C4/Search.pm

   1 package C4::Search;
   2
   3 # This file is part of Koha.
   4 #
   5 # Koha is free software; you can redistribute it and/or modify it
   6 # under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 3 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # Koha is distributed in the hope that it will be useful, but
  11 # WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License
  16 # along with Koha; if not, see <http://www.gnu.org/licenses>.
  17
  18 use strict;
  19 #use warnings; FIXME - Bug 2505
  20 require Exporter;
  21 use C4::Context;
  22 use C4::Biblio;    # GetMarcFromKohaField, GetBiblioData
  23 use C4::Koha;      # getFacets
  24 use Lingua::Stem;
  25 use C4::Search::PazPar2;
  26 use XML::Simple;
  27 use C4::Dates qw(format_date);
  28 use C4::Members qw(GetHideLostItemsPreference);
  29 use C4::XSLT;
  30 use C4::Branch;
  31 use C4::Reserves;    # GetReserveStatus
  32 use C4::Debug;
  33 use C4::Charset;
  34 use YAML;
  35 use URI::Escape;
  36 use Business::ISBN;
  37 use MARC::Record;
  38 use MARC::Field;
  39 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  40
  41 # set the version for version checking
  42 BEGIN {
  43     $VERSION = 3.07.00.049;
  44     $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
  45 }
  46
  47 =head1 NAME
  48
  49 C4::Search - Functions for searching the Koha catalog.
  50
  51 =head1 SYNOPSIS
  52
  53 See opac/opac-search.pl or catalogue/search.pl for example of usage
  54
  55 =head1 DESCRIPTION
  56
  57 This module provides searching functions for Koha's bibliographic databases
  58
  59 =head1 FUNCTIONS
  60
  61 =cut
  62
  63 @ISA    = qw(Exporter);
  64 @EXPORT = qw(
  65   &FindDuplicate
  66   &SimpleSearch
  67   &searchResults
  68   &getRecords
  69   &buildQuery
  70   &GetDistinctValues
  71   &enabled_staff_search_views
  72   &PurgeSearchHistory
  73 );
  74
  75 # make all your functions, whether exported or not;
  76
  77 =head2 FindDuplicate
  78
  79 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
  80
  81 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
  82
  83 =cut
  84
  85 sub FindDuplicate {
  86     my ($record) = @_;
  87     my $dbh = C4::Context->dbh;
  88     my $result = TransformMarcToKoha( $dbh, $record, '' );
  89     my $sth;
  90     my $query;
  91     my $search;
  92     my $type;
  93     my ( $biblionumber, $title );
  94
  95     # search duplicate on ISBN, easy and fast..
  96     # ... normalize first
  97     if ( $result->{isbn} ) {
  98         $result->{isbn} =~ s/\(.*$//;
  99         $result->{isbn} =~ s/\s+$//;
 100         $query = "isbn:$result->{isbn}";
 101     }
 102     else {
 103         my $QParser;
 104         $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
 105         my $titleindex;
 106         my $authorindex;
 107         my $op;
 108
 109         if ($QParser) {
 110             $titleindex = 'title|exact';
 111             $authorindex = 'author|exact';
 112             $op = '&&';
 113             $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 114         } else {
 115             $titleindex = 'ti,ext';
 116             $authorindex = 'au,ext';
 117             $op = 'and';
 118         }
 119
 120         $result->{title} =~ s /\\//g;
 121         $result->{title} =~ s /\"//g;
 122         $result->{title} =~ s /\(//g;
 123         $result->{title} =~ s /\)//g;
 124
 125         # FIXME: instead of removing operators, could just do
 126         # quotes around the value
 127         $result->{title} =~ s/(and|or|not)//g;
 128         $query = "$titleindex:\"$result->{title}\"";
 129         if   ( $result->{author} ) {
 130             $result->{author} =~ s /\\//g;
 131             $result->{author} =~ s /\"//g;
 132             $result->{author} =~ s /\(//g;
 133             $result->{author} =~ s /\)//g;
 134
 135             # remove valid operators
 136             $result->{author} =~ s/(and|or|not)//g;
 137             $query .= " $op $authorindex:\"$result->{author}\"";
 138         }
 139     }
 140
 141     my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
 142     my @results;
 143     if (!defined $error) {
 144         foreach my $possible_duplicate_record (@{$searchresults}) {
 145             my $marcrecord = new_record_from_zebra(
 146                 'biblioserver',
 147                 $possible_duplicate_record
 148             );
 149
 150             my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
 151
 152             # FIXME :: why 2 $biblionumber ?
 153             if ($result) {
 154                 push @results, $result->{'biblionumber'};
 155                 push @results, $result->{'title'};
 156             }
 157         }
 158     }
 159     return @results;
 160 }
 161
 162 =head2 SimpleSearch
 163
 164 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
 165
 166 This function provides a simple search API on the bibliographic catalog
 167
 168 =over 2
 169
 170 =item C<input arg:>
 171
 172     * $query can be a simple keyword or a complete CCL query
 173     * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
 174     * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0
 175     * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
 176
 177
 178 =item C<Return:>
 179
 180     Returns an array consisting of three elements
 181     * $error is undefined unless an error is detected
 182     * $results is a reference to an array of records.
 183     * $total_hits is the number of hits that would have been returned with no limit
 184
 185     If an error is returned the two other return elements are undefined. If error itself is undefined
 186     the other two elements are always defined
 187
 188 =item C<usage in the script:>
 189
 190 =back
 191
 192 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
 193
 194 if (defined $error) {
 195     $template->param(query_error => $error);
 196     warn "error: ".$error;
 197     output_html_with_http_headers $input, $cookie, $template->output;
 198     exit;
 199 }
 200
 201 my $hits = @{$marcresults};
 202 my @results;
 203
 204 for my $r ( @{$marcresults} ) {
 205     my $marcrecord = MARC::File::USMARC::decode($r);
 206     my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
 207
 208     #build the iarray of hashs for the template.
 209     push @results, {
 210         title           => $biblio->{'title'},
 211         subtitle        => $biblio->{'subtitle'},
 212         biblionumber    => $biblio->{'biblionumber'},
 213         author          => $biblio->{'author'},
 214         publishercode   => $biblio->{'publishercode'},
 215         publicationyear => $biblio->{'publicationyear'},
 216         };
 217
 218 }
 219
 220 $template->param(result=>\@results);
 221
 222 =cut
 223
 224 sub SimpleSearch {
 225     my ( $query, $offset, $max_results, $servers )  = @_;
 226
 227     return ( 'No query entered', undef, undef ) unless $query;
 228     # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
 229     my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
 230     my @zoom_queries;
 231     my @tmpresults;
 232     my @zconns;
 233     my $results = [];
 234     my $total_hits = 0;
 235
 236     my $QParser;
 237     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
 238     if ($QParser) {
 239         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 240     }
 241
 242     # Initialize & Search Zebra
 243     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 244         eval {
 245             $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 246             if ($QParser) {
 247                 $query =~ s/=/:/g;
 248                 $QParser->parse( $query );
 249                 $query = $QParser->target_syntax($servers[$i]);
 250                 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
 251             } else {
 252                 $query =~ s/:/=/g;
 253                 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
 254             }
 255             $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
 256
 257             # error handling
 258             my $error =
 259                 $zconns[$i]->errmsg() . " ("
 260               . $zconns[$i]->errcode() . ") "
 261               . $zconns[$i]->addinfo() . " "
 262               . $zconns[$i]->diagset();
 263
 264             return ( $error, undef, undef ) if $zconns[$i]->errcode();
 265         };
 266         if ($@) {
 267
 268             # caught a ZOOM::Exception
 269             my $error =
 270                 $@->message() . " ("
 271               . $@->code() . ") "
 272               . $@->addinfo() . " "
 273               . $@->diagset();
 274             warn $error." for query: $query";
 275             return ( $error, undef, undef );
 276         }
 277     }
 278
 279     _ZOOM_event_loop(
 280         \@zconns,
 281         \@tmpresults,
 282         sub {
 283             my ($i, $size) = @_;
 284             my $first_record = defined($offset) ? $offset + 1 : 1;
 285             my $hits = $tmpresults[ $i - 1 ]->size();
 286             $total_hits += $hits;
 287             my $last_record = $hits;
 288             if ( defined $max_results && $offset + $max_results < $hits ) {
 289                 $last_record = $offset + $max_results;
 290             }
 291
 292             for my $j ( $first_record .. $last_record ) {
 293                 my $record = eval {
 294                   $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
 295                   ;    # 0 indexed
 296                 };
 297                 push @{$results}, $record if defined $record;
 298             }
 299         }
 300     );
 301
 302     foreach my $zoom_query (@zoom_queries) {
 303         $zoom_query->destroy();
 304     }
 305
 306     return ( undef, $results, $total_hits );
 307 }
 308
 309 =head2 getRecords
 310
 311 ( undef, $results_hashref, \@facets_loop ) = getRecords (
 312
 313         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 314         $results_per_page, $offset,       $expanded_facet, $branches,$itemtypes,
 315         $query_type,       $scan
 316     );
