2369 =cut
2370
2371 sub enabled_staff_search_views
2372 {
2373         return (
2374                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2375                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2376                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2377         );
2378 }
2379
2380 =head2 z3950_search_args
2381
2382 $arrayref = z3950_search_args($matchpoints)
2383
2384 This function returns an array reference that contains the search parameters to be
2385 passed to the Z39.50 search script (z3950_search.pl). The array elements
2386 are hash refs whose keys are name and value, and whose values are the
2387 name of a search parameter, the value of that search parameter and the URL encoded
2388 value of that parameter.
2389
2390 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2391
2392 The search parameter values are obtained from the bibliographic record whose
2393 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2394
2395 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2396 a general purpose search argument. In this case, the returned array contains only
2397 entry: the key is 'title' and the value is derived from $matchpoints.
2398
2399 If a search parameter value is undefined or empty, it is not included in the returned
2400 array.
2401
2402 The returned array reference may be passed directly to the template parameters.
2403
2404 =over 2
2405
2406 =item C<Output arg:>
2407
2408     * $array containing hash refs as described above
2409
2410 =item C<usage in the script:>
2411
2412 =back
2413
2414 $data = Biblio::GetBiblioData($bibno);
2415 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2416
2417 *OR*
2418
2419 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2420
2421 =cut
2422
2423 sub z3950_search_args {
2424     my $bibrec = shift;
2425
2426     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2427     my $isbn = Business::ISBN->new( $isbn_string );
2428
2429     if (defined $isbn && $isbn->is_valid)
2430     {
2431         if ( ref($bibrec) ) {
2432             $bibrec->{isbn} = $isbn_string;
2433             $bibrec->{title} = undef;
2434         } else {
2435             $bibrec = { isbn => $isbn_string };
2436         }
2437     }
2438     else {
2439         $bibrec = { title => $bibrec } if !ref $bibrec;
2440     }
2441     my $array = [];
2442     for my $field (qw/ lccn isbn issn title author dewey subject /)
2443     {
2444         push @$array, { name => $field, value => $bibrec->{$field} }
2445           if defined $bibrec->{$field};
2446     }
2447     return $array;
2448 }
2449
2450 =head2 GetDistinctValues($field);
2451
2452 C<$field> is a reference to the fields array
2453
2454 =cut
2455
2456 sub GetDistinctValues {
2457     my ($fieldname,$string)=@_;
2458     # returns a reference to a hash of references to branches...
2459     if ($fieldname=~/\./){
2460                         my ($table,$column)=split /\./, $fieldname;
2461                         my $dbh = C4::Context->dbh;
2462                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2463                         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 ");
2464                         $sth->execute;
2465                         my $elements=$sth->fetchall_arrayref({});
2466                         return $elements;
2467    }
2468    else {
2469                 $string||= qq("");
2470                 my @servers=qw<biblioserver authorityserver>;
2471                 my (@zconns,@results);
2472         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2473                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2474                         $results[$i] =
2475                       $zconns[$i]->scan(
2476                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2477                       );
2478                 }
2479                 # The big moment: asynchronously retrieve results from all servers
2480                 my @elements;
2481         _ZOOM_event_loop(
2482             \@zconns,
2483             \@results,
2484             sub {
2485                 my ( $i, $size ) = @_;
2486                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2487                     my %hashscan;
2488                     @hashscan{qw(value cnt)} =
2489                       $results[ $i - 1 ]->display_term($j);
2490                     push @elements, \%hashscan;
2491                 }
2492             }
2493         );
2494                 return \@elements;
2495    }
2496 }
2497
2498 =head2 _ZOOM_event_loop
2499
2500     _ZOOM_event_loop(\@zconns, \@results, sub {
2501         my ( $i, $size ) = @_;
2502         ....
2503     } );
2504
2505 Processes a ZOOM event loop and passes control to a closure for
2506 processing the results, and destroying the resultsets.
2507
2508 =cut
2509
2510 sub _ZOOM_event_loop {
2511     my ($zconns, $results, $callback) = @_;
2512     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2513         my $ev = $zconns->[ $i - 1 ]->last_event();
2514         if ( $ev == ZOOM::Event::ZEND ) {
2515             next unless $results->[ $i - 1 ];
2516             my $size = $results->[ $i - 1 ]->size();
2517             if ( $size > 0 ) {
2518                 $callback->($i, $size);
2519             }
2520         }
2521     }
2522
2523     foreach my $result (@$results) {
2524         $result->destroy();
2525     }
2526 }
2527
2528 =head2 new_record_from_zebra
2529
2530 Given raw data from a Zebra result set, return a MARC::Record object
2531
2532 This helper function is needed to take into account all the involved
2533 system preferences and configuration variables to properly create the
2534 MARC::Record object.
2535
2536 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2537 data. If we are using DOM, then it has to be MARCXML.
2538
2539 =cut
2540
2541 sub new_record_from_zebra {
2542
2543     my $server   = shift;
2544     my $raw_data = shift;
2545     # Set the default indexing modes
2546     my $index_mode = ( $server eq 'biblioserver' )
2547                         ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2548                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2549
2550     my $marc_record =  eval {
2551         if ( $index_mode eq 'dom' ) {
2552             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2553         } else {
2554             MARC::Record->new_from_usmarc( $raw_data );
2555         }
2556     };
2557
2558     if ($@) {
2559         return;
2560     } else {
2561         return $marc_record;
2562     }
2563
2564 }
2565
2566 END { }    # module clean-up code here (global destructor)
2567
2568 1;
2569 __END__
2570
2571 =head1 AUTHOR
2572
2573 Koha Development Team <http://koha-community.org/>
2574
2575 =cut