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 Lingua::Stem;
  25 use C4::Search::PazPar2;
  26 use XML::Simple;
  27 use C4::Dates qw(format_date);
  28 use C4::Members qw(GetHideLostItemsPreference);
  29 use C4::XSLT;
  30 use C4::Branch;
  31 use C4::Reserves;    # GetReserveStatus
  32 use C4::Debug;
  33 use C4::Charset;
  34 use YAML;
  35 use URI::Escape;
  36 use Business::ISBN;
  37 use MARC::Record;
  38 use MARC::Field;
  39 use 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                     'Record-control-number',
1202                     'rcn',
1203                     'Record-type',
1204                     'rtype',
1205                     'se',
1206                     'See',
1207                     'See-also',
1208                     'sn',
1209                     'Stock-number',
1210                     'su',
1211                     'Subject',
1212                     'Subject-heading-thesaurus',
1213                     'Subject-name-personal',
1214                     'Subject-subdivision',
1215                     'Summary',
1216                     'Suppress',
1217                     'su-geo',
1218                     'su-na',
1219                     'su-to',
1220                     'su-ut',
1221                     'ut',
1222                     'Term-genre-form',
1223                     'Term-genre-form-heading',
1224                     'Term-genre-form-see',
1225                     'Term-genre-form-seealso',
1226                     'ti',
1227                     'Title',
1228                     'Title-cover',
1229                     'Title-series',
1230                     'Title-uniform',
1231                     'Title-uniform-heading',
1232                     'Title-uniform-see',
1233                     'Title-uniform-seealso',
1234                     'totalissues',
1235                     'yr',
1236
1237                     # items indexes
1238                     'acqsource',
1239                     'barcode',
1240                     'bc',
1241                     'branch',
1242                     'ccode',
1243                     'classification-source',
1244                     'cn-sort',
1245                     'coded-location-qualifier',
1246                     'copynumber',
1247                     'damaged',
1248                     'datelastborrowed',
1249                     'datelastseen',
1250                     'holdingbranch',
1251                     'homebranch',
1252                     'issues',
1253                     'item',
1254                     'itemnumber',
1255                     'itype',
1256                     'Local-classification',
1257                     'location',
1258                     'lost',
1259                     'materials-specified',
1260                     'mc-ccode',
1261                     'mc-itype',
1262                     'mc-loc',
1263                     'notforloan',
1264                     'Number-local-acquisition',
1265                     'onloan',
1266                     'price',
1267                     'renewals',
1268                     'replacementprice',
1269                     'replacementpricedate',
1270                     'reserves',
1271                     'restricted',
1272                     'stack',
1273                     'stocknumber',
1274                     'inv',
1275                     'uri',
1276                     'withdrawn',
1277
1278                     # subject related
1279                   );
1280
1281     return \@indexes;
1282 }
1283
1284 =head2 _handle_exploding_index
1285
1286     my $query = _handle_exploding_index($index, $term)
1287
1288 Callback routine to generate the search for "exploding" indexes (i.e.
1289 those indexes which are turned into multiple or-connected searches based
1290 on authority data).
1291
1292 =cut
1293
1294 sub _handle_exploding_index {
1295     my ($QParser, $filter, $params, $negate, $server) = @_;
1296     my $index = $filter;
1297     my $term = join(' ', @$params);
1298
1299     return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1300
1301     my $marcflavour = C4::Context->preference('marcflavour');
1302
1303     my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1304     my $wantedcodes = '';
1305     my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1306     my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1307     foreach my $auth (@$results) {
1308         my $record = MARC::Record->new_from_usmarc($auth);
1309         my @references = $record->field('5..');
1310         if (@references) {
1311             if ($index eq 'su-br') {
1312                 $wantedcodes = 'g';
1313             } elsif ($index eq 'su-na') {
1314                 $wantedcodes = 'h';
1315             } elsif ($index eq 'su-rl') {
1316                 $wantedcodes = '';
1317             }
1318             foreach my $reference (@references) {
1319                 my $codes = $reference->subfield($codesubfield);
1320                 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1321             }
1322         }
1323     }
1324     my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1325     return $query;
1326 }
1327
1328 =head2 parseQuery
1329
1330     ( $operators, $operands, $indexes, $limits,
1331       $sort_by, $scan, $lang ) =
1332             buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1333
1334 Shim function to ease the transition from buildQuery to a new QueryParser.
1335 This function is called at the beginning of buildQuery, and modifies
1336 buildQuery's input. If it can handle the input, it returns a query that
1337 buildQuery will not try to parse.
