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 under the
   6 # terms of the GNU General Public License as published by the Free Software
   7 # Foundation; either version 2 of the License, or (at your option) any later
   8 # version.
   9 #
  10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  12 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU General Public License along with
  15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
  16 # Suite 330, Boston, MA  02111-1307 USA
  17
  18 use strict;
  19 #use warnings; FIXME - Bug 2505
  20 require Exporter;
  21 use C4::Context;
  22 use C4::Biblio;    # GetMarcFromKohaField, GetBiblioData
  23 use C4::Koha;      # getFacets
  24 use Lingua::Stem;
  25 use C4::Search::PazPar2;
  26 use XML::Simple;
  27 use C4::Dates qw(format_date);
  28 use C4::Members qw(GetHideLostItemsPreference);
  29 use C4::XSLT;
  30 use C4::Branch;
  31 use C4::Reserves;    # GetReserveStatus
  32 use C4::Debug;
  33 use C4::Charset;
  34 use YAML;
  35 use URI::Escape;
  36 use Business::ISBN;
  37 use MARC::Record;
  38 use MARC::Field;
  39 use utf8;
  40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  41
  42 # set the version for version checking
  43 BEGIN {
  44     $VERSION = 3.07.00.049;
  45     $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
  46 }
  47
  48 =head1 NAME
  49
  50 C4::Search - Functions for searching the Koha catalog.
  51
  52 =head1 SYNOPSIS
  53
  54 See opac/opac-search.pl or catalogue/search.pl for example of usage
  55
  56 =head1 DESCRIPTION
  57
  58 This module provides searching functions for Koha's bibliographic databases
  59
  60 =head1 FUNCTIONS
  61
  62 =cut
  63
  64 @ISA    = qw(Exporter);
  65 @EXPORT = qw(
  66   &FindDuplicate
  67   &SimpleSearch
  68   &searchResults
  69   &getRecords
  70   &buildQuery
  71   &GetDistinctValues
  72   &enabled_staff_search_views
  73   &PurgeSearchHistory
  74 );
  75
  76 # make all your functions, whether exported or not;
  77
  78 =head2 FindDuplicate
  79
  80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
  81
  82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
  83
  84 =cut
  85
  86 sub FindDuplicate {
  87     my ($record) = @_;
  88     my $dbh = C4::Context->dbh;
  89     my $result = TransformMarcToKoha( $dbh, $record, '' );
  90     my $sth;
  91     my $query;
  92     my $search;
  93     my $type;
  94     my ( $biblionumber, $title );
  95
  96     # search duplicate on ISBN, easy and fast..
  97     # ... normalize first
  98     if ( $result->{isbn} ) {
  99         $result->{isbn} =~ s/\(.*$//;
 100         $result->{isbn} =~ s/\s+$//;
 101         $query = "isbn:$result->{isbn}";
 102     }
 103     else {
 104         my $QParser;
 105         $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
 106         my $titleindex;
 107         my $authorindex;
 108         my $op;
 109
 110         if ($QParser) {
 111             $titleindex = 'title|exact';
 112             $authorindex = 'author|exact';
 113             $op = '&&';
 114             $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 115         } else {
 116             $titleindex = 'ti,ext';
 117             $authorindex = 'au,ext';
 118             $op = 'and';
 119         }
 120
 121         $result->{title} =~ s /\\//g;
 122         $result->{title} =~ s /\"//g;
 123         $result->{title} =~ s /\(//g;
 124         $result->{title} =~ s /\)//g;
 125
 126         # FIXME: instead of removing operators, could just do
 127         # quotes around the value
 128         $result->{title} =~ s/(and|or|not)//g;
 129         $query = "$titleindex:\"$result->{title}\"";
 130         if   ( $result->{author} ) {
 131             $result->{author} =~ s /\\//g;
 132             $result->{author} =~ s /\"//g;
 133             $result->{author} =~ s /\(//g;
 134             $result->{author} =~ s /\)//g;
 135
 136             # remove valid operators
 137             $result->{author} =~ s/(and|or|not)//g;
 138             $query .= " $op $authorindex:\"$result->{author}\"";
 139         }
 140     }
 141
 142     my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
 143     my @results;
 144     if (!defined $error) {
 145         foreach my $possible_duplicate_record (@{$searchresults}) {
 146             my $marcrecord = new_record_from_zebra(
 147                 'biblioserver',
 148                 $possible_duplicate_record
 149             );
 150
 151             my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
 152
 153             # FIXME :: why 2 $biblionumber ?
 154             if ($result) {
 155                 push @results, $result->{'biblionumber'};
 156                 push @results, $result->{'title'};
 157             }
 158         }
 159     }
 160     return @results;
 161 }
 162
 163 =head2 SimpleSearch
 164
 165 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
 166
 167 This function provides a simple search API on the bibliographic catalog
 168
 169 =over 2
 170
 171 =item C<input arg:>
 172
 173     * $query can be a simple keyword or a complete CCL query
 174     * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
 175     * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
 176     * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
 177
 178
 179 =item C<Return:>
 180
 181     Returns an array consisting of three elements
 182     * $error is undefined unless an error is detected
 183     * $results is a reference to an array of records.
 184     * $total_hits is the number of hits that would have been returned with no limit
 185
 186     If an error is returned the two other return elements are undefined. If error itself is undefined
 187     the other two elements are always defined
 188
 189 =item C<usage in the script:>
 190
 191 =back
 192
 193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
 194
 195 if (defined $error) {
 196     $template->param(query_error => $error);
 197     warn "error: ".$error;
 198     output_html_with_http_headers $input, $cookie, $template->output;
 199     exit;
 200 }
 201
 202 my $hits = @{$marcresults};
 203 my @results;
 204
 205 for my $r ( @{$marcresults} ) {
 206     my $marcrecord = MARC::File::USMARC::decode($r);
 207     my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
 208
 209     #build the iarray of hashs for the template.
 210     push @results, {
 211         title           => $biblio->{'title'},
 212         subtitle        => $biblio->{'subtitle'},
 213         biblionumber    => $biblio->{'biblionumber'},
 214         author          => $biblio->{'author'},
 215         publishercode   => $biblio->{'publishercode'},
 216         publicationyear => $biblio->{'publicationyear'},
 217         };
 218
 219 }
 220
 221 $template->param(result=>\@results);
 222
 223 =cut
 224
 225 sub SimpleSearch {
 226     my ( $query, $offset, $max_results, $servers )  = @_;
 227
 228     return ( 'No query entered', undef, undef ) unless $query;
 229     # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
 230     my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
 231     my @zoom_queries;
 232     my @tmpresults;
 233     my @zconns;
 234     my $results = [];
 235     my $total_hits = 0;
 236
 237     my $QParser;
 238     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
 239     if ($QParser) {
 240         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 241     }
 242
 243     # Initialize & Search Zebra
 244     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 245         eval {
 246             $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 247             if ($QParser) {
 248                 $query =~ s/=/:/g;
 249                 $QParser->parse( $query );
 250                 $query = $QParser->target_syntax($servers[$i]);
 251                 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
 252             } else {
 253                 $query =~ s/:/=/g;
 254                 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
 255             }
 256             $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
 257
 258             # error handling
 259             my $error =
 260                 $zconns[$i]->errmsg() . " ("
 261               . $zconns[$i]->errcode() . ") "
 262               . $zconns[$i]->addinfo() . " "
 263               . $zconns[$i]->diagset();
 264
 265             return ( $error, undef, undef ) if $zconns[$i]->errcode();
 266         };
 267         if ($@) {
 268
 269             # caught a ZOOM::Exception
 270             my $error =
 271                 $@->message() . " ("
 272               . $@->code() . ") "
 273               . $@->addinfo() . " "
 274               . $@->diagset();
 275             warn $error." for query: $query";
 276             return ( $error, undef, undef );
 277         }
 278     }
 279
 280     _ZOOM_event_loop(
 281         \@zconns,
 282         \@tmpresults,
 283         sub {
 284             my ($i, $size) = @_;
 285             my $first_record = defined($offset) ? $offset + 1 : 1;
 286             my $hits = $tmpresults[ $i - 1 ]->size();
 287             $total_hits += $hits;
 288             my $last_record = $hits;
 289             if ( defined $max_results && $offset + $max_results < $hits ) {
 290                 $last_record = $offset + $max_results;
 291             }
 292
 293             for my $j ( $first_record .. $last_record ) {
 294                 my $record = eval {
 295                   $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
 296                   ;    # 0 indexed
 297                 };
 298                 push @{$results}, $record if defined $record;
 299             }
 300         }
 301     );
 302
 303     foreach my $zoom_query (@zoom_queries) {
 304         $zoom_query->destroy();
 305     }
 306
 307     return ( undef, $results, $total_hits );
 308 }
 309
 310 =head2 getRecords
 311
 312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
 313
 314         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 315         $results_per_page, $offset,       $expanded_facet, $branches,$itemtypes,
 316         $query_type,       $scan
 317     );
