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 # STOPWORDS
 915 sub _remove_stopwords {
 916     my ( $operand, $index ) = @_;
 917     my @stopwords_removed;
 918
 919     # phrase and exact-qualified indexes shouldn't have stopwords removed
 920     if ( $index !~ m/,(phr|ext)/ ) {
 921
 922 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
 923 #       we use IsAlpha unicode definition, to deal correctly with diacritics.
 924 #       otherwise, a French word like "leçon" would be split into "le" "çon", "le"
 925 #       is a stopword, we'd get "çon" and wouldn't find anything...
 926 #
 927                 foreach ( keys %{ C4::Context->stopwords } ) {
 928                         next if ( $_ =~ /(and|or|not)/ );    # don't remove operators
 929                         if ( my ($matched) = ($operand =~
 930                                 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
 931                         {
 932                                 $operand =~ s/\Q$matched\E/ /gi;
 933                                 push @stopwords_removed, $_;
 934                         }
 935                 }
 936         }
 937     return ( $operand, \@stopwords_removed );
 938 }
 939
 940 # TRUNCATION
 941 sub _detect_truncation {
 942     my ( $operand, $index ) = @_;
 943     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 944         @regexpr );
 945     $operand =~ s/^ //g;
 946     my @wordlist = split( /\s/, $operand );
 947     foreach my $word (@wordlist) {
 948         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 949             push @rightlefttruncated, $word;
 950         }
 951         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 952             push @lefttruncated, $word;
 953         }
 954         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 955             push @righttruncated, $word;
 956         }
 957         elsif ( index( $word, "*" ) < 0 ) {
 958             push @nontruncated, $word;
 959         }
 960         else {
 961             push @regexpr, $word;
 962         }
 963     }
 964     return (
 965         \@nontruncated,       \@righttruncated, \@lefttruncated,
 966         \@rightlefttruncated, \@regexpr
 967     );
 968 }
 969
 970 # STEMMING
 971 sub _build_stemmed_operand {
 972     my ($operand,$lang) = @_;
 973     require Lingua::Stem::Snowball ;
 974     my $stemmed_operand=q{};
 975
 976     # If operand contains a digit, it is almost certainly an identifier, and should
 977     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 978     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 979     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 980     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 981     return $operand if $operand =~ /\d/;
 982
 983 # FIXME: the locale should be set based on the user's language and/or search choice
 984     #warn "$lang";
 985     # Make sure we only use the first two letters from the language code
 986     $lang = lc(substr($lang, 0, 2));
 987     # The language codes for the two variants of Norwegian will now be "nb" and "nn",
 988     # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
 989     if ($lang eq 'nb' || $lang eq 'nn') {
 990       $lang = 'no';
 991     }
 992     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 993                                                encoding => "UTF-8" );
 994
 995     my @words = split( / /, $operand );
 996     my @stems = $stemmer->stem(\@words);
 997     for my $stem (@stems) {
 998         $stemmed_operand .= "$stem";
 999         $stemmed_operand .= "?"
1000           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
1001         $stemmed_operand .= " ";
1002     }
1003     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1004     return $stemmed_operand;
1005 }
1006
1007 # FIELD WEIGHTING
1008 sub _build_weighted_query {
1009
1010 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1011 # pretty well but could work much better if we had a smarter query parser
1012     my ( $operand, $stemmed_operand, $index ) = @_;
1013     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
1014     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1015     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
1016     $operand =~ s/"/ /g;    # Bug 7518: searches with quotation marks don't work
1017
1018     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
1019
1020     # Keyword, or, no index specified
1021     if ( ( $index eq 'kw' ) || ( !$index ) ) {
1022         $weighted_query .=
1023           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
1024         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
1025         $weighted_query .= " or Title-cover,phr,r3=\"$operand\"";    # phrase title
1026         $weighted_query .= " or ti,wrdl,r4=\"$operand\"";    # words in title
1027           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
1028           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
1029         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1030           if $fuzzy_enabled;    # add fuzzy, word list
1031         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1032           if ( $stemming and $stemmed_operand )
1033           ;                     # add stemming, right truncation
1034         $weighted_query .= " or wrdl,r9=\"$operand\"";
1035
1036         # embedded sorting: 0 a-z; 1 z-a
1037         # $weighted_query .= ") or (sort1,aut=1";
1038     }
1039
1040     # Barcode searches should skip this process
1041     elsif ( $index eq 'bc' ) {
1042         $weighted_query .= "bc=\"$operand\"";
1043     }
1044
1045     # Authority-number searches should skip this process
1046     elsif ( $index eq 'an' ) {
1047         $weighted_query .= "an=\"$operand\"";
1048     }
1049
1050     # If the index already has more than one qualifier, wrap the operand
1051     # in quotes and pass it back (assumption is that the user knows what they
1052     # are doing and won't appreciate us mucking up their query
1053     elsif ( $index =~ ',' ) {
1054         $weighted_query .= " $index=\"$operand\"";
1055     }
1056
1057     #TODO: build better cases based on specific search indexes
1058     else {
1059         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
1060           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1061         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
1062         $weighted_query .= " or $index,wrdl,r6=\"$operand\"";    # word list index
1063         $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1064           if $fuzzy_enabled;    # add fuzzy, word list
1065         $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1066           if ( $stemming and $stemmed_operand );    # add stemming, right truncation
1067     }
1068
1069     $weighted_query .= "))";                       # close rank specification
1070     return $weighted_query;