1338 =cut
1339
1340 sub parseQuery {
1341     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1342
1343     my @operators = $operators ? @$operators : ();
1344     my @indexes   = $indexes   ? @$indexes   : ();
1345     my @operands  = $operands  ? @$operands  : ();
1346     my @limits    = $limits    ? @$limits    : ();
1347     my @sort_by   = $sort_by   ? @$sort_by   : ();
1348
1349     my $query = $operands[0];
1350     my $index;
1351     my $term;
1352     my $query_desc;
1353
1354     my $QParser;
1355     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1356     undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1357     undef $QParser if (scalar @limits > 0);
1358
1359     if ($QParser)
1360     {
1361         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1362         $query = '';
1363         for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1364             next unless $operands[$ii];
1365             $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1366               if ($query);
1367             if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1368                 $query .= $operands[$ii];
1369             }
1370             elsif ( $indexes[$ii] =~ m/su-/ ) {
1371                 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1372             }
1373             else {
1374                 $query .=
1375                   ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1376             }
1377         }
1378         foreach my $limit (@limits) {
1379         }
1380         if ( scalar(@sort_by) > 0 ) {
1381             my $modifier_re =
1382               '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1383             $query =~ s/$modifier_re//g;
1384             foreach my $modifier (@sort_by) {
1385                 $query .= " #$modifier";
1386             }
1387         }
1388
1389         $query_desc = $query;
1390         $query_desc =~ s/\s+/ /g;
1391         if ( C4::Context->preference("QueryWeightFields") ) {
1392         }
1393         $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1394               { 'target_syntax_callback' => \&_handle_exploding_index } );
1395         $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1396               { 'target_syntax_callback' => \&_handle_exploding_index } );
1397         $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1398               { 'target_syntax_callback' => \&_handle_exploding_index } );
1399         $QParser->parse($query);
1400         $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1401     }
1402     else {
1403         require Koha::QueryParser::Driver::PQF;
1404         my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1405         s/$modifier_re//g for @operands;
1406     }
1407
1408     return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1409 }
1410
1411 =head2 buildQuery
1412
1413 ( $error, $query,
1414 $simple_query, $query_cgi,
1415 $query_desc, $limit,
1416 $limit_cgi, $limit_desc,
1417 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1418
1419 Build queries and limits in CCL, CGI, Human,
1420 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1421
1422 See verbose embedded documentation.
1423
1424
1425 =cut
1426
1427 sub buildQuery {
1428     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1429
1430     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1431
1432     my $query_desc;
1433     ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1434
1435     # dereference
1436     my @operators = $operators ? @$operators : ();
1437     my @indexes   = $indexes   ? @$indexes   : ();
1438     my @operands  = $operands  ? @$operands  : ();
1439     my @limits    = $limits    ? @$limits    : ();
1440     my @sort_by   = $sort_by   ? @$sort_by   : ();
1441
1442     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1443     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1444     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1445     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1446     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1447
1448     my $query        = $operands[0];
1449     my $simple_query = $operands[0];
1450
1451     # initialize the variables we're passing back
1452     my $query_cgi;
1453     my $query_type;
1454
1455     my $limit;
1456     my $limit_cgi;
1457     my $limit_desc;
1458
1459     my $stopwords_removed;    # flag to determine if stopwords have been removed
1460
1461     my $cclq       = 0;
1462     my $cclindexes = getIndexes();
1463     if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1464         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1465             my $dx = lc($1);
1466             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1467         }
1468         $query = "ccl=$query" if $cclq;
1469     }
1470
1471 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1472 # DIAGNOSTIC ONLY!!
1473     if ( $query =~ /^ccl=/ ) {
1474         my $q=$';
1475         # This is needed otherwise ccl= and &limit won't work together, and
1476         # this happens when selecting a subject on the opac-detail page
1477         @limits = grep {!/^$/} @limits;
1478         if ( @limits ) {
1479             $q .= ' and '.join(' and ', @limits);
1480         }
1481         return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $q, '', '', '', '', 'ccl' );
1482     }
1483     if ( $query =~ /^cql=/ ) {
1484         return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', '', 'cql' );
1485     }
1486     if ( $query =~ /^pqf=/ ) {
1487         if ($query_desc) {
1488             $query_cgi = "q=".uri_escape_utf8($query_desc);
1489         } else {
1490             $query_desc = $';
1491             $query_cgi = "q=pqf=".uri_escape_utf8($');
1492         }
1493         return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1494     }
1495
1496     # pass nested queries directly
1497     # FIXME: need better handling of some of these variables in this case
1498     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1499     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1500 #    if ( $query =~ /(\(|\))/ ) {
1501 #        return (
1502 #            undef,              $query, $simple_query, $query_cgi,
1503 #            $query,             $limit, $limit_cgi,    $limit_desc,
1504 #            $stopwords_removed, 'ccl'
1505 #        );
1506 #    }
1507
1508 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1509 # query operands and indexes and add stemming, truncation, field weighting, etc.
1510 # Once we do so, we'll end up with a value in $query, just like if we had an
1511 # incoming $query from the user
1512     else {
1513         $query = ""
1514           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1515         my $previous_operand
1516           ;    # a flag used to keep track if there was a previous query
1517                # if there was, we can apply the current operator
1518                # for every operand
1519         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1520
1521             # COMBINE OPERANDS, INDEXES AND OPERATORS
1522             if ( $operands[$i] ) {
1523                 $operands[$i]=~s/^\s+//;
1524
1525               # A flag to determine whether or not to add the index to the query
1526                 my $indexes_set;
1527
1528 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1529                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1530                     $weight_fields    = 0;
1531                     $stemming         = 0;
1532                     $remove_stopwords = 0;
1533                 } else {
1534                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1535                 }
1536                 my $operand = $operands[$i];
1537                 my $index   = $indexes[$i];
1538
1539                 # Add index-specific attributes
1540
1541                 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1542                 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1543                 #which is processed higher up in this sub. Other than that, year searches are typically
1544                 #handled as limits which are not processed her either.