 318
 319 The all singing, all dancing, multi-server, asynchronous, scanning,
 320 searching, record nabbing, facet-building
 321
 322 See verbse embedded documentation.
 323
 324 =cut
 325
 326 sub getRecords {
 327     my (
 328         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 329         $results_per_page, $offset,       $expanded_facet, $branches,
 330         $itemtypes,        $query_type,   $scan,           $opac
 331     ) = @_;
 332
 333     my @servers = @$servers_ref;
 334     my @sort_by = @$sort_by_ref;
 335
 336     # Initialize variables for the ZOOM connection and results object
 337     my $zconn;
 338     my @zconns;
 339     my @results;
 340     my $results_hashref = ();
 341
 342     # Initialize variables for the faceted results objects
 343     my $facets_counter = {};
 344     my $facets_info    = {};
 345     my $facets         = getFacets();
 346
 347     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 348
 349     ### LOOP THROUGH THE SERVERS
 350     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 351         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 352
 353 # perform the search, create the results objects
 354 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 355         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 356
 357         #$query_to_use = $simple_query if $scan;
 358         warn $simple_query if ( $scan and $DEBUG );
 359
 360         # Check if we've got a query_type defined, if so, use it
 361         eval {
 362             if ($query_type) {
 363                 if ($query_type =~ /^ccl/) {
 364                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 365                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 366                 } elsif ($query_type =~ /^cql/) {
 367                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 368                 } elsif ($query_type =~ /^pqf/) {
 369                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 370                 } else {
 371                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 372                 }
 373             } elsif ($scan) {
 374                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 375             } else {
 376                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 377             }
 378         };
 379         if ($@) {
 380             warn "WARNING: query problem with $query_to_use " . $@;
 381         }
 382
 383         # Concatenate the sort_by limits and pass them to the results object
 384         # Note: sort will override rank
 385         my $sort_by;
 386         foreach my $sort (@sort_by) {
 387             if ( $sort eq "author_az" || $sort eq "author_asc" ) {
 388                 $sort_by .= "1=1003 <i ";
 389             }
 390             elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
 391                 $sort_by .= "1=1003 >i ";
 392             }
 393             elsif ( $sort eq "popularity_asc" ) {
 394                 $sort_by .= "1=9003 <i ";
 395             }
 396             elsif ( $sort eq "popularity_dsc" ) {
 397                 $sort_by .= "1=9003 >i ";
 398             }
 399             elsif ( $sort eq "call_number_asc" ) {
 400                 $sort_by .= "1=8007  <i ";
 401             }
 402             elsif ( $sort eq "call_number_dsc" ) {
 403                 $sort_by .= "1=8007 >i ";
 404             }
 405             elsif ( $sort eq "pubdate_asc" ) {
 406                 $sort_by .= "1=31 <i ";
 407             }
 408             elsif ( $sort eq "pubdate_dsc" ) {
 409                 $sort_by .= "1=31 >i ";
 410             }
 411             elsif ( $sort eq "acqdate_asc" ) {
 412                 $sort_by .= "1=32 <i ";
 413             }
 414             elsif ( $sort eq "acqdate_dsc" ) {
 415                 $sort_by .= "1=32 >i ";
 416             }
 417             elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
 418                 $sort_by .= "1=4 <i ";
 419             }
 420             elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
 421                 $sort_by .= "1=4 >i ";
 422             }
 423             else {
 424                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 425             }
 426         }
 427         if ( $sort_by && !$scan && $results[$i] ) {
 428             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 429                 warn "WARNING sort $sort_by failed";
 430             }
 431         }
 432     }    # finished looping through servers
 433
 434     # The big moment: asynchronously retrieve results from all servers
 435         _ZOOM_event_loop(
 436             \@zconns,
 437             \@results,
 438             sub {
 439                 my ( $i, $size ) = @_;
 440                 my $results_hash;
 441
 442                 # loop through the results
 443                 $results_hash->{'hits'} = $size;
 444                 my $times;
 445                 if ( $offset + $results_per_page <= $size ) {
 446                     $times = $offset + $results_per_page;
 447                 }
 448                 else {
 449                     $times = $size;
 450                 }
 451
 452                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 453                     my $records_hash;
 454                     my $record;
 455
 456                     ## Check if it's an index scan
 457                     if ($scan) {
 458                         my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
 459
 460                  # here we create a minimal MARC record and hand it off to the
 461                  # template just like a normal result ... perhaps not ideal, but
 462                  # it works for now
 463                         my $tmprecord = MARC::Record->new();
 464                         $tmprecord->encoding('UTF-8');
 465                         my $tmptitle;
 466                         my $tmpauthor;
 467
 468                 # the minimal record in author/title (depending on MARC flavour)
 469                         if ( C4::Context->preference("marcflavour") eq
 470                             "UNIMARC" )
 471                         {
 472                             $tmptitle = MARC::Field->new(
 473                                 '200', ' ', ' ',
 474                                 a => $term,
 475                                 f => $occ
 476                             );
 477                             $tmprecord->append_fields($tmptitle);
 478                         }
 479                         else {
 480                             $tmptitle =
 481                               MARC::Field->new( '245', ' ', ' ', a => $term, );
 482                             $tmpauthor =
 483                               MARC::Field->new( '100', ' ', ' ', a => $occ, );
 484                             $tmprecord->append_fields($tmptitle);
 485                             $tmprecord->append_fields($tmpauthor);
 486                         }
 487                         $results_hash->{'RECORDS'}[$j] =
 488                           $tmprecord->as_usmarc();
 489                     }
 490
 491                     # not an index scan
 492                     else {
 493                         $record = $results[ $i - 1 ]->record($j)->raw();
 494                         # warn "RECORD $j:".$record;
 495                         $results_hash->{'RECORDS'}[$j] = $record;
 496                     }
 497
 498                 }
 499                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 500
 501                 # Fill the facets while we're looping, but only for the
 502                 # biblioserver and not for a scan
 503                 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
 504                     $facets_counter = GetFacets( $results[ $i - 1 ] );
 505                     $facets_info    = _get_facets_info( $facets );
 506                 }
 507
 508                 # BUILD FACETS
 509                 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 510                     for my $link_value (
 511                         sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 512                         keys %$facets_counter
 513                       )
 514                     {
 515                         my $expandable;
 516                         my $number_of_facets;
 517                         my @this_facets_array;
 518                         for my $one_facet (
 519                             sort {
 520                                 $facets_counter->{$link_value}
 521                                   ->{$b} <=> $facets_counter->{$link_value}
 522                                   ->{$a}
 523                             } keys %{ $facets_counter->{$link_value} }
 524                           )
 525                         {
 526                             $number_of_facets++;
 527                             if (   ( $number_of_facets <= 5 )
 528                                 || ( $expanded_facet eq $link_value )
 529                                 || ( $facets_info->{$link_value}->{'expanded'} )
 530                               )
 531                             {
 532
 533 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
 534                                 my $facet_link_value = $one_facet;
 535                                 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
 536
 537                                 # fix the length that will display in the label,
 538                                 my $facet_label_value = $one_facet;
 539                                 my $facet_max_length  = C4::Context->preference(
 540                                     'FacetLabelTruncationLength')
 541                                   || 20;
 542                                 $facet_label_value =
 543                                   substr( $one_facet, 0, $facet_max_length )
 544                                   . "..."
