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