1545
1546                 # Date of Publication
1547                 if ( $index =~ /yr/ ) {
1548                     #weight_fields/relevance search causes errors with date ranges
1549                     #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1550                     #In the case of YYYY-YYYY, it will return no results
1551                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1552                 }
1553
1554                 # Date of Acquisition
1555                 elsif ( $index =~ /acqdate/ ) {
1556                     #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1557                     #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1558                       #top of the results just because they have lots of item records matching that date.
1559                     #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1560                       #irrelevant here
1561                     #remove_stopwords doesn't function anymore so is irrelevant
1562                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1563                 }
1564                 # ISBN,ISSN,Standard Number, don't need special treatment
1565                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1566                     (
1567                         $stemming,      $auto_truncation,
1568                         $weight_fields, $fuzzy_enabled,
1569                         $remove_stopwords
1570                     ) = ( 0, 0, 0, 0, 0 );
1571
1572                     if ( $index eq 'nb' ) {
1573                         if ( C4::Context->preference("SearchWithISBNVariations") ) {
1574                             my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1575                             $operands[$i] = $operand =  '(nb=' . join(' OR nb=', @isbns) . ')';
1576                             $indexes[$i] = $index = '';
1577                         }
1578                     }
1579                 }
1580
1581                 if(not $index){
1582                     $index = 'kw';
1583                 }
1584
1585                 # Set default structure attribute (word list)
1586                 my $struct_attr = q{};
1587                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1588                     $struct_attr = ",wrdl";
1589                 }
1590
1591                 # Some helpful index variants
1592                 my $index_plus       = $index . $struct_attr . ':';
1593                 my $index_plus_comma = $index . $struct_attr . ',';
1594
1595                 # Remove Stopwords
1596                 if ($remove_stopwords) {
1597                     ( $operand, $stopwords_removed ) =
1598                       _remove_stopwords( $operand, $index );
1599                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1600                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1601                       if ( $stopwords_removed && $DEBUG );
1602                 }
1603
1604                 if ($auto_truncation){
1605                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1606                                                 #FIXME only valid with LTR scripts
1607                                                 $operand=join(" ",map{
1608                                                                                         (index($_,"*")>0?"$_":"$_*")
1609                                                                                          }split (/\s+/,$operand));
1610                                                 warn $operand if $DEBUG;
1611                                         }
1612                                 }
1613
1614                 # Detect Truncation
1615                 my $truncated_operand;
1616                 my( $nontruncated, $righttruncated, $lefttruncated,
1617                     $rightlefttruncated, $regexpr
1618                 ) = _detect_truncation( $operand, $index );
1619                 warn
1620 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1621                   if $DEBUG;
1622
1623                 # Apply Truncation
1624                 if (
1625                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1626                     scalar(@$rightlefttruncated) > 0 )
1627                 {
1628
1629                # Don't field weight or add the index to the query, we do it here
1630                     $indexes_set = 1;
1631                     undef $weight_fields;
1632                     my $previous_truncation_operand;
1633                     if (scalar @$nontruncated) {
1634                         $truncated_operand .= "$index_plus @$nontruncated ";
1635                         $previous_truncation_operand = 1;
1636                     }
1637                     if (scalar @$righttruncated) {
1638                         $truncated_operand .= "and " if $previous_truncation_operand;
1639                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1640                         $previous_truncation_operand = 1;
1641                     }
1642                     if (scalar @$lefttruncated) {
1643                         $truncated_operand .= "and " if $previous_truncation_operand;
1644                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1645                         $previous_truncation_operand = 1;
1646                     }
1647                     if (scalar @$rightlefttruncated) {
1648                         $truncated_operand .= "and " if $previous_truncation_operand;
1649                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1650                         $previous_truncation_operand = 1;
1651                     }
1652                 }
1653                 $operand = $truncated_operand if $truncated_operand;
1654                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1655
1656                 # Handle Stemming
1657                 my $stemmed_operand;
1658                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1659                                                                                 if $stemming;
1660
1661                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1662
1663                 # Handle Field Weighting
1664                 my $weighted_operand;
1665                 if ($weight_fields) {
1666                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1667                     $operand = $weighted_operand;
1668                     $indexes_set = 1;
1669                 }
1670
1671                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1672
1673                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1674                     query => $query,
1675                     query_cgi => $query_cgi,
1676                     query_desc => $query_desc,
1677                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1678                     parsed_operand => $operand,
1679                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1680                     index => $index,