 317
 318 The all singing, all dancing, multi-server, asynchronous, scanning,
 319 searching, record nabbing, facet-building
 320
 321 See verbse embedded documentation.
 322
 323 =cut
 324
 325 sub getRecords {
 326     my (
 327         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 328         $results_per_page, $offset,       $expanded_facet, $branches,
 329         $itemtypes,        $query_type,   $scan,           $opac
 330     ) = @_;
 331
 332     my @servers = @$servers_ref;
 333     my @sort_by = @$sort_by_ref;
 334
 335     # Initialize variables for the ZOOM connection and results object
 336     my $zconn;
 337     my @zconns;
 338     my @results;
 339     my $results_hashref = ();
 340
 341     # Initialize variables for the faceted results objects
 342     my $facets_counter = {};
 343     my $facets_info    = {};
 344     my $facets         = getFacets();
 345
 346     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 347
 348     ### LOOP THROUGH THE SERVERS
 349     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 350         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 351
 352 # perform the search, create the results objects
 353 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 354         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 355
 356         #$query_to_use = $simple_query if $scan;
 357         warn $simple_query if ( $scan and $DEBUG );
 358
 359         # Check if we've got a query_type defined, if so, use it
 360         eval {
 361             if ($query_type) {
 362                 if ($query_type =~ /^ccl/) {
 363                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 364                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 365                 } elsif ($query_type =~ /^cql/) {
 366                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 367                 } elsif ($query_type =~ /^pqf/) {
 368                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 369                 } else {
 370                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 371                 }
 372             } elsif ($scan) {
 373                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 374             } else {
 375                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 376             }
 377         };
 378         if ($@) {
 379             warn "WARNING: query problem with $query_to_use " . $@;
 380         }
 381
 382         # Concatenate the sort_by limits and pass them to the results object
 383         # Note: sort will override rank
 384         my $sort_by;
 385         foreach my $sort (@sort_by) {
 386             if ( $sort eq "author_az" || $sort eq "author_asc" ) {
 387                 $sort_by .= "1=1003 <i ";
 388             }
 389             elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
 390                 $sort_by .= "1=1003 >i ";
 391             }
 392             elsif ( $sort eq "popularity_asc" ) {
 393                 $sort_by .= "1=9003 <i ";
 394             }
 395             elsif ( $sort eq "popularity_dsc" ) {
 396                 $sort_by .= "1=9003 >i ";
 397             }
 398             elsif ( $sort eq "call_number_asc" ) {
 399                 $sort_by .= "1=8007  <i ";
 400             }
 401             elsif ( $sort eq "call_number_dsc" ) {
 402                 $sort_by .= "1=8007 >i ";
 403             }
 404             elsif ( $sort eq "pubdate_asc" ) {
 405                 $sort_by .= "1=31 <i ";
 406             }
 407             elsif ( $sort eq "pubdate_dsc" ) {
 408                 $sort_by .= "1=31 >i ";
 409             }
 410             elsif ( $sort eq "acqdate_asc" ) {
 411                 $sort_by .= "1=32 <i ";
 412             }
 413             elsif ( $sort eq "acqdate_dsc" ) {
 414                 $sort_by .= "1=32 >i ";
 415             }
 416             elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
 417                 $sort_by .= "1=4 <i ";
 418             }
 419             elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
 420                 $sort_by .= "1=4 >i ";
 421             }
 422             else {
 423                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 424             }
 425         }
 426         if ( $sort_by && !$scan && $results[$i] ) {
 427             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 428                 warn "WARNING sort $sort_by failed";
 429             }
 430         }
 431     }    # finished looping through servers
 432
 433     # The big moment: asynchronously retrieve results from all servers
 434         _ZOOM_event_loop(
 435             \@zconns,
 436             \@results,
 437             sub {
 438                 my ( $i, $size ) = @_;
 439                 my $results_hash;
 440
 441                 # loop through the results
 442                 $results_hash->{'hits'} = $size;
 443                 my $times;
 444                 if ( $offset + $results_per_page <= $size ) {
 445                     $times = $offset + $results_per_page;
 446                 }
 447                 else {
 448                     $times = $size;
 449                 }
 450
 451                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 452                     my $records_hash;
 453                     my $record;
 454
 455                     ## Check if it's an index scan
 456                     if ($scan) {
 457                         my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
 458
 459                  # here we create a minimal MARC record and hand it off to the
 460                  # template just like a normal result ... perhaps not ideal, but
 461                  # it works for now
 462                         my $tmprecord = MARC::Record->new();
 463                         $tmprecord->encoding('UTF-8');
 464                         my $tmptitle;
 465                         my $tmpauthor;
 466
 467                 # the minimal record in author/title (depending on MARC flavour)
 468                         if ( C4::Context->preference("marcflavour") eq
 469                             "UNIMARC" )
 470                         {
 471                             $tmptitle = MARC::Field->new(
 472                                 '200', ' ', ' ',
 473                                 a => $term,
 474                                 f => $occ
 475                             );
 476                             $tmprecord->append_fields($tmptitle);
 477                         }
 478                         else {
 479                             $tmptitle =
 480                               MARC::Field->new( '245', ' ', ' ', a => $term, );
 481                             $tmpauthor =
 482                               MARC::Field->new( '100', ' ', ' ', a => $occ, );
 483                             $tmprecord->append_fields($tmptitle);
 484                             $tmprecord->append_fields($tmpauthor);
 485                         }
 486                         $results_hash->{'RECORDS'}[$j] =
 487                           $tmprecord->as_usmarc();
 488                     }
 489
 490                     # not an index scan
 491                     else {
 492                         $record = $results[ $i - 1 ]->record($j)->raw();
 493                         # warn "RECORD $j:".$record;
 494                         $results_hash->{'RECORDS'}[$j] = $record;
 495                     }
 496
 497                 }
 498                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 499
 500                 # Fill the facets while we're looping, but only for the
 501                 # biblioserver and not for a scan
 502                 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
 503                     $facets_counter = GetFacets( $results[ $i - 1 ] );
 504                     $facets_info    = _get_facets_info( $facets );
 505                 }
 506
 507                 # BUILD FACETS
 508                 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 509                     for my $link_value (
 510                         sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 511                         keys %$facets_counter
 512                       )
 513                     {
 514                         my $expandable;
 515                         my $number_of_facets;
 516                         my @this_facets_array;
 517                         for my $one_facet (
 518                             sort {
 519                                 $facets_counter->{$link_value}
 520                                   ->{$b} <=> $facets_counter->{$link_value}
 521                                   ->{$a}
 522                             } keys %{ $facets_counter->{$link_value} }
 523                           )
 524                         {
 525                             $number_of_facets++;
 526                             if (   ( $number_of_facets <= 5 )
 527                                 || ( $expanded_facet eq $link_value )
 528                                 || ( $facets_info->{$link_value}->{'expanded'} )
 529                               )
 530                             {
 531
 532 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
 533                                 my $facet_link_value = $one_facet;
 534                                 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
 535
 536                                 # fix the length that will display in the label,
 537                                 my $facet_label_value = $one_facet;
 538                                 my $facet_max_length  = C4::Context->preference(
 539                                     'FacetLabelTruncationLength')
 540                                   || 20;
 541                                 $facet_label_value =
 542                                   substr( $one_facet, 0, $facet_max_length )
 543                                   . "..."
 544                                   if length($facet_label_value) >
 545                                       $facet_max_length;
 546
 547                             # if it's a branch, label by the name, not the code,
 548                                 if ( $link_value =~ /branch/ ) {
 549                                     if (   defined $branches
 550                                         && ref($branches) eq "HASH"
 551                                         && defined $branches->{$one_facet}
 552                                         && ref( $branches->{$one_facet} ) eq
 553                                         "HASH" )
 554                                     {
 555                                         $facet_label_value =
 556                                           $branches->{$one_facet}
 557                                           ->{'branchname'};
 558                                     }
 559                                     else {
 560                                         $facet_label_value = "*";
 561                                     }
 562                                 }
 563
 564                           # if it's a itemtype, label by the name, not the code,
 565                                 if ( $link_value =~ /itype/ ) {
 566                                     if (   defined $itemtypes
 567                                         && ref($itemtypes) eq "HASH"
 568                                         && defined $itemtypes->{$one_facet}
 569                                         && ref( $itemtypes->{$one_facet} ) eq
 570                                         "HASH" )
 571                                     {
 572                                         $facet_label_value =
 573                                           $itemtypes->{$one_facet}
 574                                           ->{'description'};
 575                                     }
 576                                 }
 577
 578                # also, if it's a location code, use the name instead of the code
 579                                 if ( $link_value =~ /location/ ) {
 580                                     $facet_label_value =
 581                                       GetKohaAuthorisedValueLib( 'LOC',
 582                                         $one_facet, $opac );
 583                                 }
 584
 585                 # but we're down with the whole label being in the link's title.