1071 }
1072
1073 =head2 getIndexes
1074
1075 Return an array with available indexes.
1076
1077 =cut
1078
1079 sub getIndexes{
1080     my @indexes = (
1081                     # biblio indexes
1082                     'ab',
1083                     'Abstract',
1084                     'acqdate',
1085                     'allrecords',
1086                     'an',
1087                     'Any',
1088                     'at',
1089                     'au',
1090                     'aub',
1091                     'aud',
1092                     'audience',
1093                     'auo',
1094                     'aut',
1095                     'Author',
1096                     'Author-in-order ',
1097                     'Author-personal-bibliography',
1098                     'Authority-Number',
1099                     'authtype',
1100                     'bc',
1101                     'Bib-level',
1102                     'biblionumber',
1103                     'bio',
1104                     'biography',
1105                     'callnum',
1106                     'cfn',
1107                     'Chronological-subdivision',
1108                     'cn-bib-source',
1109                     'cn-bib-sort',
1110                     'cn-class',
1111                     'cn-item',
1112                     'cn-prefix',
1113                     'cn-suffix',
1114                     'cpn',
1115                     'Code-institution',
1116                     'Conference-name',
1117                     'Conference-name-heading',
1118                     'Conference-name-see',
1119                     'Conference-name-seealso',
1120                     'Content-type',
1121                     'Control-number',
1122                     'copydate',
1123                     'Corporate-name',
1124                     'Corporate-name-heading',
1125                     'Corporate-name-see',
1126                     'Corporate-name-seealso',
1127                     'Country-publication',
1128                     'ctype',
1129                     'curriculum',
1130                     'date-entered-on-file',
1131                     'Date-of-acquisition',
1132                     'Date-of-publication',
1133                     'Date-time-last-modified',
1134                     'Dewey-classification',
1135                     'Dissertation-information',
1136                     'diss',
1137                     'dtlm',
1138                     'EAN',
1139                     'extent',
1140                     'fic',
1141                     'fiction',
1142                     'Form-subdivision',
1143                     'format',
1144                     'Geographic-subdivision',
1145                     'he',
1146                     'Heading',
1147                     'Heading-use-main-or-added-entry',
1148                     'Heading-use-series-added-entry ',
1149                     'Heading-use-subject-added-entry',
1150                     'Host-item',
1151                     'id-other',
1152                     'Illustration-code',
1153                     'Index-term-genre',
1154                     'Index-term-uncontrolled',
1155                     'ISBN',
1156                     'isbn',
1157                     'ISSN',
1158                     'issn',
1159                     'itemtype',
1160                     'kw',
1161                     'Koha-Auth-Number',
1162                     'l-format',
1163                     'language',
1164                     'language-original',
1165                     'lc-card',
1166                     'LC-card-number',
1167                     'lcn',
1168                     'lex',
1169                     'llength',
1170                     'ln',
1171                     'ln-audio',
1172                     'ln-subtitle',
1173                     'Local-classification',
1174                     'Local-number',
1175                     'Match-heading',
1176                     'Match-heading-see-from',
1177                     'Material-type',
1178                     'mc-itemtype',
1179                     'mc-rtype',
1180                     'mus',
1181                     'name',
1182                     'Music-number',
1183                     'Name-geographic',
1184                     'Name-geographic-heading',
1185                     'Name-geographic-see',
1186                     'Name-geographic-seealso',
1187                     'nb',
1188                     'Note',
1189                     'notes',
1190                     'ns',
1191                     'nt',
1192                     'pb',
1193                     'Personal-name',
1194                     'Personal-name-heading',
1195                     'Personal-name-see',
1196                     'Personal-name-seealso',
1197                     'pl',
1198                     'Place-publication',
1199                     'pn',
1200                     'popularity',
1201                     'pubdate',
1202                     'Publisher',
1203                     'Provider',
1204                     'pv',
1205                     'Record-control-number',
1206                     'rcn',
1207                     'Record-type',
1208                     'rtype',
1209                     'se',
1210                     'See',
1211                     'See-also',
1212                     'sn',
1213                     'Stock-number',
1214                     'su',
1215                     'Subject',
1216                     'Subject-heading-thesaurus',
1217                     'Subject-name-personal',
1218                     'Subject-subdivision',
1219                     'Summary',
1220                     'Suppress',
1221                     'su-geo',
1222                     'su-na',
1223                     'su-to',
1224                     'su-ut',
1225                     'ut',
1226                     'Term-genre-form',
1227                     'Term-genre-form-heading',
1228                     'Term-genre-form-see',
1229                     'Term-genre-form-seealso',
1230                     'ti',
1231                     'Title',
1232                     'Title-cover',
1233                     'Title-series',
1234                     'Title-uniform',
1235                     'Title-uniform-heading',
1236                     'Title-uniform-see',
1237                     'Title-uniform-seealso',
1238                     'totalissues',
1239                     'yr',
1240
1241                     # items indexes
1242                     'acqsource',
1243                     'barcode',
1244                     'bc',
1245                     'branch',
1246                     'ccode',
1247                     'classification-source',
1248                     'cn-sort',
1249                     'coded-location-qualifier',
1250                     'copynumber',
1251                     'damaged',
1252                     'datelastborrowed',
1253                     'datelastseen',
1254                     'holdingbranch',
1255                     'homebranch',
1256                     'issues',
1257                     'item',
1258                     'itemnumber',
1259                     'itype',
1260                     'Local-classification',
1261                     'location',
1262                     'lost',
1263                     'materials-specified',
1264                     'mc-ccode',
1265                     'mc-itype',
1266                     'mc-loc',
1267                     'notforloan',
1268                     'Number-local-acquisition',
1269                     'onloan',
1270                     'price',
1271                     'renewals',
1272                     'replacementprice',
1273                     'replacementpricedate',
1274                     'reserves',
1275                     'restricted',
1276                     'stack',
1277                     'stocknumber',
1278                     'inv',
1279                     'uri',
1280                     'withdrawn',
1281
1282                     # subject related
1283                   );
1284
1285     return \@indexes;
1286 }
1287
1288 =head2 _handle_exploding_index
1289
1290     my $query = _handle_exploding_index($index, $term)
1291
1292 Callback routine to generate the search for "exploding" indexes (i.e.
1293 those indexes which are turned into multiple or-connected searches based
1294 on authority data).
1295
1296 =cut
1297
1298 sub _handle_exploding_index {
1299     my ($QParser, $filter, $params, $negate, $server) = @_;
1300     my $index = $filter;
1301     my $term = join(' ', @$params);
1302
1303     return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1304
1305     my $marcflavour = C4::Context->preference('marcflavour');
1306
1307     my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1308     my $wantedcodes = '';
1309     my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1310     my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1311     foreach my $auth (@$results) {
1312         my $record = MARC::Record->new_from_usmarc($auth);
1313         my @references = $record->field('5..');
1314         if (@references) {
1315             if ($index eq 'su-br') {
1316                 $wantedcodes = 'g';
1317             } elsif ($index eq 'su-na') {
1318                 $wantedcodes = 'h';
1319             } elsif ($index eq 'su-rl') {
1320                 $wantedcodes = '';
1321             }
1322             foreach my $reference (@references) {
1323                 my $codes = $reference->subfield($codesubfield);
1324                 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1325             }
1326         }
1327     }
1328     my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1329     return $query;
1330 }
1331
1332 =head2 parseQuery
1333
1334     ( $operators, $operands, $indexes, $limits,
1335       $sort_by, $scan, $lang ) =
1336             buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1337
1338 Shim function to ease the transition from buildQuery to a new QueryParser.
1339 This function is called at the beginning of buildQuery, and modifies
1340 buildQuery's input. If it can handle the input, it returns a query that
1341 buildQuery will not try to parse.
