C4/Search.pm

   1 package C4::Search;
   2
   3 # This file is part of Koha.
   4 #
   5 # Koha is free software; you can redistribute it and/or modify it under the
   6 # terms of the GNU General Public License as published by the Free Software
   7 # Foundation; either version 2 of the License, or (at your option) any later
   8 # version.
   9 #
  10 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  11 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  12 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU General Public License along with
  15 # Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
  16 # Suite 330, Boston, MA  02111-1307 USA
  17
  18 use strict;
  19 #use warnings; FIXME - Bug 2505
  20 require Exporter;
  21 use C4::Context;
  22 use C4::Biblio;    # GetMarcFromKohaField, GetBiblioData
  23 use C4::Koha;      # getFacets
  24 use Lingua::Stem;
  25 use C4::Search::PazPar2;
  26 use XML::Simple;
  27 use C4::Dates qw(format_date);
  28 use C4::Members qw(GetHideLostItemsPreference);
  29 use C4::XSLT;
  30 use C4::Branch;
  31 use C4::Reserves;    # GetReserveStatus
  32 use C4::Debug;
  33 use C4::Charset;
  34 use YAML;
  35 use URI::Escape;
  36 use Business::ISBN;
  37 use MARC::Record;
  38 use MARC::Field;
  39 use utf8;
  40 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  41
  42 # set the version for version checking
  43 BEGIN {
  44     $VERSION = 3.07.00.049;
  45     $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
  46 }
  47
  48 =head1 NAME
  49
  50 C4::Search - Functions for searching the Koha catalog.
  51
  52 =head1 SYNOPSIS
  53
  54 See opac/opac-search.pl or catalogue/search.pl for example of usage
  55
  56 =head1 DESCRIPTION
  57
  58 This module provides searching functions for Koha's bibliographic databases
  59
  60 =head1 FUNCTIONS
  61
  62 =cut
  63
  64 @ISA    = qw(Exporter);
  65 @EXPORT = qw(
  66   &FindDuplicate
  67   &SimpleSearch
  68   &searchResults
  69   &getRecords
  70   &buildQuery
  71   &GetDistinctValues
  72   &enabled_staff_search_views
  73   &PurgeSearchHistory
  74 );
  75
  76 # make all your functions, whether exported or not;
  77
  78 =head2 FindDuplicate
  79
  80 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
  81
  82 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
  83
  84 =cut
  85
  86 sub FindDuplicate {
  87     my ($record) = @_;
  88     my $dbh = C4::Context->dbh;
  89     my $result = TransformMarcToKoha( $dbh, $record, '' );
  90     my $sth;
  91     my $query;
  92     my $search;
  93     my $type;
  94     my ( $biblionumber, $title );
  95
  96     # search duplicate on ISBN, easy and fast..
  97     # ... normalize first
  98     if ( $result->{isbn} ) {
  99         $result->{isbn} =~ s/\(.*$//;
 100         $result->{isbn} =~ s/\s+$//;
 101         $query = "isbn:$result->{isbn}";
 102     }
 103     else {
 104         my $QParser;
 105         $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
 106         my $titleindex;
 107         my $authorindex;
 108         my $op;
 109
 110         if ($QParser) {
 111             $titleindex = 'title|exact';
 112             $authorindex = 'author|exact';
 113             $op = '&&';
 114             $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 115         } else {
 116             $titleindex = 'ti,ext';
 117             $authorindex = 'au,ext';
 118             $op = 'and';
 119         }
 120
 121         $result->{title} =~ s /\\//g;
 122         $result->{title} =~ s /\"//g;
 123         $result->{title} =~ s /\(//g;
 124         $result->{title} =~ s /\)//g;
 125
 126         # FIXME: instead of removing operators, could just do
 127         # quotes around the value
 128         $result->{title} =~ s/(and|or|not)//g;
 129         $query = "$titleindex:\"$result->{title}\"";
 130         if   ( $result->{author} ) {
 131             $result->{author} =~ s /\\//g;
 132             $result->{author} =~ s /\"//g;
 133             $result->{author} =~ s /\(//g;
 134             $result->{author} =~ s /\)//g;
 135
 136             # remove valid operators
 137             $result->{author} =~ s/(and|or|not)//g;
 138             $query .= " $op $authorindex:\"$result->{author}\"";
 139         }
 140     }
 141
 142     my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
 143     my @results;
 144     if (!defined $error) {
 145         foreach my $possible_duplicate_record (@{$searchresults}) {
 146             my $marcrecord = new_record_from_zebra(
 147                 'biblioserver',
 148                 $possible_duplicate_record
 149             );
 150
 151             my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
 152
 153             # FIXME :: why 2 $biblionumber ?
 154             if ($result) {
 155                 push @results, $result->{'biblionumber'};
 156                 push @results, $result->{'title'};
 157             }
 158         }
 159     }
 160     return @results;
 161 }
 162
 163 =head2 SimpleSearch
 164
 165 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
 166
 167 This function provides a simple search API on the bibliographic catalog
 168
 169 =over 2
 170
 171 =item C<input arg:>
 172
 173     * $query can be a simple keyword or a complete CCL query
 174     * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
 175     * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
 176     * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
 177
 178
 179 =item C<Return:>
 180
 181     Returns an array consisting of three elements
 182     * $error is undefined unless an error is detected
 183     * $results is a reference to an array of records.
 184     * $total_hits is the number of hits that would have been returned with no limit
 185
 186     If an error is returned the two other return elements are undefined. If error itself is undefined
 187     the other two elements are always defined
 188
 189 =item C<usage in the script:>
 190
 191 =back
 192
 193 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
 194
 195 if (defined $error) {
 196     $template->param(query_error => $error);
 197     warn "error: ".$error;
 198     output_html_with_http_headers $input, $cookie, $template->output;
 199     exit;
 200 }
 201
 202 my $hits = @{$marcresults};
 203 my @results;
 204
 205 for my $r ( @{$marcresults} ) {
 206     my $marcrecord = MARC::File::USMARC::decode($r);
 207     my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
 208
 209     #build the iarray of hashs for the template.
 210     push @results, {
 211         title           => $biblio->{'title'},
 212         subtitle        => $biblio->{'subtitle'},
 213         biblionumber    => $biblio->{'biblionumber'},
 214         author          => $biblio->{'author'},
 215         publishercode   => $biblio->{'publishercode'},
 216         publicationyear => $biblio->{'publicationyear'},
 217         };
 218
 219 }
 220
 221 $template->param(result=>\@results);
 222
 223 =cut
 224
 225 sub SimpleSearch {
 226     my ( $query, $offset, $max_results, $servers )  = @_;
 227
 228     return ( 'No query entered', undef, undef ) unless $query;
 229     # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
 230     my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
 231     my @zoom_queries;
 232     my @tmpresults;
 233     my @zconns;
 234     my $results = [];
 235     my $total_hits = 0;
 236
 237     my $QParser;
 238     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') && ! ($query =~ m/\w,\w|\w=\w/));
 239     if ($QParser) {
 240         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
 241     }
 242
 243     # Initialize & Search Zebra
 244     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 245         eval {
 246             $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 247             if ($QParser) {
 248                 $query =~ s/=/:/g;
 249                 $QParser->parse( $query );
 250                 $query = $QParser->target_syntax($servers[$i]);
 251                 $zoom_queries[$i] = new ZOOM::Query::PQF( $query, $zconns[$i]);
 252             } else {
 253                 $query =~ s/:/=/g;
 254                 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
 255             }
 256             $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
 257
 258             # error handling
 259             my $error =
 260                 $zconns[$i]->errmsg() . " ("
 261               . $zconns[$i]->errcode() . ") "
 262               . $zconns[$i]->addinfo() . " "
 263               . $zconns[$i]->diagset();
 264
 265             return ( $error, undef, undef ) if $zconns[$i]->errcode();
 266         };
 267         if ($@) {
 268
 269             # caught a ZOOM::Exception
 270             my $error =
 271                 $@->message() . " ("
 272               . $@->code() . ") "
 273               . $@->addinfo() . " "
 274               . $@->diagset();
 275             warn $error." for query: $query";
 276             return ( $error, undef, undef );
 277         }
 278     }
 279
 280     _ZOOM_event_loop(
 281         \@zconns,
 282         \@tmpresults,
 283         sub {
 284             my ($i, $size) = @_;
 285             my $first_record = defined($offset) ? $offset + 1 : 1;
 286             my $hits = $tmpresults[ $i - 1 ]->size();
 287             $total_hits += $hits;
 288             my $last_record = $hits;
 289             if ( defined $max_results && $offset + $max_results < $hits ) {
 290                 $last_record = $offset + $max_results;
 291             }
 292
 293             for my $j ( $first_record .. $last_record ) {
 294                 my $record = eval {
 295                   $tmpresults[ $i - 1 ]->record( $j - 1 )->raw()
 296                   ;    # 0 indexed
 297                 };
 298                 push @{$results}, $record if defined $record;
 299             }
 300         }
 301     );
 302
 303     foreach my $zoom_query (@zoom_queries) {
 304         $zoom_query->destroy();
 305     }
 306
 307     return ( undef, $results, $total_hits );
 308 }
 309
 310 =head2 getRecords
 311
 312 ( undef, $results_hashref, \@facets_loop ) = getRecords (
 313
 314         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 315         $results_per_page, $offset,       $expanded_facet, $branches,$itemtypes,
 316         $query_type,       $scan
 317     );