 586                                 push @this_facets_array,
 587                                   {
 588                                     facet_count =>
 589                                       $facets_counter->{$link_value}
 590                                       ->{$one_facet},
 591                                     facet_label_value => $facet_label_value,
 592                                     facet_title_value => $one_facet,
 593                                     facet_link_value  => $facet_link_value,
 594                                     type_link_value   => $link_value,
 595                                   }
 596                                   if ($facet_label_value);
 597                             }
 598                         }
 599
 600                         # handle expanded option
 601                         unless ( $facets_info->{$link_value}->{'expanded'} ) {
 602                             $expandable = 1
 603                               if ( ( $number_of_facets > 5 )
 604                                 && ( $expanded_facet ne $link_value ) );
 605                         }
 606                         push @facets_loop,
 607                           {
 608                             type_link_value => $link_value,
 609                             type_id         => $link_value . "_id",
 610                             "type_label_"
 611                               . $facets_info->{$link_value}->{'label_value'} =>
 612                               1,
 613                             facets     => \@this_facets_array,
 614                             expandable => $expandable,
 615                             expand     => $link_value,
 616                           }
 617                           unless (
 618                             (
 619                                 $facets_info->{$link_value}->{'label_value'} =~
 620                                 /Libraries/
 621                             )
 622                             and ( C4::Context->preference('singleBranchMode') )
 623                           );
 624                     }
 625                 }
 626             }
 627         );
 628     return ( undef, $results_hashref, \@facets_loop );
 629 }
 630
 631 sub GetFacets {
 632
 633     my $rs = shift;
 634     my $facets;
 635
 636     my $indexing_mode    = C4::Context->config('zebra_bib_index_mode') // 'dom';
 637     my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
 638
 639     if ( $indexing_mode eq 'dom' &&
 640          $use_zebra_facets ) {
 641         $facets = _get_facets_from_zebra( $rs );
 642     } else {
 643         $facets = _get_facets_from_records( $rs );
 644     }
 645
 646     return $facets;
 647 }
 648
 649 sub _get_facets_from_records {
 650
 651     my $rs = shift;
 652
 653     my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
 654     my $facets_config  = getFacets();
 655     my $facets         = {};
 656     my $size           = $rs->size();
 657     my $jmax           = $size > $facets_maxrecs
 658                             ? $facets_maxrecs
 659                             : $size;
 660
 661     for ( my $j = 0 ; $j < $jmax ; $j++ ) {
 662
 663         my $marc_record = new_record_from_zebra (
 664                 'biblioserver',
 665                 $rs->record( $j )->raw()
 666         );
 667
 668         if ( ! defined $marc_record ) {
 669             warn "ERROR DECODING RECORD - $@: " .
 670                 $rs->record( $j )->raw();
 671             next;
 672         }
 673
 674         _get_facets_data_from_record( $marc_record, $facets_config, $facets );
 675     }
 676
 677     return $facets;
 678 }
 679
 680 =head2 _get_facets_data_from_record
 681
 682     C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
 683
 684 Internal function that extracts facets information from a MARC::Record object
 685 and populates $facets_counter for using in getRecords.
 686
 687 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
 688 facets for Zebra).
 689
 690 =cut
 691
 692 sub _get_facets_data_from_record {
 693
 694     my ( $marc_record, $facets, $facets_counter ) = @_;
 695
 696     for my $facet (@$facets) {
 697
 698         my @used_datas = ();
 699
 700         foreach my $tag ( @{ $facet->{ tags } } ) {
 701
 702             # tag number is the first three digits
 703             my $tag_num          = substr( $tag, 0, 3 );
 704             # subfields are the remainder
 705             my $subfield_letters = substr( $tag, 3 );
 706
 707             my @fields = $marc_record->field( $tag_num );
 708             foreach my $field (@fields) {
 709                 # If $field->indicator(1) eq 'z', it means it is a 'see from'
 710                 # field introduced because of IncludeSeeFromInSearches, so skip it
 711                 next if $field->indicator(1) eq 'z';
 712
 713                 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
 714
 715                 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
 716                     push @used_datas, $data;
 717                     $facets_counter->{ $facet->{ idx } }->{ $data }++;
 718                 }
 719             }
 720         }
 721     }
 722 }
 723
 724 =head2 _get_facets_from_zebra
 725
 726     my $facets = _get_facets_from_zebra( $result_set )
 727
 728 Retrieves facets for a specified result set. It loops through the facets defined
 729 in C4::Koha::getFacets and returns a hash with the following structure:
 730
 731    {  facet_idx => {
 732             facet_value => count
 733       },
 734       ...
 735    }
 736
 737 =cut
 738
 739 sub _get_facets_from_zebra {
 740
 741     my $rs = shift;
 742
 743     # save current elementSetName
 744     my $elementSetName = $rs->option( 'elementSetName' );
 745
 746     my $facets_loop = getFacets();
 747     my $facets_data  = {};
 748     # loop through defined facets and fill the facets hashref
 749     foreach my $facet ( @$facets_loop ) {
 750
 751         my $idx = $facet->{ idx };
 752         my $sep = $facet->{ sep };
 753         my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
 754         if ( $facet_values ) {
 755             # we've actually got a result
 756             $facets_data->{ $idx } = $facet_values;
 757         }
 758     }
 759     # set elementSetName to its previous value to avoid side effects
 760     $rs->option( elementSetName => $elementSetName );
 761
 762     return $facets_data;
 763 }
 764
 765 =head2 _get_facet_from_result_set
 766
 767     my $facet_values =
 768         C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
 769
 770 Internal function that extracts facet information for a specific index ($facet_idx) and
 771 returns a hash containing facet values and count:
 772
 773     {
 774         $facet_value => $count ,
 775         ...
 776     }
 777
 778 Warning: this function has the side effect of changing the elementSetName for the result
 779 set. It is a helper function for the main loop, which takes care of backing it up for
 780 restoring.