1342 =cut
1343
1344 sub parseQuery {
1345     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1346
1347     my @operators = $operators ? @$operators : ();
1348     my @indexes   = $indexes   ? @$indexes   : ();
1349     my @operands  = $operands  ? @$operands  : ();
1350     my @limits    = $limits    ? @$limits    : ();
1351     my @sort_by   = $sort_by   ? @$sort_by   : ();
1352
1353     my $query = $operands[0];
1354     my $index;
1355     my $term;
1356     my $query_desc;
1357
1358     my $QParser;
1359     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1360     undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1361     undef $QParser if (scalar @limits > 0);
1362
1363     if ($QParser)
1364     {
1365         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1366         $query = '';
1367         for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1368             next unless $operands[$ii];
1369             $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1370               if ($query);
1371             if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1372                 $query .= $operands[$ii];
1373             }
1374             elsif ( $indexes[$ii] =~ m/su-/ ) {
1375                 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1376             }
1377             else {
1378                 $query .=
1379                   ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1380             }
1381         }
1382         foreach my $limit (@limits) {
1383         }
1384         if ( scalar(@sort_by) > 0 ) {
1385             my $modifier_re =
1386               '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1387             $query =~ s/$modifier_re//g;
1388             foreach my $modifier (@sort_by) {
1389                 $query .= " #$modifier";
1390             }
1391         }
1392
1393         $query_desc = $query;
1394         $query_desc =~ s/\s+/ /g;
1395         if ( C4::Context->preference("QueryWeightFields") ) {
1396         }
1397         $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1398               { 'target_syntax_callback' => \&_handle_exploding_index } );
1399         $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1400               { 'target_syntax_callback' => \&_handle_exploding_index } );
1401         $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1402               { 'target_syntax_callback' => \&_handle_exploding_index } );
1403         $QParser->parse($query);
1404         $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1405     }
1406     else {
1407         require Koha::QueryParser::Driver::PQF;
1408         my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1409         s/$modifier_re//g for @operands;
1410     }
1411
1412     return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1413 }
1414
1415 =head2 buildQuery
1416
1417 ( $error, $query,
1418 $simple_query, $query_cgi,
1419 $query_desc, $limit,
1420 $limit_cgi, $limit_desc,
1421 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1422
1423 Build queries and limits in CCL, CGI, Human,
1424 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1425
1426 See verbose embedded documentation.
1427
1428
1429 =cut
1430
1431 sub buildQuery {
1432     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1433
1434     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1435
1436     my $query_desc;
1437     ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1438
1439     # dereference
1440     my @operators = $operators ? @$operators : ();
1441     my @indexes   = $indexes   ? @$indexes   : ();
1442     my @operands  = $operands  ? @$operands  : ();
1443     my @limits    = $limits    ? @$limits    : ();
1444     my @sort_by   = $sort_by   ? @$sort_by   : ();
1445
1446     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1447     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1448     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1449     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1450     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1451
1452     my $query        = $operands[0];
1453     my $simple_query = $operands[0];
1454
1455     # initialize the variables we're passing back
1456     my $query_cgi;
1457     my $query_type;
1458
1459     my $limit;
1460     my $limit_cgi;
1461     my $limit_desc;
1462
1463     my $stopwords_removed;    # flag to determine if stopwords have been removed
1464
1465     my $cclq       = 0;
1466     my $cclindexes = getIndexes();
1467     if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1468         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1469             my $dx = lc($1);
1470             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1471         }
1472         $query = "ccl=$query" if $cclq;
1473     }
1474
1475 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1476 # DIAGNOSTIC ONLY!!
1477     if ( $query =~ /^ccl=/ ) {
1478         my $q=$';
1479         # This is needed otherwise ccl= and &limit won't work together, and
1480         # this happens when selecting a subject on the opac-detail page
1481         @limits = grep {!/^$/} @limits;
1482         if ( @limits ) {
1483             $q .= ' and '.join(' and ', @limits);
1484         }
1485         return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $q, '', '', '', '', 'ccl' );
1486     }
1487     if ( $query =~ /^cql=/ ) {
1488         return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', '', 'cql' );
1489     }
1490     if ( $query =~ /^pqf=/ ) {
1491         if ($query_desc) {
1492             $query_cgi = "q=".uri_escape_utf8($query_desc);
1493         } else {
1494             $query_desc = $';
1495             $query_cgi = "q=pqf=".uri_escape_utf8($');
1496         }
1497         return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1498     }
1499
1500     # pass nested queries directly
1501     # FIXME: need better handling of some of these variables in this case
1502     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1503     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1504 #    if ( $query =~ /(\(|\))/ ) {
1505 #        return (
1506 #            undef,              $query, $simple_query, $query_cgi,
1507 #            $query,             $limit, $limit_cgi,    $limit_desc,
1508 #            $stopwords_removed, 'ccl'
1509 #        );
1510 #    }
1511
1512 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1513 # query operands and indexes and add stemming, truncation, field weighting, etc.
1514 # Once we do so, we'll end up with a value in $query, just like if we had an
1515 # incoming $query from the user
1516     else {
1517         $query = ""
1518           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1519         my $previous_operand
1520           ;    # a flag used to keep track if there was a previous query
1521                # if there was, we can apply the current operator
1522                # for every operand
1523         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1524
1525             # COMBINE OPERANDS, INDEXES AND OPERATORS
1526             if ( $operands[$i] ) {
1527                 $operands[$i]=~s/^\s+//;
1528
1529               # A flag to determine whether or not to add the index to the query
1530                 my $indexes_set;
1531
1532 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1533                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1534                     $weight_fields    = 0;
1535                     $stemming         = 0;
1536                     $remove_stopwords = 0;
1537                 } else {
1538                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1539                 }
1540                 my $operand = $operands[$i];
1541                 my $index   = $indexes[$i];
1542
1543                 # Add index-specific attributes
1544
1545                 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1546                 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1547                 #which is processed higher up in this sub. Other than that, year searches are typically
1548                 #handled as limits which are not processed her either.