1681                     index_plus => $index_plus,
1682                     indexes_set => $indexes_set,
1683                     previous_operand => $previous_operand,
1684                 });
1685
1686             }    #/if $operands
1687         }    # /for
1688     }
1689     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1690
1691     # add limits
1692     my %group_OR_limits;
1693     my $availability_limit;
1694     foreach my $this_limit (@limits) {
1695         next unless $this_limit;
1696         if ( $this_limit =~ /available/ ) {
1697 #
1698 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1699 ## In English:
1700 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1701             $availability_limit .=
1702 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1703             $limit_cgi  .= "&limit=available";
1704             $limit_desc .= "";
1705         }
1706
1707         # group_OR_limits, prefixed by mc-
1708         # OR every member of the group
1709         elsif ( $this_limit =~ /mc/ ) {
1710             my ($k,$v) = split(/:/, $this_limit,2);
1711             if ( $k !~ /mc-i(tem)?type/ ) {
1712                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1713                 $this_limit =~ tr/"//d;
1714                 $this_limit = $k.":'".$v."'";
1715             }
1716
1717             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1718             $limit_desc      .= " or " if $group_OR_limits{$k};
1719             $group_OR_limits{$k} .= "$this_limit";
1720             $limit_cgi       .= "&limit=" . uri_escape_utf8($this_limit);
1721             $limit_desc      .= " $this_limit";
1722         }
1723
1724         # Regular old limits
1725         else {
1726             $limit .= " and " if $limit || $query;
1727             $limit      .= "$this_limit";
1728             $limit_cgi  .= "&limit=" . uri_escape_utf8($this_limit);
1729             if ($this_limit =~ /^branch:(.+)/) {
1730                 my $branchcode = $1;
1731                 my $branchname = GetBranchName($branchcode);
1732                 if (defined $branchname) {
1733                     $limit_desc .= " branch:$branchname";
1734                 } else {
1735                     $limit_desc .= " $this_limit";
1736                 }
1737             } else {
1738                 $limit_desc .= " $this_limit";
1739             }
1740         }
1741     }
1742     foreach my $k (keys (%group_OR_limits)) {
1743         $limit .= " and " if ( $query || $limit );
1744         $limit .= "($group_OR_limits{$k})";
1745     }
1746     if ($availability_limit) {
1747         $limit .= " and " if ( $query || $limit );
1748         $limit .= "($availability_limit)";
1749     }
1750
1751     # Normalize the query and limit strings
1752     # This is flawed , means we can't search anything with : in it
1753     # if user wants to do ccl or cql, start the query with that
1754 #    $query =~ s/:/=/g;
1755     #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1756     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1757     $query =~ s/(?<=(wrdl)):/=/g;
1758     $query =~ s/(?<=(trn|phr)):/=/g;
1759     $query =~ s/(?<=(st-numeric)):/=/g;
1760     $query =~ s/(?<=(st-year)):/=/g;
1761     $query =~ s/(?<=(st-date-normalized)):/=/g;
1762     $limit =~ s/:/=/g;
1763     for ( $query, $query_desc, $limit, $limit_desc ) {
1764         s/  +/ /g;    # remove extra spaces
1765         s/^ //g;     # remove any beginning spaces
1766         s/ $//g;     # remove any ending spaces
1767         s/==/=/g;    # remove double == from query
1768     }
1769     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1770
1771     for ($query_cgi,$simple_query) {
1772         s/"//g;
1773     }
1774     # append the limit to the query
1775     $query .= " " . $limit;
1776
1777     # Warnings if DEBUG
1778     if ($DEBUG) {
1779         warn "QUERY:" . $query;
1780         warn "QUERY CGI:" . $query_cgi;
1781         warn "QUERY DESC:" . $query_desc;
1782         warn "LIMIT:" . $limit;
1783         warn "LIMIT CGI:" . $limit_cgi;
1784         warn "LIMIT DESC:" . $limit_desc;
1785         warn "---------\nLeave buildQuery\n---------";
1786     }
1787     return (
1788         undef,              $query, $simple_query, $query_cgi,
1789         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1790         $stopwords_removed, $query_type
1791     );
1792 }
1793
1794 =head2 _build_initial_query
1795
1796   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1797
1798   Build a section of the initial query containing indexes, operators, and operands.
1799
1800 =cut
1801
1802 sub _build_initial_query {
1803     my ($params) = @_;
1804
1805     my $operator = "";
1806     if ($params->{previous_operand}){
1807         #If there is a previous operand, add a supplied operator or the default 'and'
1808         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1809     }
1810
1811     #NOTE: indexes_set is typically set when doing truncation or field weighting
1812     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1813
1814     #e.g. "kw,wrdl:test"
1815     #e.g. " and kw,wrdl:test"
1816     $params->{query} .= $operator . $operand;
1817
1818     $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1819     $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1820     $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1821
1822     #e.g. " and kw,wrdl: test"
1823     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1824
1825     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1826
1827     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1828 }
1829
1830 =head2 searchResults
1831
1832   my @search_results = searchResults($search_context, $searchdesc, $hits,
1833                                      $results_per_page, $offset, $scan,
1834                                      @marcresults);
1835
1836 Format results in a form suitable for passing to the template
1837
1838 =cut
1839
1840 # IMO this subroutine is pretty messy still -- it's responsible for
1841 # building the HTML output for the template
1842 sub searchResults {
1843     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1844     my $dbh = C4::Context->dbh;
1845     my @newresults;
1846
1847     require C4::Items;
1848
1849     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1850     my ($is_opac, $hidelostitems);
1851     if ($search_context eq 'opac') {
1852         $hidelostitems = C4::Context->preference('hidelostitems');
1853         $is_opac       = 1;
1854     }
1855
1856     #Build branchnames hash
1857     #find branchname
1858     #get branch information.....
1859     my %branches;
1860     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1861     $bsth->execute();
1862     while ( my $bdata = $bsth->fetchrow_hashref ) {
1863         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1864     }
1865 # FIXME - We build an authorised values hash here, using the default framework
1866 # though it is possible to have different authvals for different fws.