 545                                   if length($facet_label_value) >
 546                                       $facet_max_length;
 547
 548                             # if it's a branch, label by the name, not the code,
 549                                 if ( $link_value =~ /branch/ ) {
 550                                     if (   defined $branches
 551                                         && ref($branches) eq "HASH"
 552                                         && defined $branches->{$one_facet}
 553                                         && ref( $branches->{$one_facet} ) eq
 554                                         "HASH" )
 555                                     {
 556                                         $facet_label_value =
 557                                           $branches->{$one_facet}
 558                                           ->{'branchname'};
 559                                     }
 560                                     else {
 561                                         $facet_label_value = "*";
 562                                     }
 563                                 }
 564
 565                           # if it's a itemtype, label by the name, not the code,
 566                                 if ( $link_value =~ /itype/ ) {
 567                                     if (   defined $itemtypes
 568                                         && ref($itemtypes) eq "HASH"
 569                                         && defined $itemtypes->{$one_facet}
 570                                         && ref( $itemtypes->{$one_facet} ) eq
 571                                         "HASH" )
 572                                     {
 573                                         $facet_label_value =
 574                                           $itemtypes->{$one_facet}
 575                                           ->{'description'};
 576                                     }
 577                                 }
 578
 579                # also, if it's a location code, use the name instead of the code
 580                                 if ( $link_value =~ /location/ ) {
 581                                     $facet_label_value =
 582                                       GetKohaAuthorisedValueLib( 'LOC',
 583                                         $one_facet, $opac );
 584                                 }
 585
 586                 # but we're down with the whole label being in the link's title.
 587                                 push @this_facets_array,
 588                                   {
 589                                     facet_count =>
 590                                       $facets_counter->{$link_value}
 591                                       ->{$one_facet},
 592                                     facet_label_value => $facet_label_value,
 593                                     facet_title_value => $one_facet,
 594                                     facet_link_value  => $facet_link_value,
 595                                     type_link_value   => $link_value,
 596                                   }
 597                                   if ($facet_label_value);
 598                             }
 599                         }
 600
 601                         # handle expanded option
 602                         unless ( $facets_info->{$link_value}->{'expanded'} ) {
 603                             $expandable = 1
 604                               if ( ( $number_of_facets > 5 )
 605                                 && ( $expanded_facet ne $link_value ) );
 606                         }
 607                         push @facets_loop,
 608                           {
 609                             type_link_value => $link_value,
 610                             type_id         => $link_value . "_id",
 611                             "type_label_"
 612                               . $facets_info->{$link_value}->{'label_value'} =>
 613                               1,
 614                             facets     => \@this_facets_array,
 615                             expandable => $expandable,
 616                             expand     => $link_value,
 617                           }
 618                           unless (
 619                             (
 620                                 $facets_info->{$link_value}->{'label_value'} =~
 621                                 /Libraries/
 622                             )
 623                             and ( C4::Context->preference('singleBranchMode') )
 624                           );
 625                     }
 626                 }
 627             }
 628         );
 629     return ( undef, $results_hashref, \@facets_loop );
 630 }
 631
 632 sub GetFacets {
 633
 634     my $rs = shift;
 635     my $facets;
 636
 637     my $indexing_mode    = C4::Context->config('zebra_bib_index_mode') // 'dom';
 638     my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
 639
 640     if ( $indexing_mode eq 'dom' &&
 641          $use_zebra_facets ) {
 642         $facets = _get_facets_from_zebra( $rs );
 643     } else {
 644         $facets = _get_facets_from_records( $rs );
 645     }
 646
 647     return $facets;
 648 }
 649
 650 sub _get_facets_from_records {
 651
 652     my $rs = shift;
 653
 654     my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
 655     my $facets_config  = getFacets();
 656     my $facets         = {};
 657     my $size           = $rs->size();
 658     my $jmax           = $size > $facets_maxrecs
 659                             ? $facets_maxrecs
 660                             : $size;
 661
 662     for ( my $j = 0 ; $j < $jmax ; $j++ ) {
 663
 664         my $marc_record = new_record_from_zebra (
 665                 'biblioserver',
 666                 $rs->record( $j )->raw()
 667         );
 668
 669         if ( ! defined $marc_record ) {
 670             warn "ERROR DECODING RECORD - $@: " .
 671                 $rs->record( $j )->raw();
 672             next;
 673         }
 674
 675         _get_facets_data_from_record( $marc_record, $facets_config, $facets );
 676     }
 677
 678     return $facets;
 679 }
 680
 681 =head2 _get_facets_data_from_record
 682
 683     C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
 684
 685 Internal function that extracts facets information from a MARC::Record object
 686 and populates $facets_counter for using in getRecords.
 687
 688 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
 689 facets for Zebra).
 690
 691 =cut
 692
 693 sub _get_facets_data_from_record {
 694
 695     my ( $marc_record, $facets, $facets_counter ) = @_;
 696
 697     for my $facet (@$facets) {
 698
 699         my @used_datas = ();
 700
 701         foreach my $tag ( @{ $facet->{ tags } } ) {
 702
 703             # tag number is the first three digits
 704             my $tag_num          = substr( $tag, 0, 3 );
 705             # subfields are the remainder
 706             my $subfield_letters = substr( $tag, 3 );
 707
 708             my @fields = $marc_record->field( $tag_num );
 709             foreach my $field (@fields) {
 710                 # If $field->indicator(1) eq 'z', it means it is a 'see from'
 711                 # field introduced because of IncludeSeeFromInSearches, so skip it
 712                 next if $field->indicator(1) eq 'z';
 713
 714                 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
 715
 716                 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
 717                     push @used_datas, $data;
 718                     $facets_counter->{ $facet->{ idx } }->{ $data }++;
 719                 }
 720             }
 721         }
 722     }
 723 }
 724
 725 =head2 _get_facets_from_zebra
 726
 727     my $facets = _get_facets_from_zebra( $result_set )
 728
 729 Retrieves facets for a specified result set. It loops through the facets defined
 730 in C4::Koha::getFacets and returns a hash with the following structure:
 731
 732    {  facet_idx => {
 733             facet_value => count
 734       },
 735       ...
 736    }
 737
 738 =cut
 739
 740 sub _get_facets_from_zebra {
 741
 742     my $rs = shift;
 743
 744     # save current elementSetName
 745     my $elementSetName = $rs->option( 'elementSetName' );
 746
 747     my $facets_loop = getFacets();
 748     my $facets_data  = {};
 749     # loop through defined facets and fill the facets hashref
 750     foreach my $facet ( @$facets_loop ) {
 751
 752         my $idx = $facet->{ idx };
 753         my $sep = $facet->{ sep };
 754         my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
 755         if ( $facet_values ) {
 756             # we've actually got a result
 757             $facets_data->{ $idx } = $facet_values;
 758         }
 759     }
 760     # set elementSetName to its previous value to avoid side effects
 761     $rs->option( elementSetName => $elementSetName );
 762
 763     return $facets_data;
 764 }
 765
 766 =head2 _get_facet_from_result_set
 767
 768     my $facet_values =
 769         C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
 770
 771 Internal function that extracts facet information for a specific index ($facet_idx) and
 772 returns a hash containing facet values and count:
 773
 774     {
 775         $facet_value => $count ,
 776         ...
 777     }
 778
 779 Warning: this function has the side effect of changing the elementSetName for the result
 780 set. It is a helper function for the main loop, which takes care of backing it up for
 781 restoring.
 782
 783 =cut
 784
 785 sub _get_facet_from_result_set {
 786
 787     my $facet_idx = shift;
 788     my $rs        = shift;
 789     my $sep       = shift;
 790
 791     my $internal_sep = '<*>';
 792
 793     return if ( ! defined $facet_idx || ! defined $rs );
 794     # zebra's facet element, untokenized index
 795     my $facet_element = 'zebra::facet::' . $facet_idx . ':0:100';
 796     # configure zebra results for retrieving the desired facet
 797     $rs->option( elementSetName => $facet_element );
 798     # get the facet record from result set
 799     my $facet = $rs->record( 0 )->raw;
 800     # if the facet has no restuls...
 801     return if !defined $facet;
 802     # TODO: benchmark DOM vs. SAX performance
 803     my $facet_dom = XML::LibXML->load_xml(
 804       string => ($facet)
 805     );
 806     my @terms = $facet_dom->getElementsByTagName('term');
 807     return if ! @terms;
 808
 809     my $facets = {};
 810     foreach my $term ( @terms ) {
 811         my $facet_value = $term->textContent;
 812         $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
 813         $facets->{ $facet_value } = $term->getAttribute( 'occur' );
 814     }
 815
 816     return $facets;
 817 }
 818
 819 =head2 _get_facets_info
 820
 821     my $facets_info = C4::Search::_get_facets_info( $facets )
 822
 823 Internal function that extracts facets information and properly builds
 824 the data structure needed to render facet labels.