1549
1550                 # Date of Publication
1551                 if ( $index =~ /yr/ ) {
1552                     #weight_fields/relevance search causes errors with date ranges
1553                     #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1554                     #In the case of YYYY-YYYY, it will return no results
1555                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1556                 }
1557
1558                 # Date of Acquisition
1559                 elsif ( $index =~ /acqdate/ ) {
1560                     #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1561                     #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1562                       #top of the results just because they have lots of item records matching that date.
1563                     #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1564                       #irrelevant here
1565                     #remove_stopwords doesn't function anymore so is irrelevant
1566                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1567                 }
1568                 # ISBN,ISSN,Standard Number, don't need special treatment
1569                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1570                     (
1571                         $stemming,      $auto_truncation,
1572                         $weight_fields, $fuzzy_enabled,
1573                         $remove_stopwords
1574                     ) = ( 0, 0, 0, 0, 0 );
1575
1576                     if ( $index eq 'nb' ) {
1577                         if ( C4::Context->preference("SearchWithISBNVariations") ) {
1578                             my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1579                             $operands[$i] = $operand =  '(nb=' . join(' OR nb=', @isbns) . ')';
1580                             $indexes[$i] = $index = '';
1581                         }
1582                     }
1583                 }
1584
1585                 if(not $index){
1586                     $index = 'kw';
1587                 }
1588
1589                 # Set default structure attribute (word list)
1590                 my $struct_attr = q{};
1591                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1592                     $struct_attr = ",wrdl";
1593                 }
1594
1595                 # Some helpful index variants
1596                 my $index_plus       = $index . $struct_attr . ':';
1597                 my $index_plus_comma = $index . $struct_attr . ',';
1598
1599                 # Remove Stopwords
1600                 if ($remove_stopwords) {
1601                     ( $operand, $stopwords_removed ) =
1602                       _remove_stopwords( $operand, $index );
1603                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1604                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1605                       if ( $stopwords_removed && $DEBUG );
1606                 }
1607
1608                 if ($auto_truncation){
1609                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1610                                                 #FIXME only valid with LTR scripts
1611                                                 $operand=join(" ",map{
1612                                                                                         (index($_,"*")>0?"$_":"$_*")
1613                                                                                          }split (/\s+/,$operand));
1614                                                 warn $operand if $DEBUG;
1615                                         }
1616                                 }
1617
1618                 # Detect Truncation
1619                 my $truncated_operand;
1620                 my( $nontruncated, $righttruncated, $lefttruncated,
1621                     $rightlefttruncated, $regexpr
1622                 ) = _detect_truncation( $operand, $index );
1623                 warn
1624 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1625                   if $DEBUG;
1626
1627                 # Apply Truncation
1628                 if (
1629                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1630                     scalar(@$rightlefttruncated) > 0 )
1631                 {
1632
1633                # Don't field weight or add the index to the query, we do it here
1634                     $indexes_set = 1;
1635                     undef $weight_fields;
1636                     my $previous_truncation_operand;
1637                     if (scalar @$nontruncated) {
1638                         $truncated_operand .= "$index_plus @$nontruncated ";
1639                         $previous_truncation_operand = 1;
1640                     }
1641                     if (scalar @$righttruncated) {
1642                         $truncated_operand .= "and " if $previous_truncation_operand;
1643                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1644                         $previous_truncation_operand = 1;
1645                     }
1646                     if (scalar @$lefttruncated) {
1647                         $truncated_operand .= "and " if $previous_truncation_operand;
1648                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1649                         $previous_truncation_operand = 1;
1650                     }
1651                     if (scalar @$rightlefttruncated) {
1652                         $truncated_operand .= "and " if $previous_truncation_operand;
1653                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1654                         $previous_truncation_operand = 1;
1655                     }
1656                 }
1657                 $operand = $truncated_operand if $truncated_operand;
1658                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1659
1660                 # Handle Stemming
1661                 my $stemmed_operand;
1662                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1663                                                                                 if $stemming;
1664
1665                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1666
1667                 # Handle Field Weighting
1668                 my $weighted_operand;
1669                 if ($weight_fields) {
1670                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1671                     $operand = $weighted_operand;
1672                     $indexes_set = 1;
1673                 }
1674
1675                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1676
1677                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1678                     query => $query,
1679                     query_cgi => $query_cgi,
1680                     query_desc => $query_desc,
1681                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1682                     parsed_operand => $operand,
1683                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1684                     index => $index,
1685                     index_plus => $index_plus,
1686                     indexes_set => $indexes_set,
1687                     previous_operand => $previous_operand,
1688                 });
1689
1690             }    #/if $operands
1691         }    # /for
1692     }
1693     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1694
1695     # add limits
1696     my %group_OR_limits;
1697     my $availability_limit;
1698     foreach my $this_limit (@limits) {
1699         next unless $this_limit;
1700         if ( $this_limit =~ /available/ ) {
1701 #
1702 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1703 ## In English:
1704 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1705             $availability_limit .=
1706 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1707             $limit_cgi  .= "&limit=available";
1708             $limit_desc .= "";
1709         }
1710
1711         # group_OR_limits, prefixed by mc-
1712         # OR every member of the group
1713         elsif ( $this_limit =~ /mc/ ) {
1714             my ($k,$v) = split(/:/, $this_limit,2);
1715             if ( $k !~ /mc-i(tem)?type/ ) {
1716                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1717                 $this_limit =~ tr/"//d;
1718                 $this_limit = $k.":'".$v."'";
1719             }
1720
1721             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1722             $limit_desc      .= " or " if $group_OR_limits{$k};
1723             $group_OR_limits{$k} .= "$this_limit";
1724             $limit_cgi       .= "&limit=" . uri_escape_utf8($this_limit);
1725             $limit_desc      .= " $this_limit";
1726         }
1727
1728         # Regular old limits
1729         else {
1730             $limit .= " and " if $limit || $query;
1731             $limit      .= "$this_limit";
1732             $limit_cgi  .= "&limit=" . uri_escape_utf8($this_limit);