 781
 782 =cut
 783
 784 sub _get_facet_from_result_set {
 785
 786     my $facet_idx = shift;
 787     my $rs        = shift;
 788     my $sep       = shift;
 789
 790     my $internal_sep  = '<*>';
 791     my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20;
 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:' . $facetMaxCount;
 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" would 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                     'diss',
1137                     'EAN',
1138                     'extent',
1139                     'fic',
1140                     'fiction',
1141                     'Form-subdivision',
1142                     'format',
1143                     'Geographic-subdivision',
1144                     'he',
1145                     'Heading',
1146                     'Heading-use-main-or-added-entry',
1147                     'Heading-use-series-added-entry ',
1148                     'Heading-use-subject-added-entry',
1149                     'Host-item',
1150                     'id-other',
1151                     'Illustration-code',
1152                     'Index-term-genre',
1153                     'Index-term-uncontrolled',
1154                     'ISBN',
1155                     'isbn',
1156                     'ISSN',
1157                     'issn',
1158                     'itemtype',
1159                     'kw',
1160                     'Koha-Auth-Number',
1161                     'l-format',
1162                     'language',
1163                     'language-original',
1164                     'lc-card',
1165                     'LC-card-number',
1166                     'lcn',
1167                     'lex',
1168                     'llength',
1169                     'ln',
1170                     'ln-audio',
1171                     'ln-subtitle',
1172                     'Local-classification',
1173                     'Local-number',
1174                     'Match-heading',
1175                     'Match-heading-see-from',
1176                     'Material-type',
1177                     'mc-itemtype',
1178                     'mc-rtype',
1179                     'mus',
1180                     'name',
1181                     'Music-number',
1182                     'Name-geographic',
1183                     'Name-geographic-heading',
1184                     'Name-geographic-see',
1185                     'Name-geographic-seealso',
1186                     'nb',
1187                     'Note',
1188                     'notes',
1189                     'ns',
1190                     'nt',
1191                     'pb',
1192                     'Personal-name',
1193                     'Personal-name-heading',
1194                     'Personal-name-see',
1195                     'Personal-name-seealso',
1196                     'pl',
1197                     'Place-publication',
1198                     'pn',
1199                     'popularity',
1200                     'pubdate',
1201                     'Publisher',
1202                     'Record-control-number',
1203                     'rcn',
1204                     'Record-type',
1205                     'rtype',
1206                     'se',
1207                     'See',
1208                     'See-also',
1209                     'sn',
1210                     'Stock-number',
1211                     'su',
1212                     'Subject',
1213                     'Subject-heading-thesaurus',
1214                     'Subject-name-personal',
1215                     'Subject-subdivision',
1216                     'Summary',
1217                     'Suppress',
1218                     'su-geo',
1219                     'su-na',
1220                     'su-to',
1221                     'su-ut',
1222                     'ut',
1223                     'Term-genre-form',
1224                     'Term-genre-form-heading',
1225                     'Term-genre-form-see',
1226                     'Term-genre-form-seealso',
1227                     'ti',
1228                     'Title',
1229                     'Title-cover',
1230                     'Title-series',
1231                     'Title-uniform',
1232                     'Title-uniform-heading',
1233                     'Title-uniform-see',
1234                     'Title-uniform-seealso',
1235                     'totalissues',
1236                     'yr',
1237
1238                     # items indexes
1239                     'acqsource',
1240                     'barcode',
1241                     'bc',
1242                     'branch',
1243                     'ccode',
1244                     'classification-source',
1245                     'cn-sort',
1246                     'coded-location-qualifier',
1247                     'copynumber',
1248                     'damaged',
1249                     'datelastborrowed',
1250                     'datelastseen',
1251                     'holdingbranch',
1252                     'homebranch',
1253                     'issues',
1254                     'item',
1255                     'itemnumber',
1256                     'itype',
1257                     'Local-classification',
1258                     'location',
1259                     'lost',
1260                     'materials-specified',
1261                     'mc-ccode',
1262                     'mc-itype',
1263                     'mc-loc',
1264                     'notforloan',
1265                     'Number-local-acquisition',
1266                     'onloan',
1267                     'price',
1268                     'renewals',
1269                     'replacementprice',
1270                     'replacementpricedate',
1271                     'reserves',
1272                     'restricted',
1273                     'stack',
1274                     'stocknumber',
1275                     'inv',
1276                     'uri',
1277                     'withdrawn',
1278
1279                     # subject related
1280                   );
1281
1282     return \@indexes;
1283 }
1284
1285 =head2 _handle_exploding_index
1286
1287     my $query = _handle_exploding_index($index, $term)
1288
1289 Callback routine to generate the search for "exploding" indexes (i.e.
1290 those indexes which are turned into multiple or-connected searches based
1291 on authority data).
1292
1293 =cut
1294
1295 sub _handle_exploding_index {
1296     my ($QParser, $filter, $params, $negate, $server) = @_;
1297     my $index = $filter;
1298     my $term = join(' ', @$params);
1299
1300     return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1301
1302     my $marcflavour = C4::Context->preference('marcflavour');
1303
1304     my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1305     my $wantedcodes = '';
1306     my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1307     my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1308     foreach my $auth (@$results) {
1309         my $record = MARC::Record->new_from_usmarc($auth);
1310         my @references = $record->field('5..');
1311         if (@references) {
1312             if ($index eq 'su-br') {
1313                 $wantedcodes = 'g';
1314             } elsif ($index eq 'su-na') {
1315                 $wantedcodes = 'h';
1316             } elsif ($index eq 'su-rl') {
1317                 $wantedcodes = '';
1318             }
1319             foreach my $reference (@references) {
1320                 my $codes = $reference->subfield($codesubfield);
1321                 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1322             }
1323         }
1324     }
1325     my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1326     return $query;
1327 }
1328
1329 =head2 parseQuery
1330
1331     ( $operators, $operands, $indexes, $limits,
1332       $sort_by, $scan, $lang ) =
1333             buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1334
1335 Shim function to ease the transition from buildQuery to a new QueryParser.
1336 This function is called at the beginning of buildQuery, and modifies
1337 buildQuery's input. If it can handle the input, it returns a query that
1338 buildQuery will not try to parse.
1339 =cut
1340
1341 sub parseQuery {
1342     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1343
1344     my @operators = $operators ? @$operators : ();
1345     my @indexes   = $indexes   ? @$indexes   : ();
1346     my @operands  = $operands  ? @$operands  : ();
1347     my @limits    = $limits    ? @$limits    : ();
1348     my @sort_by   = $sort_by   ? @$sort_by   : ();
1349
1350     my $query = $operands[0];
1351     my $index;
1352     my $term;
1353     my $query_desc;
1354
1355     my $QParser;
1356     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1357     undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1358     undef $QParser if (scalar @limits > 0);
1359
1360     if ($QParser)
1361     {
1362         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1363         $query = '';
1364         for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1365             next unless $operands[$ii];
1366             $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1367               if ($query);
1368             if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1369                 $query .= $operands[$ii];
1370             }
1371             elsif ( $indexes[$ii] =~ m/su-/ ) {
1372                 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1373             }
1374             else {
1375                 $query .=
1376                   ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1377             }
1378         }
1379         foreach my $limit (@limits) {
1380         }
1381         if ( scalar(@sort_by) > 0 ) {
1382             my $modifier_re =
1383               '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1384             $query =~ s/$modifier_re//g;
1385             foreach my $modifier (@sort_by) {
1386                 $query .= " #$modifier";
1387             }
1388         }
1389
1390         $query_desc = $query;
1391         $query_desc =~ s/\s+/ /g;
1392         if ( C4::Context->preference("QueryWeightFields") ) {
1393         }
1394         $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1395               { 'target_syntax_callback' => \&_handle_exploding_index } );
1396         $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1397               { 'target_syntax_callback' => \&_handle_exploding_index } );
1398         $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1399               { 'target_syntax_callback' => \&_handle_exploding_index } );
1400         $QParser->parse($query);
1401         $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1402     }
1403     else {
1404         require Koha::QueryParser::Driver::PQF;
1405         my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1406         s/$modifier_re//g for @operands;
1407     }
1408
1409     return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1410 }
1411
1412 =head2 buildQuery
1413
1414 ( $error, $query,
1415 $simple_query, $query_cgi,
1416 $query_desc, $limit,
1417 $limit_cgi, $limit_desc,
1418 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1419
1420 Build queries and limits in CCL, CGI, Human,
1421 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1422
1423 See verbose embedded documentation.
1424
1425
1426 =cut
1427
1428 sub buildQuery {
1429     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1430
1431     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1432
1433     my $query_desc;
1434     ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1435
1436     # dereference
1437     my @operators = $operators ? @$operators : ();
1438     my @indexes   = $indexes   ? @$indexes   : ();
1439     my @operands  = $operands  ? @$operands  : ();
1440     my @limits    = $limits    ? @$limits    : ();
1441     my @sort_by   = $sort_by   ? @$sort_by   : ();
1442
1443     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1444     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1445     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1446     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1447     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1448
1449     my $query        = $operands[0];
1450     my $simple_query = $operands[0];
1451
1452     # initialize the variables we're passing back
1453     my $query_cgi;
1454     my $query_type;
1455
1456     my $limit;
1457     my $limit_cgi;
1458     my $limit_desc;
1459
1460     my $stopwords_removed;    # flag to determine if stopwords have been removed
1461
1462     my $cclq       = 0;
1463     my $cclindexes = getIndexes();
1464     if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1465         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1466             my $dx = lc($1);
1467             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1468         }
1469         $query = "ccl=$query" if $cclq;
1470     }
1471
1472 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1473 # DIAGNOSTIC ONLY!!
1474     if ( $query =~ /^ccl=/ ) {
1475         my $q=$';
1476         # This is needed otherwise ccl= and &limit won't work together, and
1477         # this happens when selecting a subject on the opac-detail page
1478         @limits = grep {!/^$/} @limits;
1479         if ( @limits ) {
1480             $q .= ' and '.join(' and ', @limits);
1481         }
1482         return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $q, '', '', '', '', 'ccl' );
1483     }
1484     if ( $query =~ /^cql=/ ) {
1485         return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', '', 'cql' );
1486     }
1487     if ( $query =~ /^pqf=/ ) {
1488         if ($query_desc) {
1489             $query_cgi = "q=".uri_escape_utf8($query_desc);
1490         } else {
1491             $query_desc = $';
1492             $query_cgi = "q=pqf=".uri_escape_utf8($');
1493         }
1494         return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1495     }
1496
1497     # pass nested queries directly
1498     # FIXME: need better handling of some of these variables in this case
1499     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1500     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1501 #    if ( $query =~ /(\(|\))/ ) {
1502 #        return (
1503 #            undef,              $query, $simple_query, $query_cgi,
1504 #            $query,             $limit, $limit_cgi,    $limit_desc,
1505 #            $stopwords_removed, 'ccl'
1506 #        );
1507 #    }
1508
1509 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1510 # query operands and indexes and add stemming, truncation, field weighting, etc.