 825
 826 =cut
 827
 828 sub _get_facets_info {
 829
 830     my $facets = shift;
 831
 832     my $facets_info = {};
 833
 834     for my $facet ( @$facets ) {
 835         $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
 836         $facets_info->{ $facet->{ idx } }->{ expanded }    = $facet->{ expanded };
 837     }
 838
 839     return $facets_info;
 840 }
 841
 842 sub pazGetRecords {
 843     my (
 844         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 845         $results_per_page, $offset,       $expanded_facet, $branches,
 846         $query_type,       $scan
 847     ) = @_;
 848
 849     my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
 850     $paz->init();
 851     $paz->search($simple_query);
 852     sleep 1;   # FIXME: WHY?
 853
 854     # do results
 855     my $results_hashref = {};
 856     my $stats = XMLin($paz->stat);
 857     my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
 858
 859     # for a grouped search result, the number of hits
 860     # is the number of groups returned; 'bib_hits' will have
 861     # the total number of bibs.
 862     $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
 863     $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
 864
 865     HIT: foreach my $hit (@{ $results->{'hit'} }) {
 866         my $recid = $hit->{recid}->[0];
 867
 868         my $work_title = $hit->{'md-work-title'}->[0];
 869         my $work_author;
 870         if (exists $hit->{'md-work-author'}) {
 871             $work_author = $hit->{'md-work-author'}->[0];
 872         }
 873         my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
 874
 875         my $result_group = {};
 876         $result_group->{'group_label'} = $group_label;
 877         $result_group->{'group_merge_key'} = $recid;
 878
 879         my $count = 1;
 880         if (exists $hit->{count}) {
 881             $count = $hit->{count}->[0];
 882         }
 883         $result_group->{'group_count'} = $count;
 884
 885         for (my $i = 0; $i < $count; $i++) {
 886             # FIXME -- may need to worry about diacritics here
 887             my $rec = $paz->record($recid, $i);
 888             push @{ $result_group->{'RECORDS'} }, $rec;
 889         }
 890
 891         push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
 892     }
 893
 894     # pass through facets
 895     my $termlist_xml = $paz->termlist('author,subject');
 896     my $terms = XMLin($termlist_xml, forcearray => 1);
 897     my @facets_loop = ();
 898     #die Dumper($results);
 899 #    foreach my $list (sort keys %{ $terms->{'list'} }) {
 900 #        my @facets = ();
 901 #        foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
 902 #            push @facets, {
 903 #                facet_label_value => $facet->{'name'}->[0],
 904 #            };
 905 #        }
 906 #        push @facets_loop, ( {
 907 #            type_label => $list,
 908 #            facets => \@facets,
 909 #        } );
 910 #    }
 911
 912     return ( undef, $results_hashref, \@facets_loop );
 913 }
 914
 915 # STOPWORDS
 916 sub _remove_stopwords {
 917     my ( $operand, $index ) = @_;
 918     my @stopwords_removed;
 919
 920     # phrase and exact-qualified indexes shouldn't have stopwords removed
 921     if ( $index !~ m/,(phr|ext)/ ) {
 922
 923 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
 924 #       we use IsAlpha unicode definition, to deal correctly with diacritics.
 925 #       otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
 926 #       is a stopword, we'd get "çon" and wouldn't find anything...
 927 #
 928                 foreach ( keys %{ C4::Context->stopwords } ) {
 929                         next if ( $_ =~ /(and|or|not)/ );    # don't remove operators
 930                         if ( my ($matched) = ($operand =~
 931                                 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
 932                         {
 933                                 $operand =~ s/\Q$matched\E/ /gi;
 934                                 push @stopwords_removed, $_;
 935                         }
 936                 }
 937         }
 938     return ( $operand, \@stopwords_removed );
 939 }
 940
 941 # TRUNCATION
 942 sub _detect_truncation {
 943     my ( $operand, $index ) = @_;
 944     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 945         @regexpr );
 946     $operand =~ s/^ //g;
 947     my @wordlist = split( /\s/, $operand );
 948     foreach my $word (@wordlist) {
 949         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 950             push @rightlefttruncated, $word;
 951         }
 952         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 953             push @lefttruncated, $word;
 954         }
 955         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 956             push @righttruncated, $word;
 957         }
 958         elsif ( index( $word, "*" ) < 0 ) {
 959             push @nontruncated, $word;
 960         }
 961         else {
 962             push @regexpr, $word;
 963         }
 964     }
 965     return (
 966         \@nontruncated,       \@righttruncated, \@lefttruncated,
 967         \@rightlefttruncated, \@regexpr
 968     );
 969 }
 970
 971 # STEMMING
 972 sub _build_stemmed_operand {
 973     my ($operand,$lang) = @_;
 974     require Lingua::Stem::Snowball ;
 975     my $stemmed_operand=q{};
 976
 977     # If operand contains a digit, it is almost certainly an identifier, and should
 978     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 979     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 980     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 981     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 982     return $operand if $operand =~ /\d/;
 983
 984 # FIXME: the locale should be set based on the user's language and/or search choice
 985     #warn "$lang";
 986     # Make sure we only use the first two letters from the language code
 987     $lang = lc(substr($lang, 0, 2));
 988     # The language codes for the two variants of Norwegian will now be "nb" and "nn",
 989     # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
 990     if ($lang eq 'nb' || $lang eq 'nn') {
 991       $lang = 'no';
 992     }
 993     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 994                                                encoding => "UTF-8" );
 995
 996     my @words = split( / /, $operand );
 997     my @stems = $stemmer->stem(\@words);
 998     for my $stem (@stems) {
 999         $stemmed_operand .= "$stem";
1000         $stemmed_operand .= "?"
1001           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
1002         $stemmed_operand .= " ";
1003     }
1004     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1005     return $stemmed_operand;
1006 }
1007
1008 # FIELD WEIGHTING
1009 sub _build_weighted_query {
1010
1011 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1012 # pretty well but could work much better if we had a smarter query parser
1013     my ( $operand, $stemmed_operand, $index ) = @_;
1014     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
1015     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1016     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
1017     $operand =~ s/"/ /g;    # Bug 7518: searches with quotation marks don't work
1018
1019     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
1020
1021     # Keyword, or, no index specified
1022     if ( ( $index eq 'kw' ) || ( !$index ) ) {
1023         $weighted_query .=
1024           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
1025         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
1026         $weighted_query .= " or Title-cover,phr,r3=\"$operand\"";    # phrase title
1027         $weighted_query .= " or ti,wrdl,r4=\"$operand\"";    # words in title
1028           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
1029           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
1030         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1031           if $fuzzy_enabled;    # add fuzzy, word list
1032         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1033           if ( $stemming and $stemmed_operand )
1034           ;                     # add stemming, right truncation
1035         $weighted_query .= " or wrdl,r9=\"$operand\"";
1036
1037         # embedded sorting: 0 a-z; 1 z-a
1038         # $weighted_query .= ") or (sort1,aut=1";
1039     }
1040
1041     # Barcode searches should skip this process
1042     elsif ( $index eq 'bc' ) {
1043         $weighted_query .= "bc=\"$operand\"";
1044     }
1045
1046     # Authority-number searches should skip this process
1047     elsif ( $index eq 'an' ) {
1048         $weighted_query .= "an=\"$operand\"";
1049     }
1050
1051     # If the index already has more than one qualifier, wrap the operand
1052     # in quotes and pass it back (assumption is that the user knows what they
1053     # are doing and won't appreciate us mucking up their query
1054     elsif ( $index =~ ',' ) {
1055         $weighted_query .= " $index=\"$operand\"";
1056     }
1057
1058     #TODO: build better cases based on specific search indexes
1059     else {
1060         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
1061           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1062         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
1063         $weighted_query .= " or $index,wrdl,r6=\"$operand\"";    # word list index
1064         $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1065           if $fuzzy_enabled;    # add fuzzy, word list
1066         $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1067           if ( $stemming and $stemmed_operand );    # add stemming, right truncation
1068     }
1069
1070     $weighted_query .= "))";                       # close rank specification
1071     return $weighted_query;