1733             if ($this_limit =~ /^branch:(.+)/) {
1734                 my $branchcode = $1;
1735                 my $branchname = GetBranchName($branchcode);
1736                 if (defined $branchname) {
1737                     $limit_desc .= " branch:$branchname";
1738                 } else {
1739                     $limit_desc .= " $this_limit";
1740                 }
1741             } else {
1742                 $limit_desc .= " $this_limit";
1743             }
1744         }
1745     }
1746     foreach my $k (keys (%group_OR_limits)) {
1747         $limit .= " and " if ( $query || $limit );
1748         $limit .= "($group_OR_limits{$k})";
1749     }
1750     if ($availability_limit) {
1751         $limit .= " and " if ( $query || $limit );
1752         $limit .= "($availability_limit)";
1753     }
1754
1755     # Normalize the query and limit strings
1756     # This is flawed , means we can't search anything with : in it
1757     # if user wants to do ccl or cql, start the query with that
1758 #    $query =~ s/:/=/g;
1759     #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1760     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1761     $query =~ s/(?<=(wrdl)):/=/g;
1762     $query =~ s/(?<=(trn|phr)):/=/g;
1763     $query =~ s/(?<=(st-numeric)):/=/g;
1764     $query =~ s/(?<=(st-year)):/=/g;
1765     $query =~ s/(?<=(st-date-normalized)):/=/g;
1766     $limit =~ s/:/=/g;
1767     for ( $query, $query_desc, $limit, $limit_desc ) {
1768         s/  +/ /g;    # remove extra spaces
1769         s/^ //g;     # remove any beginning spaces
1770         s/ $//g;     # remove any ending spaces
1771         s/==/=/g;    # remove double == from query
1772     }
1773     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1774
1775     for ($query_cgi,$simple_query) {
1776         s/"//g;
1777     }
1778     # append the limit to the query
1779     $query .= " " . $limit;
1780
1781     # Warnings if DEBUG
1782     if ($DEBUG) {
1783         warn "QUERY:" . $query;
1784         warn "QUERY CGI:" . $query_cgi;
1785         warn "QUERY DESC:" . $query_desc;
1786         warn "LIMIT:" . $limit;
1787         warn "LIMIT CGI:" . $limit_cgi;
1788         warn "LIMIT DESC:" . $limit_desc;
1789         warn "---------\nLeave buildQuery\n---------";
1790     }
1791     return (
1792         undef,              $query, $simple_query, $query_cgi,
1793         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1794         $stopwords_removed, $query_type
1795     );
1796 }
1797
1798 =head2 _build_initial_query
1799
1800   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1801
1802   Build a section of the initial query containing indexes, operators, and operands.
1803
1804 =cut
1805
1806 sub _build_initial_query {
1807     my ($params) = @_;
1808
1809     my $operator = "";
1810     if ($params->{previous_operand}){
1811         #If there is a previous operand, add a supplied operator or the default 'and'
1812         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1813     }
1814
1815     #NOTE: indexes_set is typically set when doing truncation or field weighting
1816     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1817
1818     #e.g. "kw,wrdl:test"
1819     #e.g. " and kw,wrdl:test"
1820     $params->{query} .= $operator . $operand;
1821
1822     $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1823     $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1824     $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1825
1826     #e.g. " and kw,wrdl: test"
1827     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1828
1829     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1830
1831     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1832 }
1833
1834 =head2 searchResults
1835
1836   my @search_results = searchResults($search_context, $searchdesc, $hits,
1837                                      $results_per_page, $offset, $scan,
1838                                      @marcresults);
1839
1840 Format results in a form suitable for passing to the template
1841
1842 =cut
1843
1844 # IMO this subroutine is pretty messy still -- it's responsible for
1845 # building the HTML output for the template
1846 sub searchResults {
1847     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1848     my $dbh = C4::Context->dbh;
1849     my @newresults;
1850
1851     require C4::Items;
1852
1853     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1854     my ($is_opac, $hidelostitems);
1855     if ($search_context eq 'opac') {
1856         $hidelostitems = C4::Context->preference('hidelostitems');
1857         $is_opac       = 1;
1858     }
1859
1860     #Build branchnames hash
1861     #find branchname
1862     #get branch information.....
1863     my %branches;
1864     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1865     $bsth->execute();
1866     while ( my $bdata = $bsth->fetchrow_hashref ) {
1867         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1868     }
1869 # FIXME - We build an authorised values hash here, using the default framework
1870 # though it is possible to have different authvals for different fws.
1871
1872     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1873
1874     # get notforloan authorised value list (see $shelflocations  FIXME)
1875     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1876
1877     #Get itemtype hash
1878     my %itemtypes = %{ GetItemTypes() };
1879
1880     #search item field code
1881     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1882
1883     ## find column names of items related to MARC
1884     my %subfieldstosearch;
1885     my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1886     for my $column ( @columns ) {
1887         my ( $tagfield, $tagsubfield ) =
1888           &GetMarcFromKohaField( "items." . $column, "" );
1889         if ( defined $tagsubfield ) {
1890             $subfieldstosearch{$column} = $tagsubfield;
1891         }
1892     }
1893
1894     # handle which records to actually retrieve
1895     my $times;
1896     if ( $hits && $offset + $results_per_page <= $hits ) {
1897         $times = $offset + $results_per_page;
1898     }
1899     else {
1900         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1901     }
1902
1903     my $marcflavour = C4::Context->preference("marcflavour");
1904     # We get the biblionumber position in MARC
1905     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1906
1907     # loop through all of the records we've retrieved
1908     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1909
1910         my $marcrecord;
1911         if ($scan) {
1912             # For Scan searches we built USMARC data
1913             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1914         } else {
1915             # Normal search, render from Zebra's output
1916             $marcrecord = new_record_from_zebra(
1917                 'biblioserver',
1918                 $marcresults->[$i]
1919             );
1920
1921             if ( ! defined $marcrecord ) {
1922                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1923                 next;
1924             }
1925         }
1926
1927         my $fw = $scan
1928              ? undef
1929              : $bibliotag < 10
1930                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1931                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1932
1933         SetUTF8Flag($marcrecord);
1934         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1935         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1936         $oldbiblio->{result_number} = $i + 1;
1937
1938         # add imageurl to itemtype if there is one
1939         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1940
1941         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1942                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1943                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1944                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1945                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1946                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1947
1948                 # edition information, if any
1949         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1950         $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1951  # Build summary if there is one (the summary is defined in the itemtypes table)
1952  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1953         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1954             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1955             my @fields  = $marcrecord->fields();
1956
1957             my $newsummary;
1958             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1959                 my $tags = {};
1960                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1961                     $tag =~ /(.{3})(.)/;
1962                     if($marcrecord->field($1)){