1867
1868     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1869
1870     # get notforloan authorised value list (see $shelflocations  FIXME)
1871     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1872
1873     #Get itemtype hash
1874     my %itemtypes = %{ GetItemTypes() };
1875
1876     #search item field code
1877     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1878
1879     ## find column names of items related to MARC
1880     my %subfieldstosearch;
1881     my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns;
1882     for my $column ( @columns ) {
1883         my ( $tagfield, $tagsubfield ) =
1884           &GetMarcFromKohaField( "items." . $column, "" );
1885         if ( defined $tagsubfield ) {
1886             $subfieldstosearch{$column} = $tagsubfield;
1887         }
1888     }
1889
1890     # handle which records to actually retrieve
1891     my $times;
1892     if ( $hits && $offset + $results_per_page <= $hits ) {
1893         $times = $offset + $results_per_page;
1894     }
1895     else {
1896         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1897     }
1898
1899     my $marcflavour = C4::Context->preference("marcflavour");
1900     # We get the biblionumber position in MARC
1901     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1902
1903     # loop through all of the records we've retrieved
1904     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1905
1906         my $marcrecord;
1907         if ($scan) {
1908             # For Scan searches we built USMARC data
1909             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1910         } else {
1911             # Normal search, render from Zebra's output
1912             $marcrecord = new_record_from_zebra(
1913                 'biblioserver',
1914                 $marcresults->[$i]
1915             );
1916
1917             if ( ! defined $marcrecord ) {
1918                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1919                 next;
1920             }
1921         }
1922
1923         my $fw = $scan
1924              ? undef
1925              : $bibliotag < 10
1926                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1927                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1928
1929         SetUTF8Flag($marcrecord);
1930         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1931         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1932         $oldbiblio->{result_number} = $i + 1;
1933
1934         # add imageurl to itemtype if there is one
1935         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1936
1937         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1938                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1939                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1940                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1941                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1942                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1943
1944                 # edition information, if any
1945         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1946         $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{translated_description};
1947  # Build summary if there is one (the summary is defined in the itemtypes table)
1948  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1949         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1950             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1951             my @fields  = $marcrecord->fields();
1952
1953             my $newsummary;
1954             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1955                 my $tags = {};
1956                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1957                     $tag =~ /(.{3})(.)/;
1958                     if($marcrecord->field($1)){
1959                         my @abc = $marcrecord->field($1)->subfield($2);
1960                         $tags->{$tag} = $#abc + 1 ;
1961                     }
1962                 }
1963
1964                 # We catch how many times to repeat this line
1965                 my $max = 0;
1966                 foreach my $tag (keys(%$tags)){
1967                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1968                  }
1969
1970                 # we replace, and repeat each line
1971                 for (my $i = 0 ; $i < $max ; $i++){
1972                     my $newline = $line;
1973
1974                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1975                         $tag =~ /(.{3})(.)/;
1976
1977                         if($marcrecord->field($1)){
1978                             my @repl = $marcrecord->field($1)->subfield($2);
1979                             my $subfieldvalue = $repl[$i];
1980                             $newline =~ s/\[$tag\]/$subfieldvalue/g;
1981                         }
1982                     }
1983                     $newsummary .= "$newline\n";
1984                 }
1985             }
1986
1987             $newsummary =~ s/\[(.*?)]//g;
1988             $newsummary =~ s/\n/<br\/>/g;
1989             $oldbiblio->{summary} = $newsummary;
1990         }
1991
1992         # Pull out the items fields
1993         my @fields = $marcrecord->field($itemtag);
1994         my $marcflavor = C4::Context->preference("marcflavour");
1995         # adding linked items that belong to host records
1996         my $analyticsfield = '773';
1997         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1998             $analyticsfield = '773';
1999         } elsif ($marcflavor eq 'UNIMARC') {
2000             $analyticsfield = '461';
2001         }
2002         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2003             my $hostbiblionumber = $hostfield->subfield("0");
2004             my $linkeditemnumber = $hostfield->subfield("9");
2005             if(!$hostbiblionumber eq undef){
2006                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2007                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2008                 if(!$hostbiblio eq undef){
2009                     my @hostitems = $hostbiblio->field($itemfield);
2010                     foreach my $hostitem (@hostitems){
2011                         if ($hostitem->subfield("9") eq $linkeditemnumber){
2012                             my $linkeditem =$hostitem;
2013                             # append linked items if they exist
2014                             if (!$linkeditem eq undef){
2015                                 push (@fields, $linkeditem);}
2016                         }
2017                     }
2018                 }
2019             }
2020         }
2021
2022         # Setting item statuses for display
2023         my @available_items_loop;
2024         my @onloan_items_loop;
2025         my @other_items_loop;
2026
2027         my $available_items;
2028         my $onloan_items;
2029         my $other_items;
2030
2031         my $ordered_count         = 0;
2032         my $available_count       = 0;
2033         my $onloan_count          = 0;
2034         my $longoverdue_count     = 0;
2035         my $other_count           = 0;
2036         my $withdrawn_count        = 0;
2037         my $itemlost_count        = 0;
2038         my $hideatopac_count      = 0;
2039         my $itembinding_count     = 0;
2040         my $itemdamaged_count     = 0;
2041         my $item_in_transit_count = 0;
2042         my $can_place_holds       = 0;
2043         my $item_onhold_count     = 0;
2044         my $notforloan_count      = 0;
2045         my $items_count           = scalar(@fields);
2046         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2047         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2048         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2049
2050         # loop through every item
2051         foreach my $field (@fields) {
2052             my $item;
2053
2054             # populate the items hash
2055             foreach my $code ( keys %subfieldstosearch ) {
2056                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2057             }