1072 }
1073
1074 =head2 getIndexes
1075
1076 Return an array with available indexes.
1077
1078 =cut
1079
1080 sub getIndexes{
1081     my @indexes = (
1082                     # biblio indexes
1083                     'ab',
1084                     'Abstract',
1085                     'acqdate',
1086                     'allrecords',
1087                     'an',
1088                     'Any',
1089                     'at',
1090                     'au',
1091                     'aub',
1092                     'aud',
1093                     'audience',
1094                     'auo',
1095                     'aut',
1096                     'Author',
1097                     'Author-in-order ',
1098                     'Author-personal-bibliography',
1099                     'Authority-Number',
1100                     'authtype',
1101                     'bc',
1102                     'Bib-level',
1103                     'biblionumber',
1104                     'bio',
1105                     'biography',
1106                     'callnum',
1107                     'cfn',
1108                     'Chronological-subdivision',
1109                     'cn-bib-source',
1110                     'cn-bib-sort',
1111                     'cn-class',
1112                     'cn-item',
1113                     'cn-prefix',
1114                     'cn-suffix',
1115                     'cpn',
1116                     'Code-institution',
1117                     'Conference-name',
1118                     'Conference-name-heading',
1119                     'Conference-name-see',
1120                     'Conference-name-seealso',
1121                     'Content-type',
1122                     'Control-number',
1123                     'copydate',
1124                     'Corporate-name',
1125                     'Corporate-name-heading',
1126                     'Corporate-name-see',
1127                     'Corporate-name-seealso',
1128                     'Country-publication',
1129                     'ctype',
1130                     'curriculum',
1131                     'date-entered-on-file',
1132                     'Date-of-acquisition',
1133                     'Date-of-publication',
1134                     'Dewey-classification',
1135                     'Dissertation-information',
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, $q, "q=ccl=".uri_escape($q), $q, '', '', '', '', 'ccl' );
1482     }
1483     if ( $query =~ /^cql=/ ) {
1484         return ( undef, $', $', "q=cql=".uri_escape($'), $', '', '', '', '', 'cql' );
1485     }
1486     if ( $query =~ /^pqf=/ ) {
1487         if ($query_desc) {
1488             $query_cgi = "q=".uri_escape($query_desc);
1489         } else {
1490             $query_desc = $';
1491             $query_cgi = "q=pqf=".uri_escape($');
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                 # Date of Publication
1541                 if ( $index eq 'yr' ) {
1542                     $index .= ",st-numeric";
1543                     $indexes_set++;
1544                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1545                 }
1546
1547                 # Date of Acquisition
1548                 elsif ( $index eq 'acqdate' ) {
1549                     $index .= ",st-date-normalized";
1550                     $indexes_set++;
1551                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1552                 }
1553                 # ISBN,ISSN,Standard Number, don't need special treatment
1554                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1555                     (
1556                         $stemming,      $auto_truncation,
1557                         $weight_fields, $fuzzy_enabled,
1558                         $remove_stopwords
1559                     ) = ( 0, 0, 0, 0, 0 );
1560
1561                 }
1562
1563                 if(not $index){
1564                     $index = 'kw';
1565                 }
1566
1567                 # Set default structure attribute (word list)
1568                 my $struct_attr = q{};
1569                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1570                     $struct_attr = ",wrdl";
1571                 }
1572
1573                 # Some helpful index variants
1574                 my $index_plus       = $index . $struct_attr . ':';
1575                 my $index_plus_comma = $index . $struct_attr . ',';
1576
1577                 # Remove Stopwords
1578                 if ($remove_stopwords) {
1579                     ( $operand, $stopwords_removed ) =
1580                       _remove_stopwords( $operand, $index );
1581                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1582                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1583                       if ( $stopwords_removed && $DEBUG );
1584                 }
1585
1586                 if ($auto_truncation){
1587                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1588                                                 #FIXME only valid with LTR scripts
1589                                                 $operand=join(" ",map{
1590                                                                                         (index($_,"*")>0?"$_":"$_*")
1591                                                                                          }split (/\s+/,$operand));
1592                                                 warn $operand if $DEBUG;
1593                                         }
1594                                 }
1595
1596                 # Detect Truncation
1597                 my $truncated_operand;
1598                 my( $nontruncated, $righttruncated, $lefttruncated,
1599                     $rightlefttruncated, $regexpr
1600                 ) = _detect_truncation( $operand, $index );
1601                 warn
1602 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1603                   if $DEBUG;
1604
1605                 # Apply Truncation
1606                 if (
1607                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1608                     scalar(@$rightlefttruncated) > 0 )
1609                 {
1610
1611                # Don't field weight or add the index to the query, we do it here
1612                     $indexes_set = 1;
1613                     undef $weight_fields;
1614                     my $previous_truncation_operand;
1615                     if (scalar @$nontruncated) {
1616                         $truncated_operand .= "$index_plus @$nontruncated ";
1617                         $previous_truncation_operand = 1;
1618                     }
1619                     if (scalar @$righttruncated) {
1620                         $truncated_operand .= "and " if $previous_truncation_operand;
1621                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1622                         $previous_truncation_operand = 1;
1623                     }
1624                     if (scalar @$lefttruncated) {
1625                         $truncated_operand .= "and " if $previous_truncation_operand;
1626                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1627                         $previous_truncation_operand = 1;
1628                     }
1629                     if (scalar @$rightlefttruncated) {
1630                         $truncated_operand .= "and " if $previous_truncation_operand;
1631                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1632                         $previous_truncation_operand = 1;
1633                     }
1634                 }
1635                 $operand = $truncated_operand if $truncated_operand;
1636                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1637
1638                 # Handle Stemming
1639                 my $stemmed_operand;
1640                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1641                                                                                 if $stemming;
1642
1643                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1644
1645                 # Handle Field Weighting
1646                 my $weighted_operand;
1647                 if ($weight_fields) {
1648                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1649                     $operand = $weighted_operand;
1650                     $indexes_set = 1;
1651                 }
1652
1653                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1654
1655                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1656                     query => $query,
1657                     query_cgi => $query_cgi,
1658                     query_desc => $query_desc,
1659                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1660                     parsed_operand => $operand,
1661                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1662                     index => $index,
1663                     index_plus => $index_plus,
1664                     indexes_set => $indexes_set,
1665                     previous_operand => $previous_operand,
1666                 });
1667
1668             }    #/if $operands
1669         }    # /for
1670     }
1671     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1672
1673     # add limits
1674     my %group_OR_limits;
1675     my $availability_limit;
1676     foreach my $this_limit (@limits) {
1677         next unless $this_limit;
1678         if ( $this_limit =~ /available/ ) {
1679 #
1680 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1681 ## In English:
1682 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1683             $availability_limit .=
1684 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1685             $limit_cgi  .= "&limit=available";
1686             $limit_desc .= "";
1687         }
1688
1689         # group_OR_limits, prefixed by mc-
1690         # OR every member of the group
1691         elsif ( $this_limit =~ /mc/ ) {
1692             my ($k,$v) = split(/:/, $this_limit,2);
1693             if ( $k !~ /mc-i(tem)?type/ ) {
1694                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1695                 $this_limit =~ tr/"//d;
1696                 $this_limit = $k.":\"".$v."\"";
1697             }
1698
1699             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1700             $limit_desc      .= " or " if $group_OR_limits{$k};
1701             $group_OR_limits{$k} .= "$this_limit";
1702             $limit_cgi       .= "&limit=" . uri_escape($this_limit);
1703             $limit_desc      .= " $this_limit";
1704         }
1705
1706         # Regular old limits
1707         else {
1708             $limit .= " and " if $limit || $query;
1709             $limit      .= "$this_limit";
1710             $limit_cgi  .= "&limit=" . uri_escape($this_limit);
1711             if ($this_limit =~ /^branch:(.+)/) {
1712                 my $branchcode = $1;
1713                 my $branchname = GetBranchName($branchcode);
1714                 if (defined $branchname) {
1715                     $limit_desc .= " branch:$branchname";
1716                 } else {
1717                     $limit_desc .= " $this_limit";
1718                 }
1719             } else {
1720                 $limit_desc .= " $this_limit";
1721             }
1722         }
1723     }
1724     foreach my $k (keys (%group_OR_limits)) {
1725         $limit .= " and " if ( $query || $limit );
1726         $limit .= "($group_OR_limits{$k})";
1727     }
1728     if ($availability_limit) {
1729         $limit .= " and " if ( $query || $limit );
1730         $limit .= "($availability_limit)";
1731     }
1732
1733     # Normalize the query and limit strings
1734     # This is flawed , means we can't search anything with : in it
1735     # if user wants to do ccl or cql, start the query with that
1736 #    $query =~ s/:/=/g;
1737     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1738     $query =~ s/(?<=(wrdl)):/=/g;
1739     $query =~ s/(?<=(trn|phr)):/=/g;
1740     $limit =~ s/:/=/g;
1741     for ( $query, $query_desc, $limit, $limit_desc ) {
1742         s/  +/ /g;    # remove extra spaces
1743         s/^ //g;     # remove any beginning spaces
1744         s/ $//g;     # remove any ending spaces
1745         s/==/=/g;    # remove double == from query
1746     }
1747     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1748
1749     for ($query_cgi,$simple_query) {
1750         s/"//g;
1751     }
1752     # append the limit to the query
1753     $query .= " " . $limit;
1754
1755     # Warnings if DEBUG
1756     if ($DEBUG) {
1757         warn "QUERY:" . $query;
1758         warn "QUERY CGI:" . $query_cgi;
1759         warn "QUERY DESC:" . $query_desc;
1760         warn "LIMIT:" . $limit;
1761         warn "LIMIT CGI:" . $limit_cgi;
1762         warn "LIMIT DESC:" . $limit_desc;
1763         warn "---------\nLeave buildQuery\n---------";
1764     }
1765     return (
1766         undef,              $query, $simple_query, $query_cgi,
1767         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1768         $stopwords_removed, $query_type
1769     );