1963                         my @abc = $marcrecord->field($1)->subfield($2);
1964                         $tags->{$tag} = $#abc + 1 ;
1965                     }
1966                 }
1967
1968                 # We catch how many times to repeat this line
1969                 my $max = 0;
1970                 foreach my $tag (keys(%$tags)){
1971                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1972                  }
1973
1974                 # we replace, and repeat each line
1975                 for (my $i = 0 ; $i < $max ; $i++){
1976                     my $newline = $line;
1977
1978                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1979                         $tag =~ /(.{3})(.)/;
1980
1981                         if($marcrecord->field($1)){
1982                             my @repl = $marcrecord->field($1)->subfield($2);
1983                             my $subfieldvalue = $repl[$i];
1984                             $newline =~ s/\[$tag\]/$subfieldvalue/g;
1985                         }
1986                     }
1987                     $newsummary .= "$newline\n";
1988                 }
1989             }
1990
1991             $newsummary =~ s/\[(.*?)]//g;
1992             $newsummary =~ s/\n/<br\/>/g;
1993             $oldbiblio->{summary} = $newsummary;
1994         }
1995
1996         # Pull out the items fields
1997         my @fields = $marcrecord->field($itemtag);
1998         my $marcflavor = C4::Context->preference("marcflavour");
1999         # adding linked items that belong to host records
2000         my $analyticsfield = '773';
2001         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2002             $analyticsfield = '773';
2003         } elsif ($marcflavor eq 'UNIMARC') {
2004             $analyticsfield = '461';
2005         }
2006         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2007             my $hostbiblionumber = $hostfield->subfield("0");
2008             my $linkeditemnumber = $hostfield->subfield("9");
2009             if(!$hostbiblionumber eq undef){
2010                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2011                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2012                 if(!$hostbiblio eq undef){
2013                     my @hostitems = $hostbiblio->field($itemfield);
2014                     foreach my $hostitem (@hostitems){
2015                         if ($hostitem->subfield("9") eq $linkeditemnumber){
2016                             my $linkeditem =$hostitem;
2017                             # append linked items if they exist
2018                             if (!$linkeditem eq undef){
2019                                 push (@fields, $linkeditem);}
2020                         }
2021                     }
2022                 }
2023             }
2024         }
2025
2026         # Setting item statuses for display
2027         my @available_items_loop;
2028         my @onloan_items_loop;
2029         my @other_items_loop;
2030
2031         my $available_items;
2032         my $onloan_items;
2033         my $other_items;
2034
2035         my $ordered_count         = 0;
2036         my $available_count       = 0;
2037         my $onloan_count          = 0;
2038         my $longoverdue_count     = 0;
2039         my $other_count           = 0;
2040         my $withdrawn_count        = 0;
2041         my $itemlost_count        = 0;
2042         my $hideatopac_count      = 0;
2043         my $itembinding_count     = 0;
2044         my $itemdamaged_count     = 0;
2045         my $item_in_transit_count = 0;
2046         my $can_place_holds       = 0;
2047         my $item_onhold_count     = 0;
2048         my $notforloan_count      = 0;
2049         my $items_count           = scalar(@fields);
2050         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2051         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2052         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2053
2054         # loop through every item
2055         foreach my $field (@fields) {
2056             my $item;
2057
2058             # populate the items hash
2059             foreach my $code ( keys %subfieldstosearch ) {
2060                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2061             }
2062             $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2063
2064                 # OPAC hidden items
2065             if ($is_opac) {
2066                 # hidden because lost
2067                 if ($hidelostitems && $item->{itemlost}) {
2068                     $hideatopac_count++;
2069                     next;
2070                 }
2071                 # hidden based on OpacHiddenItems syspref
2072                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2073                 if (scalar @hi) {
2074                     push @hiddenitems, @hi;
2075                     $hideatopac_count++;
2076                     next;
2077                 }
2078             }
2079
2080             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2081             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2082
2083             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2084             if ($item->{$hbranch}) {
2085                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2086             }
2087             elsif ($item->{$otherbranch}) {     # Last resort
2088                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2089             }
2090
2091                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2092 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2093             my $userenv = C4::Context->userenv;
2094             if ( $item->{onloan}
2095                 && !( C4::Members::GetHideLostItemsPreference( $userenv->{'number'} ) && $item->{itemlost} ) )
2096             {
2097                 $onloan_count++;
2098                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2099                 $onloan_items->{$key}->{due_date} = output_pref( { dt => dt_from_string( $item->{onloan} ), dateonly => 1 } );
2100                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2101                 $onloan_items->{$key}->{branchname}     = $item->{branchname};
2102                 $onloan_items->{$key}->{location}       = $shelflocations->{ $item->{location} };
2103                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2104                 $onloan_items->{$key}->{description}    = $item->{description};
2105                 $onloan_items->{$key}->{imageurl} =
2106                   getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2107
2108                 # if something's checked out and lost, mark it as 'long overdue'
2109                 if ( $item->{itemlost} ) {
2110                     $onloan_items->{$key}->{longoverdue}++;
2111                     $longoverdue_count++;
2112                 }
2113                 else {    # can place holds as long as item isn't lost
2114                     $can_place_holds = 1;
2115                 }
2116             }
2117
2118          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2119             else {
2120
2121                 $item->{notforloan}=1 if !$item->{notforloan}  && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2122
2123                 # item is on order
2124                 if ( $item->{notforloan} < 0 ) {
2125                     $ordered_count++;
2126                 } elsif ( $item->{notforloan} > 0 ) {
2127                     $notforloan_count++;
2128                 }
2129
2130                 # is item in transit?
2131                 my $transfertwhen = '';
2132                 my ($transfertfrom, $transfertto);
2133
2134                 # is item on the reserve shelf?
2135                 my $reservestatus = '';
2136
2137                 unless ($item->{withdrawn}
2138                         || $item->{itemlost}
2139                         || $item->{damaged}
2140                         || $item->{notforloan}
2141                         || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck')
2142                         && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) {
2143
2144                     # A couple heuristics to limit how many times
2145                     # we query the database for item transfer information, sacrificing
2146                     # accuracy in some cases for speed;
2147                     #
2148                     # 1. don't query if item has one of the other statuses
2149                     # 2. don't check transit status if the bib has
2150                     #    more than 20 items
2151                     #
2152                     # FIXME: to avoid having the query the database like this, and to make
2153                     #        the in transit status count as unavailable for search limiting,
2154                     #        should map transit status to record indexed in Zebra.