2058             $item->{description} = $itemtypes{ $item->{itype} }{translated_description};
2059
2060                 # OPAC hidden items
2061             if ($is_opac) {
2062                 # hidden because lost
2063                 if ($hidelostitems && $item->{itemlost}) {
2064                     $hideatopac_count++;
2065                     next;
2066                 }
2067                 # hidden based on OpacHiddenItems syspref
2068                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2069                 if (scalar @hi) {
2070                     push @hiddenitems, @hi;
2071                     $hideatopac_count++;
2072                     next;
2073                 }
2074             }
2075
2076             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2077             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2078
2079             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2080             if ($item->{$hbranch}) {
2081                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2082             }
2083             elsif ($item->{$otherbranch}) {     # Last resort
2084                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2085             }
2086
2087                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2088 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2089             my $userenv = C4::Context->userenv;
2090             if ( $item->{onloan}
2091                 && !( C4::Members::GetHideLostItemsPreference( $userenv->{'number'} ) && $item->{itemlost} ) )
2092             {
2093                 $onloan_count++;
2094                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2095                 $onloan_items->{$key}->{due_date} = format_date( $item->{onloan} );
2096                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2097                 $onloan_items->{$key}->{branchname}     = $item->{branchname};
2098                 $onloan_items->{$key}->{location}       = $shelflocations->{ $item->{location} };
2099                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2100                 $onloan_items->{$key}->{description}    = $item->{description};
2101                 $onloan_items->{$key}->{imageurl} =
2102                   getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2103
2104                 # if something's checked out and lost, mark it as 'long overdue'
2105                 if ( $item->{itemlost} ) {
2106                     $onloan_items->{$key}->{longoverdue}++;
2107                     $longoverdue_count++;
2108                 }
2109                 else {    # can place holds as long as item isn't lost
2110                     $can_place_holds = 1;
2111                 }
2112             }
2113
2114          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2115             else {
2116
2117                 $item->{notforloan}=1 if !$item->{notforloan}  && $itemtypes{ C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype} }->{notforloan};
2118
2119                 # item is on order
2120                 if ( $item->{notforloan} < 0 ) {
2121                     $ordered_count++;
2122                 } elsif ( $item->{notforloan} > 0 ) {
2123                     $notforloan_count++;
2124                 }
2125
2126                 # is item in transit?
2127                 my $transfertwhen = '';
2128                 my ($transfertfrom, $transfertto);
2129
2130                 # is item on the reserve shelf?
2131                 my $reservestatus = '';
2132
2133                 unless ($item->{withdrawn}
2134                         || $item->{itemlost}
2135                         || $item->{damaged}
2136                         || $item->{notforloan}
2137                         || $items_count > 20) {
2138
2139                     # A couple heuristics to limit how many times
2140                     # we query the database for item transfer information, sacrificing
2141                     # accuracy in some cases for speed;
2142                     #
2143                     # 1. don't query if item has one of the other statuses
2144                     # 2. don't check transit status if the bib has
2145                     #    more than 20 items
2146                     #
2147                     # FIXME: to avoid having the query the database like this, and to make
2148                     #        the in transit status count as unavailable for search limiting,
2149                     #        should map transit status to record indexed in Zebra.
2150                     #
2151                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2152                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2153                 }
2154
2155                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2156                 if (   $item->{withdrawn}
2157                     || $item->{itemlost}
2158                     || $item->{damaged}
2159                     || $item->{notforloan}
2160                     || $reservestatus eq 'Waiting'
2161                     || ($transfertwhen ne ''))
2162                 {
2163                     $withdrawn_count++        if $item->{withdrawn};
2164                     $itemlost_count++        if $item->{itemlost};
2165                     $itemdamaged_count++     if $item->{damaged};
2166                     $item_in_transit_count++ if $transfertwhen ne '';
2167                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2168                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2169
2170                     # can place a hold on a item if
2171                     # not lost nor withdrawn
2172                     # not damaged unless AllowHoldsOnDamagedItems is true
2173                     # item is either for loan or on order (notforloan < 0)
2174                     $can_place_holds = 1
2175                       if (
2176                            !$item->{itemlost}
2177                         && !$item->{withdrawn}
2178                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2179                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2180                       );
2181
2182                     $other_count++;
2183
2184                     my $key = $prefix . $item->{status};
2185                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2186                         $other_items->{$key}->{$_} = $item->{$_};
2187                     }
2188                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2189                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2190                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2191                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2192                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2193                                         $other_items->{$key}->{description} = $item->{description};
2194                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2195                 }
2196                 # item is available
2197                 else {
2198                     $can_place_holds = 1;
2199                     $available_count++;
2200                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2201                                         foreach (qw(branchname itemcallnumber description)) {
2202                         $available_items->{$prefix}->{$_} = $item->{$_};
2203                                         }
2204                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2205                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2206                 }
2207             }
2208         }    # notforloan, item level and biblioitem level
2209
2210         # if all items are hidden, do not show the record
2211         if ($items_count > 0 && $hideatopac_count == $items_count) {
2212             next;
2213         }
2214
2215         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2216         for my $key ( sort keys %$onloan_items ) {
2217             (++$onloanitemscount > $maxitems) and last;