 318
 319 The all singing, all dancing, multi-server, asynchronous, scanning,
 320 searching, record nabbing, facet-building
 321
 322 See verbse embedded documentation.
 323
 324 =cut
 325
 326 sub getRecords {
 327     my (
 328         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 329         $results_per_page, $offset,       $expanded_facet, $branches,
 330         $itemtypes,        $query_type,   $scan,           $opac
 331     ) = @_;
 332
 333     my @servers = @$servers_ref;
 334     my @sort_by = @$sort_by_ref;
 335
 336     # Initialize variables for the ZOOM connection and results object
 337     my $zconn;
 338     my @zconns;
 339     my @results;
 340     my $results_hashref = ();
 341
 342     # Initialize variables for the faceted results objects
 343     my $facets_counter = {};
 344     my $facets_info    = {};
 345     my $facets         = getFacets();
 346
 347     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 348
 349     ### LOOP THROUGH THE SERVERS
 350     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 351         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 352
 353 # perform the search, create the results objects
 354 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 355         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 356
 357         #$query_to_use = $simple_query if $scan;
 358         warn $simple_query if ( $scan and $DEBUG );
 359
 360         # Check if we've got a query_type defined, if so, use it
 361         eval {
 362             if ($query_type) {
 363                 if ($query_type =~ /^ccl/) {
 364                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 365                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 366                 } elsif ($query_type =~ /^cql/) {
 367                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 368                 } elsif ($query_type =~ /^pqf/) {
 369                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 370                 } else {
 371                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 372                 }
 373             } elsif ($scan) {
 374                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 375             } else {
 376                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 377             }
 378         };
 379         if ($@) {
 380             warn "WARNING: query problem with $query_to_use " . $@;
 381         }
 382
 383         # Concatenate the sort_by limits and pass them to the results object
 384         # Note: sort will override rank
 385         my $sort_by;
 386         foreach my $sort (@sort_by) {
 387             if ( $sort eq "author_az" || $sort eq "author_asc" ) {
 388                 $sort_by .= "1=1003 <i ";
 389             }
 390             elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
 391                 $sort_by .= "1=1003 >i ";
 392             }
 393             elsif ( $sort eq "popularity_asc" ) {
 394                 $sort_by .= "1=9003 <i ";
 395             }
 396             elsif ( $sort eq "popularity_dsc" ) {
 397                 $sort_by .= "1=9003 >i ";
 398             }
 399             elsif ( $sort eq "call_number_asc" ) {
 400                 $sort_by .= "1=8007  <i ";
 401             }
 402             elsif ( $sort eq "call_number_dsc" ) {
 403                 $sort_by .= "1=8007 >i ";
 404             }
 405             elsif ( $sort eq "pubdate_asc" ) {
 406                 $sort_by .= "1=31 <i ";
 407             }
 408             elsif ( $sort eq "pubdate_dsc" ) {
 409                 $sort_by .= "1=31 >i ";
 410             }
 411             elsif ( $sort eq "acqdate_asc" ) {
 412                 $sort_by .= "1=32 <i ";
 413             }
 414             elsif ( $sort eq "acqdate_dsc" ) {
 415                 $sort_by .= "1=32 >i ";
 416             }
 417             elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
 418                 $sort_by .= "1=4 <i ";
 419             }
 420             elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
 421                 $sort_by .= "1=4 >i ";
 422             }
 423             else {
 424                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 425             }
 426         }
 427         if ( $sort_by && !$scan && $results[$i] ) {
 428             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 429                 warn "WARNING sort $sort_by failed";
 430             }
 431         }
 432     }    # finished looping through servers
 433
 434     # The big moment: asynchronously retrieve results from all servers
 435         _ZOOM_event_loop(
 436             \@zconns,
 437             \@results,
 438             sub {
 439                 my ( $i, $size ) = @_;
 440                 my $results_hash;
 441
 442                 # loop through the results
 443                 $results_hash->{'hits'} = $size;
 444                 my $times;
 445                 if ( $offset + $results_per_page <= $size ) {
 446                     $times = $offset + $results_per_page;
 447                 }
 448                 else {
 449                     $times = $size;
 450                 }
 451
 452                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 453                     my $records_hash;
 454                     my $record;
 455
 456                     ## Check if it's an index scan
 457                     if ($scan) {
 458                         my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j);
 459
 460                  # here we create a minimal MARC record and hand it off to the
 461                  # template just like a normal result ... perhaps not ideal, but
 462                  # it works for now
 463                         my $tmprecord = MARC::Record->new();
 464                         $tmprecord->encoding('UTF-8');
 465                         my $tmptitle;
 466                         my $tmpauthor;
 467
 468                 # the minimal record in author/title (depending on MARC flavour)
 469                         if ( C4::Context->preference("marcflavour") eq
 470                             "UNIMARC" )
 471                         {
 472                             $tmptitle = MARC::Field->new(
 473                                 '200', ' ', ' ',
 474                                 a => $term,
 475                                 f => $occ
 476                             );
 477                             $tmprecord->append_fields($tmptitle);
 478                         }
 479                         else {
 480                             $tmptitle =
 481                               MARC::Field->new( '245', ' ', ' ', a => $term, );
 482                             $tmpauthor =
 483                               MARC::Field->new( '100', ' ', ' ', a => $occ, );
 484                             $tmprecord->append_fields($tmptitle);
 485                             $tmprecord->append_fields($tmpauthor);
 486                         }
 487                         $results_hash->{'RECORDS'}[$j] =
 488                           $tmprecord->as_usmarc();
 489                     }
 490
 491                     # not an index scan
 492                     else {
 493                         $record = $results[ $i - 1 ]->record($j)->raw();
 494                         # warn "RECORD $j:".$record;
 495                         $results_hash->{'RECORDS'}[$j] = $record;
 496                     }
 497
 498                 }
 499                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 500
 501                 # Fill the facets while we're looping, but only for the
 502                 # biblioserver and not for a scan
 503                 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
 504                     $facets_counter = GetFacets( $results[ $i - 1 ] );
 505                     $facets_info    = _get_facets_info( $facets );
 506                 }
 507
 508                 # BUILD FACETS
 509                 if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 510                     for my $link_value (
 511                         sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 512                         keys %$facets_counter
 513                       )
 514                     {
 515                         my $expandable;
 516                         my $number_of_facets;
 517                         my @this_facets_array;
 518                         for my $one_facet (
 519                             sort {
 520                                 $facets_counter->{$link_value}
 521                                   ->{$b} <=> $facets_counter->{$link_value}
 522                                   ->{$a}
 523                             } keys %{ $facets_counter->{$link_value} }
 524                           )
 525                         {
 526                             $number_of_facets++;
 527                             if (   ( $number_of_facets <= 5 )
 528                                 || ( $expanded_facet eq $link_value )
 529                                 || ( $facets_info->{$link_value}->{'expanded'} )
 530                               )
 531                             {
 532
 533 # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL
 534                                 my $facet_link_value = $one_facet;
 535                                 $facet_link_value =~ s/[()!?¡¿؟]/ /g;
 536
 537                                 # fix the length that will display in the label,
 538                                 my $facet_label_value = $one_facet;
 539                                 my $facet_max_length  = C4::Context->preference(
 540                                     'FacetLabelTruncationLength')
 541                                   || 20;
 542                                 $facet_label_value =
 543                                   substr( $one_facet, 0, $facet_max_length )
 544                                   . "..."