1770 }
1771
1772 =head2 _build_initial_query
1773
1774   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1775
1776   Build a section of the initial query containing indexes, operators, and operands.
1777
1778 =cut
1779
1780 sub _build_initial_query {
1781     my ($params) = @_;
1782
1783     my $operator = "";
1784     if ($params->{previous_operand}){
1785         #If there is a previous operand, add a supplied operator or the default 'and'
1786         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1787     }
1788
1789     #NOTE: indexes_set is typically set when doing truncation or field weighting
1790     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1791
1792     #e.g. "kw,wrdl:test"
1793     #e.g. " and kw,wrdl:test"
1794     $params->{query} .= $operator . $operand;
1795
1796     $params->{query_cgi} .= "&op=".uri_escape($operator) if $operator;
1797     $params->{query_cgi} .= "&idx=".uri_escape($params->{index}) if $params->{index};
1798     $params->{query_cgi} .= "&q=".uri_escape($params->{original_operand}) if $params->{original_operand};
1799
1800     #e.g. " and kw,wrdl: test"
1801     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1802
1803     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1804
1805     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1806 }
1807
1808 =head2 searchResults
1809
1810   my @search_results = searchResults($search_context, $searchdesc, $hits,
1811                                      $results_per_page, $offset, $scan,
1812                                      @marcresults);
1813
1814 Format results in a form suitable for passing to the template
1815
1816 =cut
1817
1818 # IMO this subroutine is pretty messy still -- it's responsible for
1819 # building the HTML output for the template
1820 sub searchResults {
1821     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1822     my $dbh = C4::Context->dbh;
1823     my @newresults;
1824
1825     require C4::Items;
1826
1827     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1828     my ($is_opac, $hidelostitems);
1829     if ($search_context eq 'opac') {
1830         $hidelostitems = C4::Context->preference('hidelostitems');
1831         $is_opac       = 1;
1832     }
1833
1834     #Build branchnames hash
1835     #find branchname
1836     #get branch information.....
1837     my %branches;
1838     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1839     $bsth->execute();
1840     while ( my $bdata = $bsth->fetchrow_hashref ) {
1841         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1842     }
1843 # FIXME - We build an authorised values hash here, using the default framework
1844 # though it is possible to have different authvals for different fws.
1845
1846     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1847
1848     # get notforloan authorised value list (see $shelflocations  FIXME)
1849     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1850
1851     #Build itemtype hash
1852     #find itemtype & itemtype image
1853     my %itemtypes;
1854     $bsth =
1855       $dbh->prepare(
1856         "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1857       );
1858     $bsth->execute();
1859     while ( my $bdata = $bsth->fetchrow_hashref ) {
1860                 foreach (qw(description imageurl summary notforloan)) {
1861                 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1862                 }
1863     }
1864
1865     #search item field code
1866     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1867
1868     ## find column names of items related to MARC
1869     my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1870     $sth2->execute;
1871     my %subfieldstosearch;
1872     while ( ( my $column ) = $sth2->fetchrow ) {
1873         my ( $tagfield, $tagsubfield ) =
1874           &GetMarcFromKohaField( "items." . $column, "" );
1875         if ( defined $tagsubfield ) {
1876             $subfieldstosearch{$column} = $tagsubfield;
1877         }
1878     }
1879
1880     # handle which records to actually retrieve
1881     my $times;
1882     if ( $hits && $offset + $results_per_page <= $hits ) {
1883         $times = $offset + $results_per_page;
1884     }
1885     else {
1886         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1887     }
1888
1889     my $marcflavour = C4::Context->preference("marcflavour");
1890     # We get the biblionumber position in MARC
1891     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1892
1893     # loop through all of the records we've retrieved
1894     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1895
1896         my $marcrecord;
1897         if ($scan) {
1898             # For Scan searches we built USMARC data
1899             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1900         } else {
1901             # Normal search, render from Zebra's output
1902             $marcrecord = new_record_from_zebra(
1903                 'biblioserver',
1904                 $marcresults->[$i]
1905             );
1906
1907             if ( ! defined $marcrecord ) {
1908                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1909                 next;
1910             }
1911         }
1912
1913         my $fw = $scan
1914              ? undef
1915              : $bibliotag < 10
1916                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1917                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1918         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1919         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1920         $oldbiblio->{result_number} = $i + 1;
1921
1922         # add imageurl to itemtype if there is one
1923         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1924
1925         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1926                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1927                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1928                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1929                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1930                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1931
1932                 # edition information, if any
1933         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1934                 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1935  # Build summary if there is one (the summary is defined in the itemtypes table)
1936  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1937         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1938             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1939             my @fields  = $marcrecord->fields();
1940
1941             my $newsummary;
1942             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1943                 my $tags = {};
1944                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1945                     $tag =~ /(.{3})(.)/;
1946                     if($marcrecord->field($1)){
1947                         my @abc = $marcrecord->field($1)->subfield($2);
1948                         $tags->{$tag} = $#abc + 1 ;
1949                     }
1950                 }
1951
1952                 # We catch how many times to repeat this line
1953                 my $max = 0;
1954                 foreach my $tag (keys(%$tags)){
1955                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1956                  }
1957
1958                 # we replace, and repeat each line
1959                 for (my $i = 0 ; $i < $max ; $i++){
1960                     my $newline = $line;
1961
1962                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1963                         $tag =~ /(.{3})(.)/;
1964
1965                         if($marcrecord->field($1)){
1966                             my @repl = $marcrecord->field($1)->subfield($2);
1967                             my $subfieldvalue = $repl[$i];
1968
1969                             if (! utf8::is_utf8($subfieldvalue)) {
1970                                 utf8::decode($subfieldvalue);
1971                             }
1972
1973                              $newline =~ s/\[$tag\]/$subfieldvalue/g;
1974                         }
1975                     }
1976                     $newsummary .= "$newline\n";
1977                 }
1978             }
1979
1980             $newsummary =~ s/\[(.*?)]//g;
1981             $newsummary =~ s/\n/<br\/>/g;
1982             $oldbiblio->{summary} = $newsummary;
1983         }
1984
1985         # Pull out the items fields
1986         my @fields = $marcrecord->field($itemtag);
1987         my $marcflavor = C4::Context->preference("marcflavour");
1988         # adding linked items that belong to host records
1989         my $analyticsfield = '773';
1990         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1991             $analyticsfield = '773';
1992         } elsif ($marcflavor eq 'UNIMARC') {
1993             $analyticsfield = '461';
1994         }
1995         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1996             my $hostbiblionumber = $hostfield->subfield("0");
1997             my $linkeditemnumber = $hostfield->subfield("9");
1998             if(!$hostbiblionumber eq undef){
1999                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2000                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2001                 if(!$hostbiblio eq undef){
2002                     my @hostitems = $hostbiblio->field($itemfield);
2003                     foreach my $hostitem (@hostitems){
2004                         if ($hostitem->subfield("9") eq $linkeditemnumber){
2005                             my $linkeditem =$hostitem;
2006                             # append linked items if they exist
2007                             if (!$linkeditem eq undef){
2008                                 push (@fields, $linkeditem);}
2009                         }
2010                     }
2011                 }
2012             }
2013         }
2014
2015         # Setting item statuses for display
2016         my @available_items_loop;
2017         my @onloan_items_loop;
2018         my @other_items_loop;
2019
2020         my $available_items;
2021         my $onloan_items;
2022         my $other_items;
2023
2024         my $ordered_count         = 0;
2025         my $available_count       = 0;
2026         my $onloan_count          = 0;
2027         my $longoverdue_count     = 0;
2028         my $other_count           = 0;
2029         my $withdrawn_count        = 0;
2030         my $itemlost_count        = 0;
2031         my $hideatopac_count      = 0;
2032         my $itembinding_count     = 0;
2033         my $itemdamaged_count     = 0;
2034         my $item_in_transit_count = 0;
2035         my $can_place_holds       = 0;
2036         my $item_onhold_count     = 0;
2037         my $notforloan_count      = 0;
2038         my $items_count           = scalar(@fields);
2039         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2040         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2041         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2042
2043         # loop through every item
2044         foreach my $field (@fields) {
2045             my $item;
2046
2047             # populate the items hash
2048             foreach my $code ( keys %subfieldstosearch ) {
2049                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2050             }
2051             $item->{description} = $itemtypes{ $item->{itype} }{description};
2052
2053                 # OPAC hidden items
2054             if ($is_opac) {
2055                 # hidden because lost
2056                 if ($hidelostitems && $item->{itemlost}) {
2057                     $hideatopac_count++;
2058                     next;
2059                 }
2060                 # hidden based on OpacHiddenItems syspref
2061                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2062                 if (scalar @hi) {
2063                     push @hiddenitems, @hi;
2064                     $hideatopac_count++;
2065                     next;
2066                 }
2067             }
2068
2069             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2070             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2071
2072             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2073             if ($item->{$hbranch}) {
2074                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2075             }
2076             elsif ($item->{$otherbranch}) {     # Last resort
2077                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2078             }
2079
2080                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2081 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2082             my $userenv = C4::Context->userenv;
2083             if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
2084                 $onloan_count++;
2085                                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2086                                 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
2087                                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2088                                 $onloan_items->{$key}->{branchname} = $item->{branchname};
2089                                 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2090                                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2091                                 $onloan_items->{$key}->{description} = $item->{description};
2092                                 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2093                 # if something's checked out and lost, mark it as 'long overdue'
2094                 if ( $item->{itemlost} ) {
2095                     $onloan_items->{$prefix}->{longoverdue}++;
2096                     $longoverdue_count++;
2097                 } else {        # can place holds as long as item isn't lost
2098                     $can_place_holds = 1;
2099                 }
2100             }
2101
2102          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2103             else {
2104
2105                 # item is on order
2106                 if ( $item->{notforloan} < 0 ) {
2107                     $ordered_count++;
2108                 } elsif ( $item->{notforloan} > 0 ) {
2109                     $notforloan_count++;
2110                 }
2111
2112                 # is item in transit?