2218             push @onloan_items_loop, $onloan_items->{$key};
2219         }
2220         for my $key ( sort keys %$other_items ) {
2221             (++$otheritemscount > $maxitems) and last;
2222             push @other_items_loop, $other_items->{$key};
2223         }
2224         for my $key ( sort keys %$available_items ) {
2225             (++$availableitemscount > $maxitems) and last;
2226             push @available_items_loop, $available_items->{$key}
2227         }
2228
2229         # XSLT processing of some stuff
2230         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2231         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2232             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2233         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2234         }
2235
2236         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2237         if (!C4::Context->preference("item-level_itypes")) {
2238             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2239                 $can_place_holds = 0;
2240             }
2241         }
2242         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2243         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2244         $oldbiblio->{items_count}          = $items_count;
2245         $oldbiblio->{available_items_loop} = \@available_items_loop;
2246         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2247         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2248         $oldbiblio->{availablecount}       = $available_count;
2249         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2250         $oldbiblio->{onloancount}          = $onloan_count;
2251         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2252         $oldbiblio->{othercount}           = $other_count;
2253         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2254         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2255         $oldbiblio->{itemlostcount}        = $itemlost_count;
2256         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2257         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2258         $oldbiblio->{onholdcount}          = $item_onhold_count;
2259         $oldbiblio->{orderedcount}         = $ordered_count;
2260         $oldbiblio->{notforloancount}      = $notforloan_count;
2261
2262         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2263             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2264             my $subfields = substr $fieldspec, 3;
2265             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2266             my @alternateholdingsinfo = ();
2267             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2268             my $alternateholdingscount = 0;
2269
2270             for my $field (@holdingsfields) {
2271                 my %holding = ( holding => '' );
2272                 my $havesubfield = 0;
2273                 for my $subfield ($field->subfields()) {
2274                     if ((index $subfields, $$subfield[0]) >= 0) {
2275                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2276                         $holding{'holding'} .= $$subfield[1];
2277                         $havesubfield++;
2278                     }
2279                 }
2280                 if ($havesubfield) {
2281                     push(@alternateholdingsinfo, \%holding);
2282                     $alternateholdingscount++;
2283                 }
2284             }
2285
2286             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2287             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2288         }
2289
2290         push( @newresults, $oldbiblio );
2291     }
2292
2293     return @newresults;
2294 }
2295
2296 =head2 SearchAcquisitions
2297     Search for acquisitions
2298 =cut
2299
2300 sub SearchAcquisitions{
2301     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2302
2303     my $dbh=C4::Context->dbh;
2304     # Variable initialization
2305     my $str=qq|
2306     SELECT marcxml
2307     FROM biblio
2308     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2309     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2310     WHERE dateaccessioned BETWEEN ? AND ?
2311     |;
2312
2313     my (@params,@loopcriteria);
2314
2315     push @params, $datebegin->output("iso");
2316     push @params, $dateend->output("iso");
2317
2318     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2319         if(C4::Context->preference("item-level_itypes")){
2320             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2321         }else{
2322             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2323         }
2324         push @params, @$itemtypes;
2325     }
2326
2327     if ($criteria =~/itemtype/){
2328         if(C4::Context->preference("item-level_itypes")){
2329             $str .= "AND items.itype=? ";
2330         }else{
2331             $str .= "AND biblioitems.itemtype=? ";
2332         }
2333
2334         if(scalar(@$itemtypes) == 0){
2335             my $itypes = GetItemTypes();
2336             for my $key (keys %$itypes){
2337                 push @$itemtypes, $key;
2338             }
2339         }
2340
2341         @loopcriteria= @$itemtypes;
2342     }elsif ($criteria=~/itemcallnumber/){
2343         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2344                  OR items.itemcallnumber is NULL
2345                  OR items.itemcallnumber = '')";
2346
2347         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2348     }else {
2349         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2350         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2351     }
2352
2353     if ($orderby =~ /date_desc/){
2354         $str.=" ORDER BY dateaccessioned DESC";
2355     } else {
2356         $str.=" ORDER BY title";
2357     }
2358
2359     my $qdataacquisitions=$dbh->prepare($str);
2360
2361     my @loopacquisitions;
2362     foreach my $value(@loopcriteria){
2363         push @params,$value;
2364         my %cell;
2365         $cell{"title"}=$value;
2366         $cell{"titlecode"}=$value;
2367
2368         eval{$qdataacquisitions->execute(@params);};
2369
2370         if ($@){ warn "recentacquisitions Error :$@";}
2371         else {
2372             my @loopdata;
2373             while (my $data=$qdataacquisitions->fetchrow_hashref){
2374                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2375             }
2376             $cell{"loopdata"}=\@loopdata;
2377         }
2378         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2379         pop @params;
2380     }
2381     $qdataacquisitions->finish;
2382     return \@loopacquisitions;
2383 }
2384
2385 =head2 enabled_staff_search_views
2386
2387 %hash = enabled_staff_search_views()
2388
2389 This function returns a hash that contains three flags obtained from the system
2390 preferences, used to determine whether a particular staff search results view
2391 is enabled.
2392
2393 =over 2
2394
2395 =item C<Output arg:>
2396
2397     * $hash{can_view_MARC} is true only if the MARC view is enabled
2398     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2399     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2400
2401 =item C<usage in the script:>
2402
2403 =back
2404
2405 $template->param ( C4::Search::enabled_staff_search_views );
2406
2407 =cut
2408
2409 sub enabled_staff_search_views
2410 {
2411         return (
2412                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2413                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2414                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2415         );
2416 }
2417
2418 =head2 z3950_search_args
2419
2420 $arrayref = z3950_search_args($matchpoints)
2421
2422 This function returns an array reference that contains the search parameters to be
2423 passed to the Z39.50 search script (z3950_search.pl). The array elements
2424 are hash refs whose keys are name and value, and whose values are the
2425 name of a search parameter, the value of that search parameter and the URL encoded
2426 value of that parameter.