 545                                   if length($facet_label_value) >
 546                                       $facet_max_length;
 547
 548                             # if it's a branch, label by the name, not the code,
 549                                 if ( $link_value =~ /branch/ ) {
 550                                     if (   defined $branches
 551                                         && ref($branches) eq "HASH"
 552                                         && defined $branches->{$one_facet}
 553                                         && ref( $branches->{$one_facet} ) eq
 554                                         "HASH" )
 555                                     {
 556                                         $facet_label_value =
 557                                           $branches->{$one_facet}
 558                                           ->{'branchname'};
 559                                     }
 560                                     else {
 561                                         $facet_label_value = "*";
 562                                     }
 563                                 }
 564
 565                           # if it's a itemtype, label by the name, not the code,
 566                                 if ( $link_value =~ /itype/ ) {
 567                                     if (   defined $itemtypes
 568                                         && ref($itemtypes) eq "HASH"
 569                                         && defined $itemtypes->{$one_facet}
 570                                         && ref( $itemtypes->{$one_facet} ) eq
 571                                         "HASH" )
 572                                     {
 573                                         $facet_label_value =
 574                                           $itemtypes->{$one_facet}
 575                                           ->{'description'};
 576                                     }
 577                                 }
 578
 579                # also, if it's a location code, use the name instead of the code
 580                                 if ( $link_value =~ /location/ ) {
 581                                     $facet_label_value =
 582                                       GetKohaAuthorisedValueLib( 'LOC',
 583                                         $one_facet, $opac );
 584                                 }
 585
 586                 # but we're down with the whole label being in the link's title.
 587                                 push @this_facets_array,
 588                                   {
 589                                     facet_count =>
 590                                       $facets_counter->{$link_value}
 591                                       ->{$one_facet},
 592                                     facet_label_value => $facet_label_value,
 593                                     facet_title_value => $one_facet,
 594                                     facet_link_value  => $facet_link_value,
 595                                     type_link_value   => $link_value,
 596                                   }
 597                                   if ($facet_label_value);
 598                             }
 599                         }
 600
 601                         # handle expanded option
 602                         unless ( $facets_info->{$link_value}->{'expanded'} ) {
 603                             $expandable = 1
 604                               if ( ( $number_of_facets > 5 )
 605                                 && ( $expanded_facet ne $link_value ) );
 606                         }
 607                         push @facets_loop,
 608                           {
 609                             type_link_value => $link_value,
 610                             type_id         => $link_value . "_id",
 611                             "type_label_"
 612                               . $facets_info->{$link_value}->{'label_value'} =>
 613                               1,
 614                             facets     => \@this_facets_array,
 615                             expandable => $expandable,
 616                             expand     => $link_value,
 617                           }
 618                           unless (
 619                             (
 620                                 $facets_info->{$link_value}->{'label_value'} =~
 621                                 /Libraries/
 622                             )
 623                             and ( C4::Context->preference('singleBranchMode') )
 624                           );
 625                     }
 626                 }
 627             }
 628         );
 629     return ( undef, $results_hashref, \@facets_loop );
 630 }
 631
 632 sub GetFacets {
 633
 634     my $rs = shift;
 635     my $facets;
 636
 637     my $indexing_mode    = C4::Context->config('zebra_bib_index_mode') // 'dom';
 638     my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0;
 639
 640     if ( $indexing_mode eq 'dom' &&
 641          $use_zebra_facets ) {
 642         $facets = _get_facets_from_zebra( $rs );
 643     } else {
 644         $facets = _get_facets_from_records( $rs );
 645     }
 646
 647     return $facets;
 648 }
 649
 650 sub _get_facets_from_records {
 651
 652     my $rs = shift;
 653
 654     my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20;
 655     my $facets_config  = getFacets();
 656     my $facets         = {};
 657     my $size           = $rs->size();
 658     my $jmax           = $size > $facets_maxrecs
 659                             ? $facets_maxrecs
 660                             : $size;
 661
 662     for ( my $j = 0 ; $j < $jmax ; $j++ ) {
 663
 664         my $marc_record = new_record_from_zebra (
 665                 'biblioserver',
 666                 $rs->record( $j )->raw()
 667         );
 668
 669         if ( ! defined $marc_record ) {
 670             warn "ERROR DECODING RECORD - $@: " .
 671                 $rs->record( $j )->raw();
 672             next;
 673         }
 674
 675         _get_facets_data_from_record( $marc_record, $facets_config, $facets );
 676     }
 677
 678     return $facets;
 679 }
 680
 681 =head2 _get_facets_data_from_record
 682
 683     C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter );
 684
 685 Internal function that extracts facets information from a MARC::Record object
 686 and populates $facets_counter for using in getRecords.
 687
 688 $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured
 689 facets for Zebra).
 690
 691 =cut
 692
 693 sub _get_facets_data_from_record {
 694
 695     my ( $marc_record, $facets, $facets_counter ) = @_;
 696
 697     for my $facet (@$facets) {
 698
 699         my @used_datas = ();
 700
 701         foreach my $tag ( @{ $facet->{ tags } } ) {
 702
 703             # tag number is the first three digits
 704             my $tag_num          = substr( $tag, 0, 3 );
 705             # subfields are the remainder
 706             my $subfield_letters = substr( $tag, 3 );
 707
 708             my @fields = $marc_record->field( $tag_num );
 709             foreach my $field (@fields) {
 710                 # If $field->indicator(1) eq 'z', it means it is a 'see from'
 711                 # field introduced because of IncludeSeeFromInSearches, so skip it
 712                 next if $field->indicator(1) eq 'z';
 713
 714                 my $data = $field->as_string( $subfield_letters, $facet->{ sep } );
 715
 716                 unless ( grep { /^\Q$data\E$/ } @used_datas ) {
 717                     push @used_datas, $data;
 718                     $facets_counter->{ $facet->{ idx } }->{ $data }++;
 719                 }
 720             }
 721         }
 722     }
 723 }
 724
 725 =head2 _get_facets_from_zebra
 726
 727     my $facets = _get_facets_from_zebra( $result_set )
 728
 729 Retrieves facets for a specified result set. It loops through the facets defined
 730 in C4::Koha::getFacets and returns a hash with the following structure:
 731
 732    {  facet_idx => {
 733             facet_value => count
 734       },
 735       ...
 736    }
 737
 738 =cut
 739
 740 sub _get_facets_from_zebra {
 741
 742     my $rs = shift;
 743
 744     # save current elementSetName
 745     my $elementSetName = $rs->option( 'elementSetName' );
 746
 747     my $facets_loop = getFacets();
 748     my $facets_data  = {};
 749     # loop through defined facets and fill the facets hashref
 750     foreach my $facet ( @$facets_loop ) {
 751
 752         my $idx = $facet->{ idx };
 753         my $sep = $facet->{ sep };
 754         my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep );
 755         if ( $facet_values ) {
 756             # we've actually got a result
 757             $facets_data->{ $idx } = $facet_values;
 758         }
 759     }
 760     # set elementSetName to its previous value to avoid side effects
 761     $rs->option( elementSetName => $elementSetName );
 762
 763     return $facets_data;
 764 }
 765
 766 =head2 _get_facet_from_result_set
 767
 768     my $facet_values =
 769         C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep )
 770
 771 Internal function that extracts facet information for a specific index ($facet_idx) and
 772 returns a hash containing facet values and count:
 773
 774     {
 775         $facet_value => $count ,
 776         ...
 777     }
 778
 779 Warning: this function has the side effect of changing the elementSetName for the result
 780 set. It is a helper function for the main loop, which takes care of backing it up for
 781 restoring.
 782
 783 =cut
 784
 785 sub _get_facet_from_result_set {
 786
 787     my $facet_idx = shift;
 788     my $rs        = shift;
 789     my $sep       = shift;
 790
 791     my $internal_sep = '<*>';
 792
 793     return if ( ! defined $facet_idx || ! defined $rs );
 794     # zebra's facet element, untokenized index
 795     my $facet_element = 'zebra::facet::' . $facet_idx . ':0:100';
 796     # configure zebra results for retrieving the desired facet
 797     $rs->option( elementSetName => $facet_element );
 798     # get the facet record from result set
 799     my $facet = $rs->record( 0 )->raw;
 800     # if the facet has no restuls...
 801     return if !defined $facet;
 802     # TODO: benchmark DOM vs. SAX performance
 803     my $facet_dom = XML::LibXML->load_xml(
 804       string => ($facet)
 805     );
 806     my @terms = $facet_dom->getElementsByTagName('term');
 807     return if ! @terms;
 808
 809     my $facets = {};
 810     foreach my $term ( @terms ) {
 811         my $facet_value = $term->textContent;
 812         $facet_value =~ s/\Q$internal_sep\E/$sep/ if defined $sep;
 813         $facets->{ $facet_value } = $term->getAttribute( 'occur' );
 814     }
 815
 816     return $facets;
 817 }
 818
 819 =head2 _get_facets_info
 820
 821     my $facets_info = C4::Search::_get_facets_info( $facets )
 822
 823 Internal function that extracts facets information and properly builds
 824 the data structure needed to render facet labels.