2155                     #
2156                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2157                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2158                 }
2159
2160                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2161                 if (   $item->{withdrawn}
2162                     || $item->{itemlost}
2163                     || $item->{damaged}
2164                     || $item->{notforloan}
2165                     || $reservestatus eq 'Waiting'
2166                     || ($transfertwhen ne ''))
2167                 {
2168                     $withdrawn_count++        if $item->{withdrawn};
2169                     $itemlost_count++        if $item->{itemlost};
2170                     $itemdamaged_count++     if $item->{damaged};
2171                     $item_in_transit_count++ if $transfertwhen ne '';
2172                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2173                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2174
2175                     # can place a hold on a item if
2176                     # not lost nor withdrawn
2177                     # not damaged unless AllowHoldsOnDamagedItems is true
2178                     # item is either for loan or on order (notforloan < 0)
2179                     $can_place_holds = 1
2180                       if (
2181                            !$item->{itemlost}
2182                         && !$item->{withdrawn}
2183                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2184                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2185                       );
2186
2187                     $other_count++;
2188
2189                     my $key = $prefix . $item->{status};
2190                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2191                         $other_items->{$key}->{$_} = $item->{$_};
2192                     }
2193                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2194                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2195                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2196                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2197                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2198                                         $other_items->{$key}->{description} = $item->{description};
2199                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2200                 }
2201                 # item is available
2202                 else {
2203                     $can_place_holds = 1;
2204                     $available_count++;
2205                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2206                                         foreach (qw(branchname itemcallnumber description)) {
2207                         $available_items->{$prefix}->{$_} = $item->{$_};
2208                                         }
2209                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2210                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2211                 }
2212             }
2213         }    # notforloan, item level and biblioitem level
2214
2215         # if all items are hidden, do not show the record
2216         if ($items_count > 0 && $hideatopac_count == $items_count) {
2217             next;
2218         }
2219
2220         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2221         for my $key ( sort keys %$onloan_items ) {
2222             (++$onloanitemscount > $maxitems) and last;
2223             push @onloan_items_loop, $onloan_items->{$key};
2224         }
2225         for my $key ( sort keys %$other_items ) {
2226             (++$otheritemscount > $maxitems) and last;
2227             push @other_items_loop, $other_items->{$key};
2228         }
2229         for my $key ( sort keys %$available_items ) {
2230             (++$availableitemscount > $maxitems) and last;
2231             push @available_items_loop, $available_items->{$key}
2232         }
2233
2234         # XSLT processing of some stuff
2235         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2236         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2237             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2238         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2239         }
2240
2241         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2242         if (!C4::Context->preference("item-level_itypes")) {
2243             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2244                 $can_place_holds = 0;
2245             }
2246         }
2247         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2248         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2249         $oldbiblio->{items_count}          = $items_count;
2250         $oldbiblio->{available_items_loop} = \@available_items_loop;
2251         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2252         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2253         $oldbiblio->{availablecount}       = $available_count;
2254         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2255         $oldbiblio->{onloancount}          = $onloan_count;
2256         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2257         $oldbiblio->{othercount}           = $other_count;
2258         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2259         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2260         $oldbiblio->{itemlostcount}        = $itemlost_count;
2261         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2262         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2263         $oldbiblio->{onholdcount}          = $item_onhold_count;
2264         $oldbiblio->{orderedcount}         = $ordered_count;
2265         $oldbiblio->{notforloancount}      = $notforloan_count;
2266
2267         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2268             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2269             my $subfields = substr $fieldspec, 3;
2270             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2271             my @alternateholdingsinfo = ();
2272             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2273             my $alternateholdingscount = 0;
2274
2275             for my $field (@holdingsfields) {
2276                 my %holding = ( holding => '' );
2277                 my $havesubfield = 0;
2278                 for my $subfield ($field->subfields()) {
2279                     if ((index $subfields, $$subfield[0]) >= 0) {
2280                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2281                         $holding{'holding'} .= $$subfield[1];
2282                         $havesubfield++;
2283                     }
2284                 }
2285                 if ($havesubfield) {
2286                     push(@alternateholdingsinfo, \%holding);
2287                     $alternateholdingscount++;
2288                 }
2289             }
2290
2291             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2292             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2293         }
2294
2295         push( @newresults, $oldbiblio );
2296     }
2297
2298     return @newresults;
2299 }
2300
2301 =head2 SearchAcquisitions
2302     Search for acquisitions
2303 =cut
2304
2305 sub SearchAcquisitions{
2306     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2307
2308     my $dbh=C4::Context->dbh;
2309     # Variable initialization
2310     my $str=qq|
2311     SELECT marcxml
2312     FROM biblio
2313     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2314     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2315     WHERE dateaccessioned BETWEEN ? AND ?
2316     |;
2317
2318     my (@params,@loopcriteria);
2319
2320     push @params, $datebegin->output("iso");
2321     push @params, $dateend->output("iso");
2322
2323     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2324         if(C4::Context->preference("item-level_itypes")){
2325             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2326         }else{
2327             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2328         }
2329         push @params, @$itemtypes;
2330     }
2331
2332     if ($criteria =~/itemtype/){
2333         if(C4::Context->preference("item-level_itypes")){
2334             $str .= "AND items.itype=? ";
2335         }else{
2336             $str .= "AND biblioitems.itemtype=? ";
2337         }
2338
2339         if(scalar(@$itemtypes) == 0){
2340             my $itypes = GetItemTypes();
2341             for my $key (keys %$itypes){
2342                 push @$itemtypes, $key;
2343             }
2344         }
2345
2346         @loopcriteria= @$itemtypes;
2347     }elsif ($criteria=~/itemcallnumber/){
2348         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2349                  OR items.itemcallnumber is NULL
2350                  OR items.itemcallnumber = '')";
2351
2352         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2353     }else {
2354         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2355         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2356     }
2357
2358     if ($orderby =~ /date_desc/){
2359         $str.=" ORDER BY dateaccessioned DESC";
2360     } else {
2361         $str.=" ORDER BY title";
2362     }
2363
2364     my $qdataacquisitions=$dbh->prepare($str);
2365
2366     my @loopacquisitions;
2367     foreach my $value(@loopcriteria){
2368         push @params,$value;
2369         my %cell;
2370         $cell{"title"}=$value;
2371         $cell{"titlecode"}=$value;
2372
2373         eval{$qdataacquisitions->execute(@params);};
2374
2375         if ($@){ warn "recentacquisitions Error :$@";}
2376         else {
2377             my @loopdata;
2378             while (my $data=$qdataacquisitions->fetchrow_hashref){
2379                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2380             }
2381             $cell{"loopdata"}=\@loopdata;
2382         }
2383         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2384         pop @params;
2385     }
2386     $qdataacquisitions->finish;
2387     return \@loopacquisitions;
2388 }
2389
2390 =head2 enabled_staff_search_views
2391
2392 %hash = enabled_staff_search_views()
2393
2394 This function returns a hash that contains three flags obtained from the system
2395 preferences, used to determine whether a particular staff search results view
2396 is enabled.