2113                 my $transfertwhen = '';
2114                 my ($transfertfrom, $transfertto);
2115
2116                 # is item on the reserve shelf?
2117                 my $reservestatus = '';
2118
2119                 unless ($item->{withdrawn}
2120                         || $item->{itemlost}
2121                         || $item->{damaged}
2122                         || $item->{notforloan}
2123                         || $items_count > 20) {
2124
2125                     # A couple heuristics to limit how many times
2126                     # we query the database for item transfer information, sacrificing
2127                     # accuracy in some cases for speed;
2128                     #
2129                     # 1. don't query if item has one of the other statuses
2130                     # 2. don't check transit status if the bib has
2131                     #    more than 20 items
2132                     #
2133                     # FIXME: to avoid having the query the database like this, and to make
2134                     #        the in transit status count as unavailable for search limiting,
2135                     #        should map transit status to record indexed in Zebra.
2136                     #
2137                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2138                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber}, $oldbiblio->{biblionumber} );
2139                 }
2140
2141                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2142                 if (   $item->{withdrawn}
2143                     || $item->{itemlost}
2144                     || $item->{damaged}
2145                     || $item->{notforloan}
2146                     || $reservestatus eq 'Waiting'
2147                     || ($transfertwhen ne ''))
2148                 {
2149                     $withdrawn_count++        if $item->{withdrawn};
2150                     $itemlost_count++        if $item->{itemlost};
2151                     $itemdamaged_count++     if $item->{damaged};
2152                     $item_in_transit_count++ if $transfertwhen ne '';
2153                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2154                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2155
2156                     # can place a hold on a item if
2157                     # not lost nor withdrawn
2158                     # not damaged unless AllowHoldsOnDamagedItems is true
2159                     # item is either for loan or on order (notforloan < 0)
2160                     $can_place_holds = 1
2161                       if (
2162                            !$item->{itemlost}
2163                         && !$item->{withdrawn}
2164                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2165                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2166                       );
2167
2168                     $other_count++;
2169
2170                     my $key = $prefix . $item->{status};
2171                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2172                         $other_items->{$key}->{$_} = $item->{$_};
2173                     }
2174                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2175                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2176                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2177                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2178                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2179                                         $other_items->{$key}->{description} = $item->{description};
2180                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2181                 }
2182                 # item is available
2183                 else {
2184                     $can_place_holds = 1;
2185                     $available_count++;
2186                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2187                                         foreach (qw(branchname itemcallnumber description)) {
2188                         $available_items->{$prefix}->{$_} = $item->{$_};
2189                                         }
2190                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2191                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2192                 }
2193             }
2194         }    # notforloan, item level and biblioitem level
2195
2196         # if all items are hidden, do not show the record
2197         if ($items_count > 0 && $hideatopac_count == $items_count) {
2198             next;
2199         }
2200
2201         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2202         for my $key ( sort keys %$onloan_items ) {
2203             (++$onloanitemscount > $maxitems) and last;
2204             push @onloan_items_loop, $onloan_items->{$key};
2205         }
2206         for my $key ( sort keys %$other_items ) {
2207             (++$otheritemscount > $maxitems) and last;
2208             push @other_items_loop, $other_items->{$key};
2209         }
2210         for my $key ( sort keys %$available_items ) {
2211             (++$availableitemscount > $maxitems) and last;
2212             push @available_items_loop, $available_items->{$key}
2213         }
2214
2215         # XSLT processing of some stuff
2216         SetUTF8Flag($marcrecord);
2217         warn $marcrecord->as_formatted if $DEBUG;
2218         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2219         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2220             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2221         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2222         }
2223
2224         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2225         if (!C4::Context->preference("item-level_itypes")) {
2226             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2227                 $can_place_holds = 0;
2228             }
2229         }
2230         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2231         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2232         $oldbiblio->{items_count}          = $items_count;
2233         $oldbiblio->{available_items_loop} = \@available_items_loop;
2234         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2235         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2236         $oldbiblio->{availablecount}       = $available_count;
2237         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2238         $oldbiblio->{onloancount}          = $onloan_count;
2239         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2240         $oldbiblio->{othercount}           = $other_count;
2241         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2242         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2243         $oldbiblio->{itemlostcount}        = $itemlost_count;
2244         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2245         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2246         $oldbiblio->{onholdcount}          = $item_onhold_count;
2247         $oldbiblio->{orderedcount}         = $ordered_count;
2248         $oldbiblio->{notforloancount}      = $notforloan_count;
2249
2250         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2251             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2252             my $subfields = substr $fieldspec, 3;
2253             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2254             my @alternateholdingsinfo = ();
2255             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2256             my $alternateholdingscount = 0;
2257
2258             for my $field (@holdingsfields) {
2259                 my %holding = ( holding => '' );
2260                 my $havesubfield = 0;
2261                 for my $subfield ($field->subfields()) {
2262                     if ((index $subfields, $$subfield[0]) >= 0) {
2263                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2264                         $holding{'holding'} .= $$subfield[1];
2265                         $havesubfield++;
2266                     }
2267                 }
2268                 if ($havesubfield) {
2269                     push(@alternateholdingsinfo, \%holding);
2270                     $alternateholdingscount++;
2271                 }
2272             }
2273
2274             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2275             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2276         }
2277
2278         push( @newresults, $oldbiblio );
2279     }
2280
2281     return @newresults;
2282 }
2283
2284 =head2 SearchAcquisitions
2285     Search for acquisitions
2286 =cut
2287
2288 sub SearchAcquisitions{
2289     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2290
2291     my $dbh=C4::Context->dbh;
2292     # Variable initialization
2293     my $str=qq|
2294     SELECT marcxml
2295     FROM biblio
2296     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2297     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2298     WHERE dateaccessioned BETWEEN ? AND ?