 825
 826 =cut
 827
 828 sub _get_facets_info {
 829
 830     my $facets = shift;
 831
 832     my $facets_info = {};
 833
 834     for my $facet ( @$facets ) {
 835         $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label };
 836         $facets_info->{ $facet->{ idx } }->{ expanded }    = $facet->{ expanded };
 837     }
 838
 839     return $facets_info;
 840 }
 841
 842 sub pazGetRecords {
 843     my (
 844         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 845         $results_per_page, $offset,       $expanded_facet, $branches,
 846         $query_type,       $scan
 847     ) = @_;
 848
 849     my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
 850     $paz->init();
 851     $paz->search($simple_query);
 852     sleep 1;   # FIXME: WHY?
 853
 854     # do results
 855     my $results_hashref = {};
 856     my $stats = XMLin($paz->stat);
 857     my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
 858
 859     # for a grouped search result, the number of hits
 860     # is the number of groups returned; 'bib_hits' will have
 861     # the total number of bibs.
 862     $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
 863     $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
 864
 865     HIT: foreach my $hit (@{ $results->{'hit'} }) {
 866         my $recid = $hit->{recid}->[0];
 867
 868         my $work_title = $hit->{'md-work-title'}->[0];
 869         my $work_author;
 870         if (exists $hit->{'md-work-author'}) {
 871             $work_author = $hit->{'md-work-author'}->[0];
 872         }
 873         my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
 874
 875         my $result_group = {};
 876         $result_group->{'group_label'} = $group_label;
 877         $result_group->{'group_merge_key'} = $recid;
 878
 879         my $count = 1;
 880         if (exists $hit->{count}) {
 881             $count = $hit->{count}->[0];
 882         }
 883         $result_group->{'group_count'} = $count;
 884
 885         for (my $i = 0; $i < $count; $i++) {
 886             # FIXME -- may need to worry about diacritics here
 887             my $rec = $paz->record($recid, $i);
 888             push @{ $result_group->{'RECORDS'} }, $rec;
 889         }
 890
 891         push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
 892     }
 893
 894     # pass through facets
 895     my $termlist_xml = $paz->termlist('author,subject');
 896     my $terms = XMLin($termlist_xml, forcearray => 1);
 897     my @facets_loop = ();
 898     #die Dumper($results);
 899 #    foreach my $list (sort keys %{ $terms->{'list'} }) {
 900 #        my @facets = ();
 901 #        foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
 902 #            push @facets, {
 903 #                facet_label_value => $facet->{'name'}->[0],
 904 #            };
 905 #        }
 906 #        push @facets_loop, ( {
 907 #            type_label => $list,
 908 #            facets => \@facets,
 909 #        } );
 910 #    }
 911
 912     return ( undef, $results_hashref, \@facets_loop );
 913 }
 914
 915 # STOPWORDS
 916 sub _remove_stopwords {
 917     my ( $operand, $index ) = @_;
 918     my @stopwords_removed;
 919
 920     # phrase and exact-qualified indexes shouldn't have stopwords removed
 921     if ( $index !~ m/,(phr|ext)/ ) {
 922
 923 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
 924 #       we use IsAlpha unicode definition, to deal correctly with diacritics.
 925 #       otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
 926 #       is a stopword, we'd get "çon" and wouldn't find anything...
 927 #
 928                 foreach ( keys %{ C4::Context->stopwords } ) {
 929                         next if ( $_ =~ /(and|or|not)/ );    # don't remove operators
 930                         if ( my ($matched) = ($operand =~
 931                                 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
 932                         {
 933                                 $operand =~ s/\Q$matched\E/ /gi;
 934                                 push @stopwords_removed, $_;
 935                         }
 936                 }
 937         }
 938     return ( $operand, \@stopwords_removed );
 939 }
 940
 941 # TRUNCATION
 942 sub _detect_truncation {
 943     my ( $operand, $index ) = @_;
 944     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 945         @regexpr );
 946     $operand =~ s/^ //g;
 947     my @wordlist = split( /\s/, $operand );
 948     foreach my $word (@wordlist) {
 949         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 950             push @rightlefttruncated, $word;
 951         }
 952         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 953             push @lefttruncated, $word;
 954         }
 955         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 956             push @righttruncated, $word;
 957         }
 958         elsif ( index( $word, "*" ) < 0 ) {
 959             push @nontruncated, $word;
 960         }
 961         else {
 962             push @regexpr, $word;
 963         }
 964     }
 965     return (
 966         \@nontruncated,       \@righttruncated, \@lefttruncated,
 967         \@rightlefttruncated, \@regexpr
 968     );
 969 }
 970
 971 # STEMMING
 972 sub _build_stemmed_operand {
 973     my ($operand,$lang) = @_;
 974     require Lingua::Stem::Snowball ;
 975     my $stemmed_operand=q{};
 976
 977     # If operand contains a digit, it is almost certainly an identifier, and should
 978     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 979     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 980     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 981     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 982     return $operand if $operand =~ /\d/;
 983
 984 # FIXME: the locale should be set based on the user's language and/or search choice
 985     #warn "$lang";
 986     # Make sure we only use the first two letters from the language code
 987     $lang = lc(substr($lang, 0, 2));
 988     # The language codes for the two variants of Norwegian will now be "nb" and "nn",
 989     # none of which Lingua::Stem::Snowball can use, so we need to "translate" them
 990     if ($lang eq 'nb' || $lang eq 'nn') {
 991       $lang = 'no';
 992     }
 993     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 994                                                encoding => "UTF-8" );
 995
 996     my @words = split( / /, $operand );
 997     my @stems = $stemmer->stem(\@words);
 998     for my $stem (@stems) {
 999         $stemmed_operand .= "$stem";
1000         $stemmed_operand .= "?"
1001           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
1002         $stemmed_operand .= " ";
1003     }
1004     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
1005     return $stemmed_operand;
1006 }
1007
1008 # FIELD WEIGHTING
1009 sub _build_weighted_query {
1010
1011 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
1012 # pretty well but could work much better if we had a smarter query parser
1013     my ( $operand, $stemmed_operand, $index ) = @_;
1014     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
1015     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
1016     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
1017     $operand =~ s/"/ /g;    # Bug 7518: searches with quotation marks don't work
1018
1019     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
1020
1021     # Keyword, or, no index specified
1022     if ( ( $index eq 'kw' ) || ( !$index ) ) {
1023         $weighted_query .=
1024           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
1025         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
1026         $weighted_query .= " or Title-cover,phr,r3=\"$operand\"";    # phrase title
1027         $weighted_query .= " or ti,wrdl,r4=\"$operand\"";    # words in title
1028           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
1029           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
1030         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
1031           if $fuzzy_enabled;    # add fuzzy, word list
1032         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
1033           if ( $stemming and $stemmed_operand )
1034           ;                     # add stemming, right truncation
1035         $weighted_query .= " or wrdl,r9=\"$operand\"";
1036
1037         # embedded sorting: 0 a-z; 1 z-a
1038         # $weighted_query .= ") or (sort1,aut=1";
1039     }
1040
1041     # Barcode searches should skip this process
1042     elsif ( $index eq 'bc' ) {
1043         $weighted_query .= "bc=\"$operand\"";
1044     }
1045
1046     # Authority-number searches should skip this process
1047     elsif ( $index eq 'an' ) {
1048         $weighted_query .= "an=\"$operand\"";
1049     }
1050
1051     # If the index already has more than one qualifier, wrap the operand
1052     # in quotes and pass it back (assumption is that the user knows what they
1053     # are doing and won't appreciate us mucking up their query
1054     elsif ( $index =~ ',' ) {
1055         $weighted_query .= " $index=\"$operand\"";
1056     }
1057
1058     #TODO: build better cases based on specific search indexes
1059     else {
1060         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
1061           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
1062         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
1063         $weighted_query .= " or $index,wrdl,r6=\"$operand\"";    # word list index
1064         $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\""
1065           if $fuzzy_enabled;    # add fuzzy, word list
1066         $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\""
1067           if ( $stemming and $stemmed_operand );    # add stemming, right truncation
1068     }
1069
1070     $weighted_query .= "))";                       # close rank specification
1071     return $weighted_query;