1511 # Once we do so, we'll end up with a value in $query, just like if we had an
1512 # incoming $query from the user
1513     else {
1514         $query = ""
1515           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1516         my $previous_operand
1517           ;    # a flag used to keep track if there was a previous query
1518                # if there was, we can apply the current operator
1519                # for every operand
1520         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1521
1522             # COMBINE OPERANDS, INDEXES AND OPERATORS
1523             if ( $operands[$i] ) {
1524                 $operands[$i]=~s/^\s+//;
1525
1526               # A flag to determine whether or not to add the index to the query
1527                 my $indexes_set;
1528
1529 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1530                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1531                     $weight_fields    = 0;
1532                     $stemming         = 0;
1533                     $remove_stopwords = 0;
1534                 } else {
1535                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1536                 }
1537                 my $operand = $operands[$i];
1538                 my $index   = $indexes[$i];
1539
1540                 # Add index-specific attributes
1541
1542                 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1543                 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1544                 #which is processed higher up in this sub. Other than that, year searches are typically
1545                 #handled as limits which are not processed her either.
1546
1547                 # Date of Publication
1548                 if ( $index =~ /yr/ ) {
1549                     #weight_fields/relevance search causes errors with date ranges
1550                     #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1551                     #In the case of YYYY-YYYY, it will return no results
1552                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1553                 }
1554
1555                 # Date of Acquisition
1556                 elsif ( $index =~ /acqdate/ ) {
1557                     #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1558                     #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1559                       #top of the results just because they have lots of item records matching that date.
1560                     #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1561                       #irrelevant here
1562                     #remove_stopwords doesn't function anymore so is irrelevant
1563                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1564                 }
1565                 # ISBN,ISSN,Standard Number, don't need special treatment
1566                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1567                     (
1568                         $stemming,      $auto_truncation,
1569                         $weight_fields, $fuzzy_enabled,
1570                         $remove_stopwords
1571                     ) = ( 0, 0, 0, 0, 0 );
1572
1573                     if ( $index eq 'nb' ) {
1574                         if ( C4::Context->preference("SearchWithISBNVariations") ) {
1575                             my @isbns = C4::Koha::GetVariationsOfISBN( $operand );
1576                             $operands[$i] = $operand =  '(nb=' . join(' OR nb=', @isbns) . ')';
1577                             $indexes[$i] = $index = '';
1578                         }
1579                     }
1580                 }
1581
1582                 if(not $index){
1583                     $index = 'kw';
1584                 }
1585
1586                 # Set default structure attribute (word list)
1587                 my $struct_attr = q{};
1588                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1589                     $struct_attr = ",wrdl";
1590                 }
1591
1592                 # Some helpful index variants
1593                 my $index_plus       = $index . $struct_attr . ':';
1594                 my $index_plus_comma = $index . $struct_attr . ',';
1595
1596                 # Remove Stopwords
1597                 if ($remove_stopwords) {
1598                     ( $operand, $stopwords_removed ) =
1599                       _remove_stopwords( $operand, $index );
1600                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1601                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1602                       if ( $stopwords_removed && $DEBUG );
1603                 }
1604
1605                 if ($auto_truncation){
1606                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1607                                                 #FIXME only valid with LTR scripts
1608                                                 $operand=join(" ",map{
1609                                                                                         (index($_,"*")>0?"$_":"$_*")
1610                                                                                          }split (/\s+/,$operand));
1611                                                 warn $operand if $DEBUG;
1612                                         }
1613                                 }
1614
1615                 # Detect Truncation
1616                 my $truncated_operand;
1617                 my( $nontruncated, $righttruncated, $lefttruncated,
1618                     $rightlefttruncated, $regexpr
1619                 ) = _detect_truncation( $operand, $index );
1620                 warn
1621 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1622                   if $DEBUG;
1623
1624                 # Apply Truncation
1625                 if (
1626                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1627                     scalar(@$rightlefttruncated) > 0 )
1628                 {
1629
1630                # Don't field weight or add the index to the query, we do it here
1631                     $indexes_set = 1;
1632                     undef $weight_fields;
1633                     my $previous_truncation_operand;
1634                     if (scalar @$nontruncated) {
1635                         $truncated_operand .= "$index_plus @$nontruncated ";
1636                         $previous_truncation_operand = 1;
1637                     }
1638                     if (scalar @$righttruncated) {
1639                         $truncated_operand .= "and " if $previous_truncation_operand;
1640                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1641                         $previous_truncation_operand = 1;
1642                     }
1643                     if (scalar @$lefttruncated) {
1644                         $truncated_operand .= "and " if $previous_truncation_operand;
1645                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1646                         $previous_truncation_operand = 1;
1647                     }
1648                     if (scalar @$rightlefttruncated) {
1649                         $truncated_operand .= "and " if $previous_truncation_operand;
1650                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1651                         $previous_truncation_operand = 1;
1652                     }
1653                 }
1654                 $operand = $truncated_operand if $truncated_operand;
1655                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1656
1657                 # Handle Stemming
1658                 my $stemmed_operand;
1659                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1660                                                                                 if $stemming;
1661
1662                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1663
1664                 # Handle Field Weighting
1665                 my $weighted_operand;
1666                 if ($weight_fields) {
1667                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1668                     $operand = $weighted_operand;
1669                     $indexes_set = 1;
1670                 }
1671
1672                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1673
1674                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1675                     query => $query,
1676                     query_cgi => $query_cgi,
1677                     query_desc => $query_desc,
1678                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1679                     parsed_operand => $operand,
1680                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1681                     index => $index,
1682                     index_plus => $index_plus,
1683                     indexes_set => $indexes_set,
1684                     previous_operand => $previous_operand,
1685                 });
1686
1687             }    #/if $operands
1688         }    # /for
1689     }
1690     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1691
1692     # add limits
1693     my %group_OR_limits;
1694     my $availability_limit;
1695     foreach my $this_limit (@limits) {
1696         next unless $this_limit;
1697         if ( $this_limit =~ /available/ ) {
1698 #
1699 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1700 ## In English:
1701 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1702             $availability_limit .=
1703 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1704             $limit_cgi  .= "&limit=available";
1705             $limit_desc .= "";
1706         }
1707
1708         # group_OR_limits, prefixed by mc-
1709         # OR every member of the group
1710         elsif ( $this_limit =~ /mc/ ) {
1711             my ($k,$v) = split(/:/, $this_limit,2);
1712             if ( $k !~ /mc-i(tem)?type/ ) {
1713                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1714                 $this_limit =~ tr/"//d;
1715                 $this_limit = $k.":'".$v."'";
1716             }
1717
1718             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1719             $limit_desc      .= " or " if $group_OR_limits{$k};
1720             $group_OR_limits{$k} .= "$this_limit";
1721             $limit_cgi       .= "&limit=" . uri_escape_utf8($this_limit);
1722             $limit_desc      .= " $this_limit";
1723         }
1724
1725         # Regular old limits
1726         else {
1727             $limit .= " and " if $limit || $query;
1728             $limit      .= "$this_limit";
1729             $limit_cgi  .= "&limit=" . uri_escape_utf8($this_limit);
1730             if ($this_limit =~ /^branch:(.+)/) {
1731                 my $branchcode = $1;
1732                 my $branchname = GetBranchName($branchcode);
1733                 if (defined $branchname) {
1734                     $limit_desc .= " branch:$branchname";
1735                 } else {
1736                     $limit_desc .= " $this_limit";
1737                 }
1738             } else {
1739                 $limit_desc .= " $this_limit";
1740             }
1741         }
1742     }
1743     foreach my $k (keys (%group_OR_limits)) {
1744         $limit .= " and " if ( $query || $limit );
1745         $limit .= "($group_OR_limits{$k})";
1746     }
1747     if ($availability_limit) {
1748         $limit .= " and " if ( $query || $limit );
1749         $limit .= "($availability_limit)";
1750     }
1751
1752     # Normalize the query and limit strings
1753     # This is flawed , means we can't search anything with : in it
1754     # if user wants to do ccl or cql, start the query with that
1755 #    $query =~ s/:/=/g;
1756     #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1757     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1758     $query =~ s/(?<=(wrdl)):/=/g;
1759     $query =~ s/(?<=(trn|phr)):/=/g;
1760     $query =~ s/(?<=(st-numeric)):/=/g;
1761     $query =~ s/(?<=(st-year)):/=/g;
1762     $query =~ s/(?<=(st-date-normalized)):/=/g;
1763     $limit =~ s/:/=/g;
1764     for ( $query, $query_desc, $limit, $limit_desc ) {
1765         s/  +/ /g;    # remove extra spaces
1766         s/^ //g;     # remove any beginning spaces
1767         s/ $//g;     # remove any ending spaces
1768         s/==/=/g;    # remove double == from query
1769     }
1770     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1771
1772     for ($query_cgi,$simple_query) {
1773         s/"//g;
1774     }
1775     # append the limit to the query
1776     $query .= " " . $limit;
1777
1778     # Warnings if DEBUG
1779     if ($DEBUG) {
1780         warn "QUERY:" . $query;
1781         warn "QUERY CGI:" . $query_cgi;
1782         warn "QUERY DESC:" . $query_desc;
1783         warn "LIMIT:" . $limit;
1784         warn "LIMIT CGI:" . $limit_cgi;
1785         warn "LIMIT DESC:" . $limit_desc;
1786         warn "---------\nLeave buildQuery\n---------";
1787     }
1788     return (
1789         undef,              $query, $simple_query, $query_cgi,
1790         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1791         $stopwords_removed, $query_type
1792     );