2397
2398 =over 2
2399
2400 =item C<Output arg:>
2401
2402     * $hash{can_view_MARC} is true only if the MARC view is enabled
2403     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2404     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2405
2406 =item C<usage in the script:>
2407
2408 =back
2409
2410 $template->param ( C4::Search::enabled_staff_search_views );
2411
2412 =cut
2413
2414 sub enabled_staff_search_views
2415 {
2416         return (
2417                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2418                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2419                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2420         );
2421 }
2422
2423 =head2 z3950_search_args
2424
2425 $arrayref = z3950_search_args($matchpoints)
2426
2427 This function returns an array reference that contains the search parameters to be
2428 passed to the Z39.50 search script (z3950_search.pl). The array elements
2429 are hash refs whose keys are name and value, and whose values are the
2430 name of a search parameter, the value of that search parameter and the URL encoded
2431 value of that parameter.
2432
2433 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2434
2435 The search parameter values are obtained from the bibliographic record whose
2436 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2437
2438 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2439 a general purpose search argument. In this case, the returned array contains only
2440 entry: the key is 'title' and the value is derived from $matchpoints.
2441
2442 If a search parameter value is undefined or empty, it is not included in the returned
2443 array.
2444
2445 The returned array reference may be passed directly to the template parameters.
2446
2447 =over 2
2448
2449 =item C<Output arg:>
2450
2451     * $array containing hash refs as described above
2452
2453 =item C<usage in the script:>
2454
2455 =back
2456
2457 $data = Biblio::GetBiblioData($bibno);
2458 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2459
2460 *OR*
2461
2462 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2463
2464 =cut
2465
2466 sub z3950_search_args {
2467     my $bibrec = shift;
2468
2469     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2470     my $isbn = Business::ISBN->new( $isbn_string );
2471
2472     if (defined $isbn && $isbn->is_valid)
2473     {
2474         if ( ref($bibrec) ) {
2475             $bibrec->{isbn} = $isbn_string;
2476             $bibrec->{title} = undef;
2477         } else {
2478             $bibrec = { isbn => $isbn_string };
2479         }
2480     }
2481     else {
2482         $bibrec = { title => $bibrec } if !ref $bibrec;
2483     }
2484     my $array = [];
2485     for my $field (qw/ lccn isbn issn title author dewey subject /)
2486     {
2487         push @$array, { name => $field, value => $bibrec->{$field} }
2488           if defined $bibrec->{$field};
2489     }
2490     return $array;
2491 }
2492
2493 =head2 GetDistinctValues($field);
2494
2495 C<$field> is a reference to the fields array
2496
2497 =cut
2498
2499 sub GetDistinctValues {
2500     my ($fieldname,$string)=@_;
2501     # returns a reference to a hash of references to branches...
2502     if ($fieldname=~/\./){
2503                         my ($table,$column)=split /\./, $fieldname;
2504                         my $dbh = C4::Context->dbh;
2505                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2506                         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 ");
2507                         $sth->execute;
2508                         my $elements=$sth->fetchall_arrayref({});
2509                         return $elements;
2510    }
2511    else {
2512                 $string||= qq("");
2513                 my @servers=qw<biblioserver authorityserver>;
2514                 my (@zconns,@results);
2515         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2516                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2517                         $results[$i] =
2518                       $zconns[$i]->scan(
2519                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2520                       );
2521                 }
2522                 # The big moment: asynchronously retrieve results from all servers
2523                 my @elements;
2524         _ZOOM_event_loop(
2525             \@zconns,
2526             \@results,
2527             sub {
2528                 my ( $i, $size ) = @_;
2529                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2530                     my %hashscan;
2531                     @hashscan{qw(value cnt)} =
2532                       $results[ $i - 1 ]->display_term($j);
2533                     push @elements, \%hashscan;
2534                 }
2535             }
2536         );
2537                 return \@elements;
2538    }
2539 }
2540
2541 =head2 _ZOOM_event_loop
2542
2543     _ZOOM_event_loop(\@zconns, \@results, sub {
2544         my ( $i, $size ) = @_;
2545         ....
2546     } );
2547
2548 Processes a ZOOM event loop and passes control to a closure for
2549 processing the results, and destroying the resultsets.
2550
2551 =cut
2552
2553 sub _ZOOM_event_loop {
2554     my ($zconns, $results, $callback) = @_;
2555     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2556         my $ev = $zconns->[ $i - 1 ]->last_event();
2557         if ( $ev == ZOOM::Event::ZEND ) {
2558             next unless $results->[ $i - 1 ];
2559             my $size = $results->[ $i - 1 ]->size();
2560             if ( $size > 0 ) {
2561                 $callback->($i, $size);
2562             }
2563         }
2564     }
2565
2566     foreach my $result (@$results) {
2567         $result->destroy();
2568     }
2569 }
2570
2571 =head2 new_record_from_zebra
2572
2573 Given raw data from a Zebra result set, return a MARC::Record object
2574
2575 This helper function is needed to take into account all the involved
2576 system preferences and configuration variables to properly create the
2577 MARC::Record object.
2578
2579 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2580 data. If we are using DOM, then it has to be MARCXML.
2581
2582 =cut
2583
2584 sub new_record_from_zebra {
2585
2586     my $server   = shift;
2587     my $raw_data = shift;
2588     # Set the default indexing modes
2589     my $index_mode = ( $server eq 'biblioserver' )
2590                         ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2591                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2592
2593     my $marc_record =  eval {
2594         if ( $index_mode eq 'dom' ) {
2595             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2596         } else {
2597             MARC::Record->new_from_usmarc( $raw_data );
2598         }
2599     };
2600
2601     if ($@) {
2602         return;
2603     } else {
2604         return $marc_record;
2605     }
2606
2607 }
2608
2609 END { }    # module clean-up code here (global destructor)
2610
2611 1;
2612 __END__
2613
2614 =head1 AUTHOR
2615
2616 Koha Development Team <http://koha-community.org/>
2617
2618 =cut