1072 }
1073
1074 =head2 getIndexes
1075
1076 Return an array with available indexes.
1077
1078 =cut
1079
1080 sub getIndexes{
1081     my @indexes = (
1082                     # biblio indexes
1083                     'ab',
1084                     'Abstract',
1085                     'acqdate',
1086                     'allrecords',
1087                     'an',
1088                     'Any',
1089                     'at',
1090                     'au',
1091                     'aub',
1092                     'aud',
1093                     'audience',
1094                     'auo',
1095                     'aut',
1096                     'Author',
1097                     'Author-in-order ',
1098                     'Author-personal-bibliography',
1099                     'Authority-Number',
1100                     'authtype',
1101                     'bc',
1102                     'Bib-level',
1103                     'biblionumber',
1104                     'bio',
1105                     'biography',
1106                     'callnum',
1107                     'cfn',
1108                     'Chronological-subdivision',
1109                     'cn-bib-source',
1110                     'cn-bib-sort',
1111                     'cn-class',
1112                     'cn-item',
1113                     'cn-prefix',
1114                     'cn-suffix',
1115                     'cpn',
1116                     'Code-institution',
1117                     'Conference-name',
1118                     'Conference-name-heading',
1119                     'Conference-name-see',
1120                     'Conference-name-seealso',
1121                     'Content-type',
1122                     'Control-number',
1123                     'copydate',
1124                     'Corporate-name',
1125                     'Corporate-name-heading',
1126                     'Corporate-name-see',
1127                     'Corporate-name-seealso',
1128                     'Country-publication',
1129                     'ctype',
1130                     'curriculum',
1131                     'date-entered-on-file',
1132                     'Date-of-acquisition',
1133                     'Date-of-publication',
1134                     'Dewey-classification',
1135                     'Dissertation-information',
1136                     'EAN',
1137                     'extent',
1138                     'fic',
1139                     'fiction',
1140                     'Form-subdivision',
1141                     'format',
1142                     'Geographic-subdivision',
1143                     'he',
1144                     'Heading',
1145                     'Heading-use-main-or-added-entry',
1146                     'Heading-use-series-added-entry ',
1147                     'Heading-use-subject-added-entry',
1148                     'Host-item',
1149                     'id-other',
1150                     'Illustration-code',
1151                     'Index-term-genre',
1152                     'Index-term-uncontrolled',
1153                     'ISBN',
1154                     'isbn',
1155                     'ISSN',
1156                     'issn',
1157                     'itemtype',
1158                     'kw',
1159                     'Koha-Auth-Number',
1160                     'l-format',
1161                     'language',
1162                     'language-original',
1163                     'lc-card',
1164                     'LC-card-number',
1165                     'lcn',
1166                     'lex',
1167                     'llength',
1168                     'ln',
1169                     'ln-audio',
1170                     'ln-subtitle',
1171                     'Local-classification',
1172                     'Local-number',
1173                     'Match-heading',
1174                     'Match-heading-see-from',
1175                     'Material-type',
1176                     'mc-itemtype',
1177                     'mc-rtype',
1178                     'mus',
1179                     'name',
1180                     'Music-number',
1181                     'Name-geographic',
1182                     'Name-geographic-heading',
1183                     'Name-geographic-see',
1184                     'Name-geographic-seealso',
1185                     'nb',
1186                     'Note',
1187                     'notes',
1188                     'ns',
1189                     'nt',
1190                     'pb',
1191                     'Personal-name',
1192                     'Personal-name-heading',
1193                     'Personal-name-see',
1194                     'Personal-name-seealso',
1195                     'pl',
1196                     'Place-publication',
1197                     'pn',
1198                     'popularity',
1199                     'pubdate',
1200                     'Publisher',
1201                     'Record-control-number',
1202                     'rcn',
1203                     'Record-type',
1204                     'rtype',
1205                     'se',
1206                     'See',
1207                     'See-also',
1208                     'sn',
1209                     'Stock-number',
1210                     'su',
1211                     'Subject',
1212                     'Subject-heading-thesaurus',
1213                     'Subject-name-personal',
1214                     'Subject-subdivision',
1215                     'Summary',
1216                     'Suppress',
1217                     'su-geo',
1218                     'su-na',
1219                     'su-to',
1220                     'su-ut',
1221                     'ut',
1222                     'Term-genre-form',
1223                     'Term-genre-form-heading',
1224                     'Term-genre-form-see',
1225                     'Term-genre-form-seealso',
1226                     'ti',
1227                     'Title',
1228                     'Title-cover',
1229                     'Title-series',
1230                     'Title-uniform',
1231                     'Title-uniform-heading',
1232                     'Title-uniform-see',
1233                     'Title-uniform-seealso',
1234                     'totalissues',
1235                     'yr',
1236
1237                     # items indexes
1238                     'acqsource',
1239                     'barcode',
1240                     'bc',
1241                     'branch',
1242                     'ccode',
1243                     'classification-source',
1244                     'cn-sort',
1245                     'coded-location-qualifier',
1246                     'copynumber',
1247                     'damaged',
1248                     'datelastborrowed',
1249                     'datelastseen',
1250                     'holdingbranch',
1251                     'homebranch',
1252                     'issues',
1253                     'item',
1254                     'itemnumber',
1255                     'itype',
1256                     'Local-classification',
1257                     'location',
1258                     'lost',
1259                     'materials-specified',
1260                     'mc-ccode',
1261                     'mc-itype',
1262                     'mc-loc',
1263                     'notforloan',
1264                     'Number-local-acquisition',
1265                     'onloan',
1266                     'price',
1267                     'renewals',
1268                     'replacementprice',
1269                     'replacementpricedate',
1270                     'reserves',
1271                     'restricted',
1272                     'stack',
1273                     'stocknumber',
1274                     'inv',
1275                     'uri',
1276                     'withdrawn',
1277
1278                     # subject related
1279                   );
1280
1281     return \@indexes;
1282 }
1283
1284 =head2 _handle_exploding_index
1285
1286     my $query = _handle_exploding_index($index, $term)
1287
1288 Callback routine to generate the search for "exploding" indexes (i.e.
1289 those indexes which are turned into multiple or-connected searches based
1290 on authority data).
1291
1292 =cut
1293
1294 sub _handle_exploding_index {
1295     my ($QParser, $filter, $params, $negate, $server) = @_;
1296     my $index = $filter;
1297     my $term = join(' ', @$params);
1298
1299     return unless ($index =~ m/(su-br|su-na|su-rl)/ && $term);
1300
1301     my $marcflavour = C4::Context->preference('marcflavour');
1302
1303     my $codesubfield = $marcflavour eq 'UNIMARC' ? '5' : 'w';
1304     my $wantedcodes = '';
1305     my @subqueries = ( "\@attr 1=Subject \@attr 4=1 \"$term\"");
1306     my ($error, $results, $total_hits) = SimpleSearch( "he:$term", undef, undef, [ "authorityserver" ] );
1307     foreach my $auth (@$results) {
1308         my $record = MARC::Record->new_from_usmarc($auth);
1309         my @references = $record->field('5..');
1310         if (@references) {
1311             if ($index eq 'su-br') {
1312                 $wantedcodes = 'g';
1313             } elsif ($index eq 'su-na') {
1314                 $wantedcodes = 'h';
1315             } elsif ($index eq 'su-rl') {
1316                 $wantedcodes = '';
1317             }
1318             foreach my $reference (@references) {
1319                 my $codes = $reference->subfield($codesubfield);
1320                 push @subqueries, '@attr 1=Subject @attr 4=1 "' . $reference->as_string('abcdefghijlmnopqrstuvxyz') . '"' if (($codes && $codes eq $wantedcodes) || !$wantedcodes);
1321             }
1322         }
1323     }
1324     my $query = ' @or ' x (scalar(@subqueries) - 1) . join(' ', @subqueries);
1325     return $query;
1326 }
1327
1328 =head2 parseQuery
1329
1330     ( $operators, $operands, $indexes, $limits,
1331       $sort_by, $scan, $lang ) =
1332             buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1333
1334 Shim function to ease the transition from buildQuery to a new QueryParser.
1335 This function is called at the beginning of buildQuery, and modifies
1336 buildQuery's input. If it can handle the input, it returns a query that
1337 buildQuery will not try to parse.