1793 }
1794
1795 =head2 _build_initial_query
1796
1797   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1798
1799   Build a section of the initial query containing indexes, operators, and operands.
1800
1801 =cut
1802
1803 sub _build_initial_query {
1804     my ($params) = @_;
1805
1806     my $operator = "";
1807     if ($params->{previous_operand}){
1808         #If there is a previous operand, add a supplied operator or the default 'and'
1809         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1810     }
1811
1812     #NOTE: indexes_set is typically set when doing truncation or field weighting
1813     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1814
1815     #e.g. "kw,wrdl:test"
1816     #e.g. " and kw,wrdl:test"
1817     $params->{query} .= $operator . $operand;
1818
1819     $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator;
1820     $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index};
1821     $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if $params->{original_operand};
1822
1823     #e.g. " and kw,wrdl: test"
1824     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1825
1826     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1827
1828     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1829 }
1830
1831 =head2 searchResults
1832
1833   my @search_results = searchResults($search_context, $searchdesc, $hits,
1834                                      $results_per_page, $offset, $scan,
1835                                      @marcresults);
1836
1837 Format results in a form suitable for passing to the template
1838
1839 =cut
1840
1841 # IMO this subroutine is pretty messy still -- it's responsible for
1842 # building the HTML output for the template
1843 sub searchResults {
1844     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1845     my $dbh = C4::Context->dbh;
1846     my @newresults;
1847
1848     require C4::Items;
1849
1850     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1851     my ($is_opac, $hidelostitems);
1852     if ($search_context eq 'opac') {
1853         $hidelostitems = C4::Context->preference('hidelostitems');
1854         $is_opac       = 1;
1855     }
1856
1857     #Build branchnames hash
1858     #find branchname
1859     #get branch information.....
1860     my %branches;
1861     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1862     $bsth->execute();
1863     while ( my $bdata = $bsth->fetchrow_hashref ) {
1864         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1865     }
1866 # FIXME - We build an authorised values hash here, using the default framework
1867 # though it is possible to have different authvals for different fws.
1868
1869     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1870
1871     # get notforloan authorised value list (see $shelflocations  FIXME)
1872     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1873
1874     #Build itemtype hash
1875     #find itemtype & itemtype image
1876     my %itemtypes;
1877     $bsth =
1878       $dbh->prepare(
1879         "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1880       );
1881     $bsth->execute();
1882     while ( my $bdata = $bsth->fetchrow_hashref ) {
1883                 foreach (qw(description imageurl summary notforloan)) {
1884                 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1885                 }
1886     }
1887
1888     #search item field code
1889     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1890
1891     ## find column names of items related to MARC
1892     my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1893     $sth2->execute;
1894     my %subfieldstosearch;
1895     while ( ( my $column ) = $sth2->fetchrow ) {
1896         my ( $tagfield, $tagsubfield ) =
1897           &GetMarcFromKohaField( "items." . $column, "" );
1898         if ( defined $tagsubfield ) {
1899             $subfieldstosearch{$column} = $tagsubfield;
1900         }
1901     }
1902
1903     # handle which records to actually retrieve
1904     my $times;
1905     if ( $hits && $offset + $results_per_page <= $hits ) {
1906         $times = $offset + $results_per_page;
1907     }
1908     else {
1909         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1910     }
1911
1912     my $marcflavour = C4::Context->preference("marcflavour");
1913     # We get the biblionumber position in MARC
1914     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1915
1916     # loop through all of the records we've retrieved
1917     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1918
1919         my $marcrecord;
1920         if ($scan) {
1921             # For Scan searches we built USMARC data
1922             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1923         } else {
1924             # Normal search, render from Zebra's output
1925             $marcrecord = new_record_from_zebra(
1926                 'biblioserver',
1927                 $marcresults->[$i]
1928             );
1929
1930             if ( ! defined $marcrecord ) {
1931                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1932                 next;
1933             }
1934         }
1935
1936         my $fw = $scan
1937              ? undef
1938              : $bibliotag < 10
1939                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1940                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1941
1942         SetUTF8Flag($marcrecord);
1943         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1944         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1945         $oldbiblio->{result_number} = $i + 1;
1946
1947         # add imageurl to itemtype if there is one
1948         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1949
1950         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1951                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1952                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1953                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1954                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1955                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1956
1957                 # edition information, if any
1958         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1959                 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1960  # Build summary if there is one (the summary is defined in the itemtypes table)
1961  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1962         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1963             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1964             my @fields  = $marcrecord->fields();
1965
1966             my $newsummary;
1967             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1968                 my $tags = {};
1969                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1970                     $tag =~ /(.{3})(.)/;
1971                     if($marcrecord->field($1)){
1972                         my @abc = $marcrecord->field($1)->subfield($2);
1973                         $tags->{$tag} = $#abc + 1 ;
1974                     }
1975                 }
1976
1977                 # We catch how many times to repeat this line
1978                 my $max = 0;
1979                 foreach my $tag (keys(%$tags)){
1980                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1981                  }
1982
1983                 # we replace, and repeat each line
1984                 for (my $i = 0 ; $i < $max ; $i++){
1985                     my $newline = $line;
1986
1987                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1988                         $tag =~ /(.{3})(.)/;
1989
1990                         if($marcrecord->field($1)){
1991                             my @repl = $marcrecord->field($1)->subfield($2);
1992                             my $subfieldvalue = $repl[$i];
1993                             $newline =~ s/\[$tag\]/$subfieldvalue/g;
1994                         }
1995                     }
1996                     $newsummary .= "$newline\n";
1997                 }
1998             }
1999
2000             $newsummary =~ s/\[(.*?)]//g;
2001             $newsummary =~ s/\n/<br\/>/g;
2002             $oldbiblio->{summary} = $newsummary;
2003         }
2004
2005         # Pull out the items fields
2006         my @fields = $marcrecord->field($itemtag);
2007         my $marcflavor = C4::Context->preference("marcflavour");
2008         # adding linked items that belong to host records
2009         my $analyticsfield = '773';
2010         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2011             $analyticsfield = '773';
2012         } elsif ($marcflavor eq 'UNIMARC') {
2013             $analyticsfield = '461';
2014         }
2015         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2016             my $hostbiblionumber = $hostfield->subfield("0");
2017             my $linkeditemnumber = $hostfield->subfield("9");
2018             if(!$hostbiblionumber eq undef){
2019                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2020                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2021                 if(!$hostbiblio eq undef){
2022                     my @hostitems = $hostbiblio->field($itemfield);
2023                     foreach my $hostitem (@hostitems){
2024                         if ($hostitem->subfield("9") eq $linkeditemnumber){
2025                             my $linkeditem =$hostitem;
2026                             # append linked items if they exist
2027                             if (!$linkeditem eq undef){
2028                                 push (@fields, $linkeditem);}
2029                         }
2030                     }
2031                 }
2032             }
2033         }
2034
2035         # Setting item statuses for display
2036         my @available_items_loop;
2037         my @onloan_items_loop;
2038         my @other_items_loop;
2039
2040         my $available_items;
2041         my $onloan_items;
2042         my $other_items;
2043
2044         my $ordered_count         = 0;
2045         my $available_count       = 0;
2046         my $onloan_count          = 0;
2047         my $longoverdue_count     = 0;
2048         my $other_count           = 0;
2049         my $withdrawn_count        = 0;
2050         my $itemlost_count        = 0;
2051         my $hideatopac_count      = 0;
2052         my $itembinding_count     = 0;
2053         my $itemdamaged_count     = 0;
2054         my $item_in_transit_count = 0;
2055         my $can_place_holds       = 0;
2056         my $item_onhold_count     = 0;
2057         my $notforloan_count      = 0;
2058         my $items_count           = scalar(@fields);
2059         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2060         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2061         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2062
2063         # loop through every item
2064         foreach my $field (@fields) {
2065             my $item;
2066
2067             # populate the items hash
2068             foreach my $code ( keys %subfieldstosearch ) {
2069                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2070             }
2071             $item->{description} = $itemtypes{ $item->{itype} }{description};
2072
2073                 # OPAC hidden items
2074             if ($is_opac) {
2075                 # hidden because lost
2076                 if ($hidelostitems && $item->{itemlost}) {
2077                     $hideatopac_count++;
2078                     next;
2079                 }
2080                 # hidden based on OpacHiddenItems syspref