2299     |;
2300
2301     my (@params,@loopcriteria);
2302
2303     push @params, $datebegin->output("iso");
2304     push @params, $dateend->output("iso");
2305
2306     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2307         if(C4::Context->preference("item-level_itypes")){
2308             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2309         }else{
2310             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2311         }
2312         push @params, @$itemtypes;
2313     }
2314
2315     if ($criteria =~/itemtype/){
2316         if(C4::Context->preference("item-level_itypes")){
2317             $str .= "AND items.itype=? ";
2318         }else{
2319             $str .= "AND biblioitems.itemtype=? ";
2320         }
2321
2322         if(scalar(@$itemtypes) == 0){
2323             my $itypes = GetItemTypes();
2324             for my $key (keys %$itypes){
2325                 push @$itemtypes, $key;
2326             }
2327         }
2328
2329         @loopcriteria= @$itemtypes;
2330     }elsif ($criteria=~/itemcallnumber/){
2331         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2332                  OR items.itemcallnumber is NULL
2333                  OR items.itemcallnumber = '')";
2334
2335         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2336     }else {
2337         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2338         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2339     }
2340
2341     if ($orderby =~ /date_desc/){
2342         $str.=" ORDER BY dateaccessioned DESC";
2343     } else {
2344         $str.=" ORDER BY title";
2345     }
2346
2347     my $qdataacquisitions=$dbh->prepare($str);
2348
2349     my @loopacquisitions;
2350     foreach my $value(@loopcriteria){
2351         push @params,$value;
2352         my %cell;
2353         $cell{"title"}=$value;
2354         $cell{"titlecode"}=$value;
2355
2356         eval{$qdataacquisitions->execute(@params);};
2357
2358         if ($@){ warn "recentacquisitions Error :$@";}
2359         else {
2360             my @loopdata;
2361             while (my $data=$qdataacquisitions->fetchrow_hashref){
2362                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2363             }
2364             $cell{"loopdata"}=\@loopdata;
2365         }
2366         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2367         pop @params;
2368     }
2369     $qdataacquisitions->finish;
2370     return \@loopacquisitions;
2371 }
2372
2373 =head2 enabled_staff_search_views
2374
2375 %hash = enabled_staff_search_views()
2376
2377 This function returns a hash that contains three flags obtained from the system
2378 preferences, used to determine whether a particular staff search results view
2379 is enabled.
2380
2381 =over 2
2382
2383 =item C<Output arg:>
2384
2385     * $hash{can_view_MARC} is true only if the MARC view is enabled
2386     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2387     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2388
2389 =item C<usage in the script:>
2390
2391 =back
2392
2393 $template->param ( C4::Search::enabled_staff_search_views );
2394
2395 =cut
2396
2397 sub enabled_staff_search_views
2398 {
2399         return (
2400                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2401                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2402                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2403         );
2404 }
2405
2406 sub PurgeSearchHistory{
2407     my ($pSearchhistory)=@_;
2408     my $dbh = C4::Context->dbh;
2409     my $sth = $dbh->prepare("DELETE FROM search_history WHERE time < DATE_SUB( NOW(), INTERVAL ? DAY )");
2410     $sth->execute($pSearchhistory) or die $dbh->errstr;
2411 }
2412
2413 =head2 z3950_search_args
2414
2415 $arrayref = z3950_search_args($matchpoints)
2416
2417 This function returns an array reference that contains the search parameters to be
2418 passed to the Z39.50 search script (z3950_search.pl). The array elements
2419 are hash refs whose keys are name and value, and whose values are the
2420 name of a search parameter, the value of that search parameter and the URL encoded
2421 value of that parameter.
2422
2423 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2424
2425 The search parameter values are obtained from the bibliographic record whose
2426 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2427
2428 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2429 a general purpose search argument. In this case, the returned array contains only
2430 entry: the key is 'title' and the value is derived from $matchpoints.
2431
2432 If a search parameter value is undefined or empty, it is not included in the returned
2433 array.
2434
2435 The returned array reference may be passed directly to the template parameters.
2436
2437 =over 2
2438
2439 =item C<Output arg:>
2440
2441     * $array containing hash refs as described above
2442
2443 =item C<usage in the script:>
2444
2445 =back
2446
2447 $data = Biblio::GetBiblioData($bibno);
2448 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2449
2450 *OR*
2451
2452 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2453
2454 =cut
2455
2456 sub z3950_search_args {
2457     my $bibrec = shift;
2458
2459     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2460     my $isbn = Business::ISBN->new( $isbn_string );
2461
2462     if (defined $isbn && $isbn->is_valid)
2463     {
2464         if ( ref($bibrec) ) {
2465             $bibrec->{isbn} = $isbn_string;
2466             $bibrec->{title} = undef;
2467         } else {
2468             $bibrec = { isbn => $isbn_string };
2469         }
2470     }
2471     else {
2472         $bibrec = { title => $bibrec } if !ref $bibrec;
2473     }
2474     my $array = [];
2475     for my $field (qw/ lccn isbn issn title author dewey subject /)
2476     {
2477         push @$array, { name => $field, value => $bibrec->{$field} }
2478           if defined $bibrec->{$field};
2479     }
2480     return $array;
2481 }
2482
2483 =head2 GetDistinctValues($field);
2484
2485 C<$field> is a reference to the fields array
2486
2487 =cut
2488
2489 sub GetDistinctValues {
2490     my ($fieldname,$string)=@_;
2491     # returns a reference to a hash of references to branches...
2492     if ($fieldname=~/\./){
2493                         my ($table,$column)=split /\./, $fieldname;
2494                         my $dbh = C4::Context->dbh;
2495                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2496                         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 ");
2497                         $sth->execute;
2498                         my $elements=$sth->fetchall_arrayref({});
2499                         return $elements;
2500    }
2501    else {
2502                 $string||= qq("");
2503                 my @servers=qw<biblioserver authorityserver>;
2504                 my (@zconns,@results);
2505         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2506                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2507                         $results[$i] =
2508                       $zconns[$i]->scan(
2509                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2510                       );
2511                 }
2512                 # The big moment: asynchronously retrieve results from all servers
2513                 my @elements;
2514         _ZOOM_event_loop(
2515             \@zconns,
2516             \@results,
2517             sub {
2518                 my ( $i, $size ) = @_;
2519                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2520                     my %hashscan;
2521                     @hashscan{qw(value cnt)} =
2522                       $results[ $i - 1 ]->display_term($j);
2523                     push @elements, \%hashscan;
2524                 }
2525             }
2526         );
2527                 return \@elements;
2528    }
2529 }
2530
2531 =head2 _ZOOM_event_loop
2532
2533     _ZOOM_event_loop(\@zconns, \@results, sub {
2534         my ( $i, $size ) = @_;
2535         ....
2536     } );
2537
2538 Processes a ZOOM event loop and passes control to a closure for
2539 processing the results, and destroying the resultsets.
2540
2541 =cut
2542
2543 sub _ZOOM_event_loop {
2544     my ($zconns, $results, $callback) = @_;
2545     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2546         my $ev = $zconns->[ $i - 1 ]->last_event();
2547         if ( $ev == ZOOM::Event::ZEND ) {
2548             next unless $results->[ $i - 1 ];
2549             my $size = $results->[ $i - 1 ]->size();
2550             if ( $size > 0 ) {
2551                 $callback->($i, $size);
2552             }
2553         }
2554     }
2555
2556     foreach my $result (@$results) {
2557         $result->destroy();
2558     }
2559 }
2560
2561 =head2 new_record_from_zebra
2562
2563 Given raw data from a Zebra result set, return a MARC::Record object
2564
2565 This helper function is needed to take into account all the involved
2566 system preferences and configuration variables to properly create the
2567 MARC::Record object.
2568
2569 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2570 data. If we are using DOM, then it has to be MARCXML.
2571
2572 =cut
2573
2574 sub new_record_from_zebra {
2575
2576     my $server   = shift;
2577     my $raw_data = shift;
2578     # Set the default indexing modes
2579     my $index_mode = ( $server eq 'biblioserver' )
2580                         ? C4::Context->config('zebra_bib_index_mode') // 'grs1'
2581                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2582
2583     my $marc_record =  eval {
2584         if ( $index_mode eq 'dom' ) {
2585             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2586         } else {
2587             MARC::Record->new_from_usmarc( $raw_data );
2588         }
2589     };
2590
2591     if ($@) {
2592         return;
2593     } else {
2594         return $marc_record;
2595     }
2596
2597 }
2598
2599 END { }    # module clean-up code here (global destructor)
2600
2601 1;
2602 __END__
2603
2604 =head1 AUTHOR
2605
2606 Koha Development Team <http://koha-community.org/>
2607
2608 =cut