1338 =cut
1339
1340 sub parseQuery {
1341     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1342
1343     my @operators = $operators ? @$operators : ();
1344     my @indexes   = $indexes   ? @$indexes   : ();
1345     my @operands  = $operands  ? @$operands  : ();
1346     my @limits    = $limits    ? @$limits    : ();
1347     my @sort_by   = $sort_by   ? @$sort_by   : ();
1348
1349     my $query = $operands[0];
1350     my $index;
1351     my $term;
1352     my $query_desc;
1353
1354     my $QParser;
1355     $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser') || $query =~ s/^qp=//);
1356     undef $QParser if ($query =~ m/^(ccl=|pqf=|cql=)/ || grep (/\w,\w|\w=\w/, @operands, @indexes) );
1357     undef $QParser if (scalar @limits > 0);
1358
1359     if ($QParser)
1360     {
1361         $QParser->custom_data->{'QueryAutoTruncate'} = C4::Context->preference('QueryAutoTruncate');
1362         $query = '';
1363         for ( my $ii = 0 ; $ii <= @operands ; $ii++ ) {
1364             next unless $operands[$ii];
1365             $query .= $operators[ $ii - 1 ] eq 'or' ? ' || ' : ' && '
1366               if ($query);
1367             if ( $operands[$ii] =~ /^[^"]\W*[-|_\w]*:\w.*[^"]$/ ) {
1368                 $query .= $operands[$ii];
1369             }
1370             elsif ( $indexes[$ii] =~ m/su-/ ) {
1371                 $query .= $indexes[$ii] . '(' . $operands[$ii] . ')';
1372             }
1373             else {
1374                 $query .=
1375                   ( $indexes[$ii] ? "$indexes[$ii]:" : '' ) . $operands[$ii];
1376             }
1377         }
1378         foreach my $limit (@limits) {
1379         }
1380         if ( scalar(@sort_by) > 0 ) {
1381             my $modifier_re =
1382               '#(' . join( '|', @{ $QParser->modifiers } ) . ')';
1383             $query =~ s/$modifier_re//g;
1384             foreach my $modifier (@sort_by) {
1385                 $query .= " #$modifier";
1386             }
1387         }
1388
1389         $query_desc = $query;
1390         $query_desc =~ s/\s+/ /g;
1391         if ( C4::Context->preference("QueryWeightFields") ) {
1392         }
1393         $QParser->add_bib1_filter_map( 'su-br' => 'biblioserver' =>
1394               { 'target_syntax_callback' => \&_handle_exploding_index } );
1395         $QParser->add_bib1_filter_map( 'su-na' => 'biblioserver' =>
1396               { 'target_syntax_callback' => \&_handle_exploding_index } );
1397         $QParser->add_bib1_filter_map( 'su-rl' => 'biblioserver' =>
1398               { 'target_syntax_callback' => \&_handle_exploding_index } );
1399         $QParser->parse($query);
1400         $operands[0] = "pqf=" . $QParser->target_syntax('biblioserver');
1401     }
1402     else {
1403         require Koha::QueryParser::Driver::PQF;
1404         my $modifier_re = '#(' . join( '|', @{Koha::QueryParser::Driver::PQF->modifiers}) . ')';
1405         s/$modifier_re//g for @operands;
1406     }
1407
1408     return ( $operators, \@operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc);
1409 }
1410
1411 =head2 buildQuery
1412
1413 ( $error, $query,
1414 $simple_query, $query_cgi,
1415 $query_desc, $limit,
1416 $limit_cgi, $limit_desc,
1417 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1418
1419 Build queries and limits in CCL, CGI, Human,
1420 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1421
1422 See verbose embedded documentation.
1423
1424
1425 =cut
1426
1427 sub buildQuery {
1428     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1429
1430     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1431
1432     my $query_desc;
1433     ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang, $query_desc) = parseQuery($operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1434
1435     # dereference
1436     my @operators = $operators ? @$operators : ();
1437     my @indexes   = $indexes   ? @$indexes   : ();
1438     my @operands  = $operands  ? @$operands  : ();
1439     my @limits    = $limits    ? @$limits    : ();
1440     my @sort_by   = $sort_by   ? @$sort_by   : ();
1441
1442     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1443     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1444     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1445     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1446     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1447
1448     my $query        = $operands[0];
1449     my $simple_query = $operands[0];
1450
1451     # initialize the variables we're passing back
1452     my $query_cgi;
1453     my $query_type;
1454
1455     my $limit;
1456     my $limit_cgi;
1457     my $limit_desc;
1458
1459     my $stopwords_removed;    # flag to determine if stopwords have been removed
1460
1461     my $cclq       = 0;
1462     my $cclindexes = getIndexes();
1463     if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) {
1464         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1465             my $dx = lc($1);
1466             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1467         }
1468         $query = "ccl=$query" if $cclq;
1469     }
1470
1471 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1472 # DIAGNOSTIC ONLY!!
1473     if ( $query =~ /^ccl=/ ) {
1474         my $q=$';
1475         # This is needed otherwise ccl= and &limit won't work together, and
1476         # this happens when selecting a subject on the opac-detail page
1477         @limits = grep {!/^$/} @limits;
1478         if ( @limits ) {
1479             $q .= ' and '.join(' and ', @limits);
1480         }
1481         return ( undef, $q, $q, "q=ccl=".uri_escape($q), $q, '', '', '', '', 'ccl' );
1482     }
1483     if ( $query =~ /^cql=/ ) {
1484         return ( undef, $', $', "q=cql=".uri_escape($'), $', '', '', '', '', 'cql' );
1485     }
1486     if ( $query =~ /^pqf=/ ) {
1487         if ($query_desc) {
1488             $query_cgi = "q=".uri_escape($query_desc);
1489         } else {
1490             $query_desc = $';
1491             $query_cgi = "q=pqf=".uri_escape($');
1492         }
1493         return ( undef, $', $', $query_cgi, $query_desc, '', '', '', '', 'pqf' );
1494     }
1495
1496     # pass nested queries directly
1497     # FIXME: need better handling of some of these variables in this case
1498     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1499     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1500 #    if ( $query =~ /(\(|\))/ ) {
1501 #        return (
1502 #            undef,              $query, $simple_query, $query_cgi,
1503 #            $query,             $limit, $limit_cgi,    $limit_desc,
1504 #            $stopwords_removed, 'ccl'
1505 #        );
1506 #    }
1507
1508 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1509 # query operands and indexes and add stemming, truncation, field weighting, etc.
1510 # Once we do so, we'll end up with a value in $query, just like if we had an
1511 # incoming $query from the user
1512     else {
1513         $query = ""
1514           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1515         my $previous_operand
1516           ;    # a flag used to keep track if there was a previous query
1517                # if there was, we can apply the current operator
1518                # for every operand
1519         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1520
1521             # COMBINE OPERANDS, INDEXES AND OPERATORS
1522             if ( $operands[$i] ) {
1523                 $operands[$i]=~s/^\s+//;
1524
1525               # A flag to determine whether or not to add the index to the query
1526                 my $indexes_set;
1527
1528 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1529                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1530                     $weight_fields    = 0;
1531                     $stemming         = 0;
1532                     $remove_stopwords = 0;
1533                 } else {
1534                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1535                 }
1536                 my $operand = $operands[$i];
1537                 my $index   = $indexes[$i];
1538
1539                 # Add index-specific attributes
1540
1541                 #Afaik, this 'yr' condition will only ever be met in the staff client advanced search
1542                 #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query,
1543                 #which is processed higher up in this sub. Other than that, year searches are typically
1544                 #handled as limits which are not processed her either.
1545
1546                 # Date of Publication
1547                 if ( $index =~ /yr/ ) {
1548                     #weight_fields/relevance search causes errors with date ranges
1549                     #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range)
1550                     #In the case of YYYY-YYYY, it will return no results
1551                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1552                 }
1553
1554                 # Date of Acquisition
1555                 elsif ( $index =~ /acqdate/ ) {
1556                     #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format
1557                     #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the
1558                       #top of the results just because they have lots of item records matching that date.