2081                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2082                 if (scalar @hi) {
2083                     push @hiddenitems, @hi;
2084                     $hideatopac_count++;
2085                     next;
2086                 }
2087             }
2088
2089             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2090             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2091
2092             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2093             if ($item->{$hbranch}) {
2094                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2095             }
2096             elsif ($item->{$otherbranch}) {     # Last resort
2097                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2098             }
2099
2100                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2101 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2102             my $userenv = C4::Context->userenv;
2103             if ( $item->{onloan}
2104                 && !( C4::Members::GetHideLostItemsPreference( $userenv->{'number'} ) && $item->{itemlost} ) )
2105             {
2106                 $onloan_count++;
2107                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2108                 $onloan_items->{$key}->{due_date} = format_date( $item->{onloan} );
2109                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2110                 $onloan_items->{$key}->{branchname}     = $item->{branchname};
2111                 $onloan_items->{$key}->{location}       = $shelflocations->{ $item->{location} };
2112                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2113                 $onloan_items->{$key}->{description}    = $item->{description};
2114                 $onloan_items->{$key}->{imageurl} =
2115                   getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2116
2117                 # if something's checked out and lost, mark it as 'long overdue'
2118                 if ( $item->{itemlost} ) {
2119                     $onloan_items->{$key}->{longoverdue}++;
2120                     $longoverdue_count++;
2121                 }
2122                 else {    # can place holds as long as item isn't lost
2123                     $can_place_holds = 1;
2124                 }
2125             }
2126
2127          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2128             else {
2129
2130                 # item is on order
2131                 if ( $item->{notforloan} < 0 ) {
2132                     $ordered_count++;
2133                 } elsif ( $item->{notforloan} > 0 ) {
2134                     $notforloan_count++;
2135                 }
2136
2137                 # is item in transit?
2138                 my $transfertwhen = '';
2139                 my ($transfertfrom, $transfertto);
2140
2141                 # is item on the reserve shelf?
2142                 my $reservestatus = '';
2143
2144                 unless ($item->{withdrawn}
2145                         || $item->{itemlost}
2146                         || $item->{damaged}
2147                         || $item->{notforloan}
2148                         || $items_count > 20) {
2149
2150                     # A couple heuristics to limit how many times
2151                     # we query the database for item transfer information, sacrificing
2152                     # accuracy in some cases for speed;
2153                     #
2154                     # 1. don't query if item has one of the other statuses
2155                     # 2. don't check transit status if the bib has
2156                     #    more than 20 items
2157                     #
2158                     # FIXME: to avoid having the query the database like this, and to make
2159                     #        the in transit status count as unavailable for search limiting,
2160                     #        should map transit status to record indexed in Zebra.
2161                     #
2162                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2163                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} );
2164                 }
2165
2166                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2167                 if (   $item->{withdrawn}
2168                     || $item->{itemlost}
2169                     || $item->{damaged}
2170                     || $item->{notforloan}
2171                     || $reservestatus eq 'Waiting'
2172                     || ($transfertwhen ne ''))
2173                 {
2174                     $withdrawn_count++        if $item->{withdrawn};
2175                     $itemlost_count++        if $item->{itemlost};
2176                     $itemdamaged_count++     if $item->{damaged};
2177                     $item_in_transit_count++ if $transfertwhen ne '';
2178                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2179                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2180
2181                     # can place a hold on a item if
2182                     # not lost nor withdrawn
2183                     # not damaged unless AllowHoldsOnDamagedItems is true
2184                     # item is either for loan or on order (notforloan < 0)
2185                     $can_place_holds = 1
2186                       if (
2187                            !$item->{itemlost}
2188                         && !$item->{withdrawn}
2189                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2190                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2191                       );
2192
2193                     $other_count++;
2194
2195                     my $key = $prefix . $item->{status};
2196                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2197                         $other_items->{$key}->{$_} = $item->{$_};
2198                     }
2199                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2200                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2201                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2202                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2203                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2204                                         $other_items->{$key}->{description} = $item->{description};
2205                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2206                 }
2207                 # item is available
2208                 else {
2209                     $can_place_holds = 1;
2210                     $available_count++;
2211                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2212                                         foreach (qw(branchname itemcallnumber description)) {
2213                         $available_items->{$prefix}->{$_} = $item->{$_};
2214                                         }
2215                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2216                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2217                 }
2218             }
2219         }    # notforloan, item level and biblioitem level
2220
2221         # if all items are hidden, do not show the record
2222         if ($items_count > 0 && $hideatopac_count == $items_count) {
2223             next;
2224         }
2225
2226         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2227         for my $key ( sort keys %$onloan_items ) {
2228             (++$onloanitemscount > $maxitems) and last;
2229             push @onloan_items_loop, $onloan_items->{$key};
2230         }
2231         for my $key ( sort keys %$other_items ) {
2232             (++$otheritemscount > $maxitems) and last;
2233             push @other_items_loop, $other_items->{$key};
2234         }
2235         for my $key ( sort keys %$available_items ) {
2236             (++$availableitemscount > $maxitems) and last;
2237             push @available_items_loop, $available_items->{$key}
2238         }
2239
2240         # XSLT processing of some stuff
2241         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2242         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2243             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2244         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2245         }
2246
2247         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2248         if (!C4::Context->preference("item-level_itypes")) {
2249             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2250                 $can_place_holds = 0;
2251             }
2252         }
2253         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2254         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2255         $oldbiblio->{items_count}          = $items_count;
2256         $oldbiblio->{available_items_loop} = \@available_items_loop;
2257         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2258         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2259         $oldbiblio->{availablecount}       = $available_count;
2260         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2261         $oldbiblio->{onloancount}          = $onloan_count;
2262         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2263         $oldbiblio->{othercount}           = $other_count;
2264         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2265         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2266         $oldbiblio->{itemlostcount}        = $itemlost_count;
2267         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2268         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2269         $oldbiblio->{onholdcount}          = $item_onhold_count;
2270         $oldbiblio->{orderedcount}         = $ordered_count;
2271         $oldbiblio->{notforloancount}      = $notforloan_count;
2272
2273         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2274             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2275             my $subfields = substr $fieldspec, 3;
2276             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2277             my @alternateholdingsinfo = ();
2278             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2279             my $alternateholdingscount = 0;
2280
2281             for my $field (@holdingsfields) {
2282                 my %holding = ( holding => '' );
2283                 my $havesubfield = 0;
2284                 for my $subfield ($field->subfields()) {
2285                     if ((index $subfields, $$subfield[0]) >= 0) {
2286                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2287                         $holding{'holding'} .= $$subfield[1];
2288                         $havesubfield++;
2289                     }
2290                 }
2291                 if ($havesubfield) {
2292                     push(@alternateholdingsinfo, \%holding);
2293                     $alternateholdingscount++;
2294                 }
2295             }
2296
2297             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2298             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2299         }
2300
2301         push( @newresults, $oldbiblio );
2302     }
2303
2304     return @newresults;
2305 }
2306
2307 =head2 SearchAcquisitions
2308     Search for acquisitions
2309 =cut
2310
2311 sub SearchAcquisitions{
2312     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2313
2314     my $dbh=C4::Context->dbh;
2315     # Variable initialization
2316     my $str=qq|
2317     SELECT marcxml
2318     FROM biblio
2319     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2320     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2321     WHERE dateaccessioned BETWEEN ? AND ?