2427
2428 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2429
2430 The search parameter values are obtained from the bibliographic record whose
2431 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2432
2433 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2434 a general purpose search argument. In this case, the returned array contains only
2435 entry: the key is 'title' and the value is derived from $matchpoints.
2436
2437 If a search parameter value is undefined or empty, it is not included in the returned
2438 array.
2439
2440 The returned array reference may be passed directly to the template parameters.
2441
2442 =over 2
2443
2444 =item C<Output arg:>
2445
2446     * $array containing hash refs as described above
2447
2448 =item C<usage in the script:>
2449
2450 =back
2451
2452 $data = Biblio::GetBiblioData($bibno);
2453 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2454
2455 *OR*
2456
2457 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2458
2459 =cut
2460
2461 sub z3950_search_args {
2462     my $bibrec = shift;
2463
2464     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2465     my $isbn = Business::ISBN->new( $isbn_string );
2466
2467     if (defined $isbn && $isbn->is_valid)
2468     {
2469         if ( ref($bibrec) ) {
2470             $bibrec->{isbn} = $isbn_string;
2471             $bibrec->{title} = undef;
2472         } else {
2473             $bibrec = { isbn => $isbn_string };
2474         }
2475     }
2476     else {
2477         $bibrec = { title => $bibrec } if !ref $bibrec;
2478     }
2479     my $array = [];
2480     for my $field (qw/ lccn isbn issn title author dewey subject /)
2481     {
2482         push @$array, { name => $field, value => $bibrec->{$field} }
2483           if defined $bibrec->{$field};
2484     }
2485     return $array;
2486 }
2487
2488 =head2 GetDistinctValues($field);
2489
2490 C<$field> is a reference to the fields array
2491
2492 =cut
2493
2494 sub GetDistinctValues {
2495     my ($fieldname,$string)=@_;
2496     # returns a reference to a hash of references to branches...
2497     if ($fieldname=~/\./){
2498                         my ($table,$column)=split /\./, $fieldname;
2499                         my $dbh = C4::Context->dbh;
2500                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2501                         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 ");
2502                         $sth->execute;
2503                         my $elements=$sth->fetchall_arrayref({});
2504                         return $elements;
2505    }
2506    else {
2507                 $string||= qq("");
2508                 my @servers=qw<biblioserver authorityserver>;
2509                 my (@zconns,@results);
2510         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2511                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2512                         $results[$i] =
2513                       $zconns[$i]->scan(
2514                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2515                       );
2516                 }
2517                 # The big moment: asynchronously retrieve results from all servers
2518                 my @elements;
2519         _ZOOM_event_loop(
2520             \@zconns,
2521             \@results,
2522             sub {
2523                 my ( $i, $size ) = @_;
2524                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2525                     my %hashscan;
2526                     @hashscan{qw(value cnt)} =
2527                       $results[ $i - 1 ]->display_term($j);
2528                     push @elements, \%hashscan;
2529                 }
2530             }
2531         );
2532                 return \@elements;
2533    }
2534 }
2535
2536 =head2 _ZOOM_event_loop
2537
2538     _ZOOM_event_loop(\@zconns, \@results, sub {
2539         my ( $i, $size ) = @_;
2540         ....
2541     } );
2542
2543 Processes a ZOOM event loop and passes control to a closure for
2544 processing the results, and destroying the resultsets.
2545
2546 =cut
2547
2548 sub _ZOOM_event_loop {
2549     my ($zconns, $results, $callback) = @_;
2550     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2551         my $ev = $zconns->[ $i - 1 ]->last_event();
2552         if ( $ev == ZOOM::Event::ZEND ) {
2553             next unless $results->[ $i - 1 ];
2554             my $size = $results->[ $i - 1 ]->size();
2555             if ( $size > 0 ) {
2556                 $callback->($i, $size);
2557             }
2558         }
2559     }
2560
2561     foreach my $result (@$results) {
2562         $result->destroy();
2563     }
2564 }
2565
2566 =head2 new_record_from_zebra
2567
2568 Given raw data from a Zebra result set, return a MARC::Record object
2569
2570 This helper function is needed to take into account all the involved
2571 system preferences and configuration variables to properly create the
2572 MARC::Record object.
2573
2574 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2575 data. If we are using DOM, then it has to be MARCXML.
2576
2577 =cut
2578
2579 sub new_record_from_zebra {
2580
2581     my $server   = shift;
2582     my $raw_data = shift;
2583     # Set the default indexing modes
2584     my $index_mode = ( $server eq 'biblioserver' )
2585                         ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2586                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2587
2588     my $marc_record =  eval {
2589         if ( $index_mode eq 'dom' ) {
2590             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2591         } else {
2592             MARC::Record->new_from_usmarc( $raw_data );
2593         }
2594     };
2595
2596     if ($@) {
2597         return;
2598     } else {
2599         return $marc_record;
2600     }
2601
2602 }
2603
2604 END { }    # module clean-up code here (global destructor)
2605
2606 1;
2607 __END__
2608
2609 =head1 AUTHOR
2610
2611 Koha Development Team <http://koha-community.org/>
2612
2613 =cut