1559                     #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so
1560                       #irrelevant here
1561                     #remove_stopwords doesn't function anymore so is irrelevant
1562                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1563                 }
1564                 # ISBN,ISSN,Standard Number, don't need special treatment
1565                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1566                     (
1567                         $stemming,      $auto_truncation,
1568                         $weight_fields, $fuzzy_enabled,
1569                         $remove_stopwords
1570                     ) = ( 0, 0, 0, 0, 0 );
1571
1572                 }
1573
1574                 if(not $index){
1575                     $index = 'kw';
1576                 }
1577
1578                 # Set default structure attribute (word list)
1579                 my $struct_attr = q{};
1580                 unless ( $indexes_set || !$index || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) {
1581                     $struct_attr = ",wrdl";
1582                 }
1583
1584                 # Some helpful index variants
1585                 my $index_plus       = $index . $struct_attr . ':';
1586                 my $index_plus_comma = $index . $struct_attr . ',';
1587
1588                 # Remove Stopwords
1589                 if ($remove_stopwords) {
1590                     ( $operand, $stopwords_removed ) =
1591                       _remove_stopwords( $operand, $index );
1592                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1593                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1594                       if ( $stopwords_removed && $DEBUG );
1595                 }
1596
1597                 if ($auto_truncation){
1598                         unless ( $index =~ /,(st-|phr|ext)/ ) {
1599                                                 #FIXME only valid with LTR scripts
1600                                                 $operand=join(" ",map{
1601                                                                                         (index($_,"*")>0?"$_":"$_*")
1602                                                                                          }split (/\s+/,$operand));
1603                                                 warn $operand if $DEBUG;
1604                                         }
1605                                 }
1606
1607                 # Detect Truncation
1608                 my $truncated_operand;
1609                 my( $nontruncated, $righttruncated, $lefttruncated,
1610                     $rightlefttruncated, $regexpr
1611                 ) = _detect_truncation( $operand, $index );
1612                 warn
1613 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1614                   if $DEBUG;
1615
1616                 # Apply Truncation
1617                 if (
1618                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1619                     scalar(@$rightlefttruncated) > 0 )
1620                 {
1621
1622                # Don't field weight or add the index to the query, we do it here
1623                     $indexes_set = 1;
1624                     undef $weight_fields;
1625                     my $previous_truncation_operand;
1626                     if (scalar @$nontruncated) {
1627                         $truncated_operand .= "$index_plus @$nontruncated ";
1628                         $previous_truncation_operand = 1;
1629                     }
1630                     if (scalar @$righttruncated) {
1631                         $truncated_operand .= "and " if $previous_truncation_operand;
1632                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1633                         $previous_truncation_operand = 1;
1634                     }
1635                     if (scalar @$lefttruncated) {
1636                         $truncated_operand .= "and " if $previous_truncation_operand;
1637                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1638                         $previous_truncation_operand = 1;
1639                     }
1640                     if (scalar @$rightlefttruncated) {
1641                         $truncated_operand .= "and " if $previous_truncation_operand;
1642                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1643                         $previous_truncation_operand = 1;
1644                     }
1645                 }
1646                 $operand = $truncated_operand if $truncated_operand;
1647                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1648
1649                 # Handle Stemming
1650                 my $stemmed_operand;
1651                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1652                                                                                 if $stemming;
1653
1654                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1655
1656                 # Handle Field Weighting
1657                 my $weighted_operand;
1658                 if ($weight_fields) {
1659                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1660                     $operand = $weighted_operand;
1661                     $indexes_set = 1;
1662                 }
1663
1664                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1665
1666                 ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({
1667                     query => $query,
1668                     query_cgi => $query_cgi,
1669                     query_desc => $query_desc,
1670                     operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '',
1671                     parsed_operand => $operand,
1672                     original_operand => ($operands[$i]) ? $operands[$i] : '',
1673                     index => $index,
1674                     index_plus => $index_plus,
1675                     indexes_set => $indexes_set,
1676                     previous_operand => $previous_operand,
1677                 });
1678
1679             }    #/if $operands
1680         }    # /for
1681     }
1682     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1683
1684     # add limits
1685     my %group_OR_limits;
1686     my $availability_limit;
1687     foreach my $this_limit (@limits) {
1688         next unless $this_limit;
1689         if ( $this_limit =~ /available/ ) {
1690 #
1691 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1692 ## In English:
1693 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1694             $availability_limit .=
1695 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1696             $limit_cgi  .= "&limit=available";
1697             $limit_desc .= "";
1698         }
1699
1700         # group_OR_limits, prefixed by mc-
1701         # OR every member of the group
1702         elsif ( $this_limit =~ /mc/ ) {
1703             my ($k,$v) = split(/:/, $this_limit,2);
1704             if ( $k !~ /mc-i(tem)?type/ ) {
1705                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1706                 $this_limit =~ tr/"//d;
1707                 $this_limit = $k.":\"".$v."\"";
1708             }
1709
1710             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1711             $limit_desc      .= " or " if $group_OR_limits{$k};
1712             $group_OR_limits{$k} .= "$this_limit";
1713             $limit_cgi       .= "&limit=" . uri_escape($this_limit);
1714             $limit_desc      .= " $this_limit";
1715         }
1716
1717         # Regular old limits
1718         else {
1719             $limit .= " and " if $limit || $query;
1720             $limit      .= "$this_limit";
1721             $limit_cgi  .= "&limit=" . uri_escape($this_limit);
1722             if ($this_limit =~ /^branch:(.+)/) {
1723                 my $branchcode = $1;
1724                 my $branchname = GetBranchName($branchcode);
1725                 if (defined $branchname) {
1726                     $limit_desc .= " branch:$branchname";
1727                 } else {
1728                     $limit_desc .= " $this_limit";
1729                 }
1730             } else {
1731                 $limit_desc .= " $this_limit";
1732             }
1733         }
1734     }
1735     foreach my $k (keys (%group_OR_limits)) {
1736         $limit .= " and " if ( $query || $limit );
1737         $limit .= "($group_OR_limits{$k})";
1738     }
1739     if ($availability_limit) {
1740         $limit .= " and " if ( $query || $limit );
1741         $limit .= "($availability_limit)";
1742     }
1743
1744     # Normalize the query and limit strings
1745     # This is flawed , means we can't search anything with : in it
1746     # if user wants to do ccl or cql, start the query with that
1747 #    $query =~ s/:/=/g;
1748     #NOTE: We use several several different regexps here as you can't have variable length lookback assertions
1749     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1750     $query =~ s/(?<=(wrdl)):/=/g;
1751     $query =~ s/(?<=(trn|phr)):/=/g;
1752     $query =~ s/(?<=(st-numeric)):/=/g;
1753     $query =~ s/(?<=(st-year)):/=/g;
1754     $query =~ s/(?<=(st-date-normalized)):/=/g;
1755     $limit =~ s/:/=/g;
1756     for ( $query, $query_desc, $limit, $limit_desc ) {
1757         s/  +/ /g;    # remove extra spaces
1758         s/^ //g;     # remove any beginning spaces
1759         s/ $//g;     # remove any ending spaces
1760         s/==/=/g;    # remove double == from query
1761     }
1762     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1763
1764     for ($query_cgi,$simple_query) {
1765         s/"//g;
1766     }
1767     # append the limit to the query
1768     $query .= " " . $limit;
1769
1770     # Warnings if DEBUG
1771     if ($DEBUG) {
1772         warn "QUERY:" . $query;
1773         warn "QUERY CGI:" . $query_cgi;
1774         warn "QUERY DESC:" . $query_desc;
1775         warn "LIMIT:" . $limit;
1776         warn "LIMIT CGI:" . $limit_cgi;
1777         warn "LIMIT DESC:" . $limit_desc;
1778         warn "---------\nLeave buildQuery\n---------";
1779     }
1780     return (
1781         undef,              $query, $simple_query, $query_cgi,
1782         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1783         $stopwords_removed, $query_type
1784     );
1785 }
1786
1787 =head2 _build_initial_query
1788
1789   ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params);
1790
1791   Build a section of the initial query containing indexes, operators, and operands.
1792
1793 =cut
1794
1795 sub _build_initial_query {
1796     my ($params) = @_;
1797
1798     my $operator = "";
1799     if ($params->{previous_operand}){
1800         #If there is a previous operand, add a supplied operator or the default 'and'
1801         $operator = ($params->{operator}) ? " ".($params->{operator})." " : ' and ';
1802     }
1803
1804     #NOTE: indexes_set is typically set when doing truncation or field weighting
1805     my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand};
1806
1807     #e.g. "kw,wrdl:test"
1808     #e.g. " and kw,wrdl:test"
1809     $params->{query} .= $operator . $operand;
1810
1811     $params->{query_cgi} .= "&op=".uri_escape($operator) if $operator;
1812     $params->{query_cgi} .= "&idx=".uri_escape($params->{index}) if $params->{index};
1813     $params->{query_cgi} .= "&q=".uri_escape($params->{original_operand}) if $params->{original_operand};
1814
1815     #e.g. " and kw,wrdl: test"
1816     $params->{query_desc} .= $operator . $params->{index_plus} . " " . $params->{original_operand};
1817
1818     $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one
1819
1820     return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand});
1821 }
1822
1823 =head2 searchResults
1824
1825   my @search_results = searchResults($search_context, $searchdesc, $hits,
1826                                      $results_per_page, $offset, $scan,
1827                                      @marcresults);
1828
1829 Format results in a form suitable for passing to the template
1830
1831 =cut
1832
1833 # IMO this subroutine is pretty messy still -- it's responsible for
1834 # building the HTML output for the template
1835 sub searchResults {
1836     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1837     my $dbh = C4::Context->dbh;
1838     my @newresults;
1839
1840     require C4::Items;
1841
1842     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1843     my ($is_opac, $hidelostitems);
1844     if ($search_context eq 'opac') {
1845         $hidelostitems = C4::Context->preference('hidelostitems');
1846         $is_opac       = 1;
1847     }
1848
1849     #Build branchnames hash
1850     #find branchname
1851     #get branch information.....