2322     |;
2323
2324     my (@params,@loopcriteria);
2325
2326     push @params, $datebegin->output("iso");
2327     push @params, $dateend->output("iso");
2328
2329     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2330         if(C4::Context->preference("item-level_itypes")){
2331             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2332         }else{
2333             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2334         }
2335         push @params, @$itemtypes;
2336     }
2337
2338     if ($criteria =~/itemtype/){
2339         if(C4::Context->preference("item-level_itypes")){
2340             $str .= "AND items.itype=? ";
2341         }else{
2342             $str .= "AND biblioitems.itemtype=? ";
2343         }
2344
2345         if(scalar(@$itemtypes) == 0){
2346             my $itypes = GetItemTypes();
2347             for my $key (keys %$itypes){
2348                 push @$itemtypes, $key;
2349             }
2350         }
2351
2352         @loopcriteria= @$itemtypes;
2353     }elsif ($criteria=~/itemcallnumber/){
2354         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2355                  OR items.itemcallnumber is NULL
2356                  OR items.itemcallnumber = '')";
2357
2358         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2359     }else {
2360         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2361         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2362     }
2363
2364     if ($orderby =~ /date_desc/){
2365         $str.=" ORDER BY dateaccessioned DESC";
2366     } else {
2367         $str.=" ORDER BY title";
2368     }
2369
2370     my $qdataacquisitions=$dbh->prepare($str);
2371
2372     my @loopacquisitions;
2373     foreach my $value(@loopcriteria){
2374         push @params,$value;
2375         my %cell;
2376         $cell{"title"}=$value;
2377         $cell{"titlecode"}=$value;
2378
2379         eval{$qdataacquisitions->execute(@params);};
2380
2381         if ($@){ warn "recentacquisitions Error :$@";}
2382         else {
2383             my @loopdata;
2384             while (my $data=$qdataacquisitions->fetchrow_hashref){
2385                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2386             }
2387             $cell{"loopdata"}=\@loopdata;
2388         }
2389         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2390         pop @params;
2391     }
2392     $qdataacquisitions->finish;
2393     return \@loopacquisitions;
2394 }
2395
2396 =head2 enabled_staff_search_views
2397
2398 %hash = enabled_staff_search_views()
2399
2400 This function returns a hash that contains three flags obtained from the system
2401 preferences, used to determine whether a particular staff search results view
2402 is enabled.
2403
2404 =over 2
2405
2406 =item C<Output arg:>
2407
2408     * $hash{can_view_MARC} is true only if the MARC view is enabled
2409     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2410     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2411
2412 =item C<usage in the script:>
2413
2414 =back
2415
2416 $template->param ( C4::Search::enabled_staff_search_views );
2417
2418 =cut
2419
2420 sub enabled_staff_search_views
2421 {
2422         return (
2423                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2424                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2425                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2426         );
2427 }
2428
2429 sub PurgeSearchHistory{
2430     my ($pSearchhistory)=@_;
2431     my $dbh = C4::Context->dbh;
2432     my $sth = $dbh->prepare("DELETE FROM search_history WHERE time < DATE_SUB( NOW(), INTERVAL ? DAY )");
2433     $sth->execute($pSearchhistory) or die $dbh->errstr;
2434 }
2435
2436 =head2 z3950_search_args
2437
2438 $arrayref = z3950_search_args($matchpoints)
2439
2440 This function returns an array reference that contains the search parameters to be
2441 passed to the Z39.50 search script (z3950_search.pl). The array elements
2442 are hash refs whose keys are name and value, and whose values are the
2443 name of a search parameter, the value of that search parameter and the URL encoded
2444 value of that parameter.
2445
2446 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2447
2448 The search parameter values are obtained from the bibliographic record whose
2449 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2450
2451 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2452 a general purpose search argument. In this case, the returned array contains only
2453 entry: the key is 'title' and the value is derived from $matchpoints.
2454
2455 If a search parameter value is undefined or empty, it is not included in the returned
2456 array.
2457
2458 The returned array reference may be passed directly to the template parameters.
2459
2460 =over 2
2461
2462 =item C<Output arg:>
2463
2464     * $array containing hash refs as described above
2465
2466 =item C<usage in the script:>
2467
2468 =back
2469
2470 $data = Biblio::GetBiblioData($bibno);
2471 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2472
2473 *OR*
2474
2475 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2476
2477 =cut
2478
2479 sub z3950_search_args {
2480     my $bibrec = shift;
2481
2482     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2483     my $isbn = Business::ISBN->new( $isbn_string );
2484
2485     if (defined $isbn && $isbn->is_valid)
2486     {
2487         if ( ref($bibrec) ) {
2488             $bibrec->{isbn} = $isbn_string;
2489             $bibrec->{title} = undef;
2490         } else {
2491             $bibrec = { isbn => $isbn_string };
2492         }
2493     }
2494     else {
2495         $bibrec = { title => $bibrec } if !ref $bibrec;
2496     }
2497     my $array = [];
2498     for my $field (qw/ lccn isbn issn title author dewey subject /)
2499     {
2500         push @$array, { name => $field, value => $bibrec->{$field} }
2501           if defined $bibrec->{$field};
2502     }
2503     return $array;
2504 }
2505
2506 =head2 GetDistinctValues($field);
2507
2508 C<$field> is a reference to the fields array
2509
2510 =cut
2511
2512 sub GetDistinctValues {
2513     my ($fieldname,$string)=@_;
2514     # returns a reference to a hash of references to branches...
2515     if ($fieldname=~/\./){
2516                         my ($table,$column)=split /\./, $fieldname;
2517                         my $dbh = C4::Context->dbh;
2518                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2519                         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 ");
2520                         $sth->execute;
2521                         my $elements=$sth->fetchall_arrayref({});
2522                         return $elements;
2523    }
2524    else {
2525                 $string||= qq("");
2526                 my @servers=qw<biblioserver authorityserver>;
2527                 my (@zconns,@results);
2528         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2529                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2530                         $results[$i] =
2531                       $zconns[$i]->scan(
2532                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2533                       );
2534                 }
2535                 # The big moment: asynchronously retrieve results from all servers
2536                 my @elements;
2537         _ZOOM_event_loop(
2538             \@zconns,
2539             \@results,
2540             sub {
2541                 my ( $i, $size ) = @_;
2542                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2543                     my %hashscan;
2544                     @hashscan{qw(value cnt)} =
2545                       $results[ $i - 1 ]->display_term($j);
2546                     push @elements, \%hashscan;
2547                 }
2548             }
2549         );
2550                 return \@elements;
2551    }
2552 }
2553
2554 =head2 _ZOOM_event_loop
2555
2556     _ZOOM_event_loop(\@zconns, \@results, sub {
2557         my ( $i, $size ) = @_;
2558         ....
2559     } );
2560
2561 Processes a ZOOM event loop and passes control to a closure for
2562 processing the results, and destroying the resultsets.
2563
2564 =cut
2565
2566 sub _ZOOM_event_loop {
2567     my ($zconns, $results, $callback) = @_;
2568     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2569         my $ev = $zconns->[ $i - 1 ]->last_event();
2570         if ( $ev == ZOOM::Event::ZEND ) {
2571             next unless $results->[ $i - 1 ];
2572             my $size = $results->[ $i - 1 ]->size();
2573             if ( $size > 0 ) {
2574                 $callback->($i, $size);
2575             }
2576         }
2577     }
2578
2579     foreach my $result (@$results) {
2580         $result->destroy();
2581     }
2582 }
2583
2584 =head2 new_record_from_zebra
2585
2586 Given raw data from a Zebra result set, return a MARC::Record object
2587
2588 This helper function is needed to take into account all the involved
2589 system preferences and configuration variables to properly create the
2590 MARC::Record object.
2591
2592 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2593 data. If we are using DOM, then it has to be MARCXML.
2594
2595 =cut
2596
2597 sub new_record_from_zebra {
2598
2599     my $server   = shift;
2600     my $raw_data = shift;
2601     # Set the default indexing modes
2602     my $index_mode = ( $server eq 'biblioserver' )
2603                         ? C4::Context->config('zebra_bib_index_mode') // 'dom'
2604                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2605
2606     my $marc_record =  eval {
2607         if ( $index_mode eq 'dom' ) {
2608             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2609         } else {
2610             MARC::Record->new_from_usmarc( $raw_data );
2611         }
2612     };
2613
2614     if ($@) {
2615         return;
2616     } else {
2617         return $marc_record;
2618     }
2619
2620 }
2621
2622 END { }    # module clean-up code here (global destructor)
2623
2624 1;
2625 __END__
2626
2627 =head1 AUTHOR
2628
2629 Koha Development Team <http://koha-community.org/>
2630
2631 =cut