1852     my %branches;
1853     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1854     $bsth->execute();
1855     while ( my $bdata = $bsth->fetchrow_hashref ) {
1856         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1857     }
1858 # FIXME - We build an authorised values hash here, using the default framework
1859 # though it is possible to have different authvals for different fws.
1860
1861     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1862
1863     # get notforloan authorised value list (see $shelflocations  FIXME)
1864     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1865
1866     #Build itemtype hash
1867     #find itemtype & itemtype image
1868     my %itemtypes;
1869     $bsth =
1870       $dbh->prepare(
1871         "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1872       );
1873     $bsth->execute();
1874     while ( my $bdata = $bsth->fetchrow_hashref ) {
1875                 foreach (qw(description imageurl summary notforloan)) {
1876                 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1877                 }
1878     }
1879
1880     #search item field code
1881     my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber", "" );
1882
1883     ## find column names of items related to MARC
1884     my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1885     $sth2->execute;
1886     my %subfieldstosearch;
1887     while ( ( my $column ) = $sth2->fetchrow ) {
1888         my ( $tagfield, $tagsubfield ) =
1889           &GetMarcFromKohaField( "items." . $column, "" );
1890         if ( defined $tagsubfield ) {
1891             $subfieldstosearch{$column} = $tagsubfield;
1892         }
1893     }
1894
1895     # handle which records to actually retrieve
1896     my $times;
1897     if ( $hits && $offset + $results_per_page <= $hits ) {
1898         $times = $offset + $results_per_page;
1899     }
1900     else {
1901         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1902     }
1903
1904     my $marcflavour = C4::Context->preference("marcflavour");
1905     # We get the biblionumber position in MARC
1906     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1907
1908     # loop through all of the records we've retrieved
1909     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1910
1911         my $marcrecord;
1912         if ($scan) {
1913             # For Scan searches we built USMARC data
1914             $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]);
1915         } else {
1916             # Normal search, render from Zebra's output
1917             $marcrecord = new_record_from_zebra(
1918                 'biblioserver',
1919                 $marcresults->[$i]
1920             );
1921
1922             if ( ! defined $marcrecord ) {
1923                 warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i];
1924                 next;
1925             }
1926         }
1927
1928         my $fw = $scan
1929              ? undef
1930              : $bibliotag < 10
1931                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1932                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1933         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1934         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1935         $oldbiblio->{result_number} = $i + 1;
1936
1937         # add imageurl to itemtype if there is one
1938         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1939
1940         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1941                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1942                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1943                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1944                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1945                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1946
1947                 # edition information, if any
1948         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1949                 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1950  # Build summary if there is one (the summary is defined in the itemtypes table)
1951  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1952         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1953             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1954             my @fields  = $marcrecord->fields();
1955
1956             my $newsummary;
1957             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1958                 my $tags = {};
1959                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1960                     $tag =~ /(.{3})(.)/;
1961                     if($marcrecord->field($1)){
1962                         my @abc = $marcrecord->field($1)->subfield($2);
1963                         $tags->{$tag} = $#abc + 1 ;
1964                     }
1965                 }
1966
1967                 # We catch how many times to repeat this line
1968                 my $max = 0;
1969                 foreach my $tag (keys(%$tags)){
1970                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1971                  }
1972
1973                 # we replace, and repeat each line
1974                 for (my $i = 0 ; $i < $max ; $i++){
1975                     my $newline = $line;
1976
1977                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1978                         $tag =~ /(.{3})(.)/;
1979
1980                         if($marcrecord->field($1)){
1981                             my @repl = $marcrecord->field($1)->subfield($2);
1982                             my $subfieldvalue = $repl[$i];
1983
1984                             if (! utf8::is_utf8($subfieldvalue)) {
1985                                 utf8::decode($subfieldvalue);
1986                             }
1987
1988                              $newline =~ s/\[$tag\]/$subfieldvalue/g;
1989                         }
1990                     }
1991                     $newsummary .= "$newline\n";
1992                 }
1993             }
1994
1995             $newsummary =~ s/\[(.*?)]//g;
1996             $newsummary =~ s/\n/<br\/>/g;
1997             $oldbiblio->{summary} = $newsummary;
1998         }
1999
2000         # Pull out the items fields
2001         my @fields = $marcrecord->field($itemtag);
2002         my $marcflavor = C4::Context->preference("marcflavour");
2003         # adding linked items that belong to host records
2004         my $analyticsfield = '773';
2005         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
2006             $analyticsfield = '773';
2007         } elsif ($marcflavor eq 'UNIMARC') {
2008             $analyticsfield = '461';
2009         }
2010         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
2011             my $hostbiblionumber = $hostfield->subfield("0");
2012             my $linkeditemnumber = $hostfield->subfield("9");
2013             if(!$hostbiblionumber eq undef){
2014                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
2015                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
2016                 if(!$hostbiblio eq undef){
2017                     my @hostitems = $hostbiblio->field($itemfield);
2018                     foreach my $hostitem (@hostitems){
2019                         if ($hostitem->subfield("9") eq $linkeditemnumber){
2020                             my $linkeditem =$hostitem;
2021                             # append linked items if they exist
2022                             if (!$linkeditem eq undef){
2023                                 push (@fields, $linkeditem);}
2024                         }
2025                     }
2026                 }
2027             }
2028         }
2029
2030         # Setting item statuses for display
2031         my @available_items_loop;
2032         my @onloan_items_loop;
2033         my @other_items_loop;
2034
2035         my $available_items;
2036         my $onloan_items;
2037         my $other_items;
2038
2039         my $ordered_count         = 0;
2040         my $available_count       = 0;
2041         my $onloan_count          = 0;
2042         my $longoverdue_count     = 0;
2043         my $other_count           = 0;
2044         my $withdrawn_count        = 0;
2045         my $itemlost_count        = 0;
2046         my $hideatopac_count      = 0;
2047         my $itembinding_count     = 0;
2048         my $itemdamaged_count     = 0;
2049         my $item_in_transit_count = 0;
2050         my $can_place_holds       = 0;
2051         my $item_onhold_count     = 0;
2052         my $notforloan_count      = 0;
2053         my $items_count           = scalar(@fields);
2054         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
2055         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
2056         my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref
2057
2058         # loop through every item
2059         foreach my $field (@fields) {
2060             my $item;
2061
2062             # populate the items hash
2063             foreach my $code ( keys %subfieldstosearch ) {
2064                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
2065             }
2066             $item->{description} = $itemtypes{ $item->{itype} }{description};
2067
2068                 # OPAC hidden items
2069             if ($is_opac) {
2070                 # hidden because lost
2071                 if ($hidelostitems && $item->{itemlost}) {
2072                     $hideatopac_count++;
2073                     next;
2074                 }
2075                 # hidden based on OpacHiddenItems syspref
2076                 my @hi = C4::Items::GetHiddenItemnumbers($item);
2077                 if (scalar @hi) {
2078                     push @hiddenitems, @hi;
2079                     $hideatopac_count++;
2080                     next;
2081                 }
2082             }
2083
2084             my $hbranch     = C4::Context->preference('StaffSearchResultsDisplayBranch');
2085             my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch';
2086
2087             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
2088             if ($item->{$hbranch}) {
2089                 $item->{'branchname'} = $branches{$item->{$hbranch}};
2090             }
2091             elsif ($item->{$otherbranch}) {     # Last resort
2092                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
2093             }
2094
2095                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
2096 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
2097             my $userenv = C4::Context->userenv;
2098             if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
2099                 $onloan_count++;
2100                                 my $key = $prefix . $item->{onloan} . $item->{barcode};
2101                                 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
2102                                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
2103                                 $onloan_items->{$key}->{branchname} = $item->{branchname};
2104                                 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2105                                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
2106                                 $onloan_items->{$key}->{description} = $item->{description};
2107                                 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2108                 # if something's checked out and lost, mark it as 'long overdue'
2109                 if ( $item->{itemlost} ) {
2110                     $onloan_items->{$prefix}->{longoverdue}++;
2111                     $longoverdue_count++;
2112                 } else {        # can place holds as long as item isn't lost
2113                     $can_place_holds = 1;
2114                 }
2115             }
2116
2117          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
2118             else {
2119
2120                 # item is on order
2121                 if ( $item->{notforloan} < 0 ) {
2122                     $ordered_count++;
2123                 } elsif ( $item->{notforloan} > 0 ) {
2124                     $notforloan_count++;
2125                 }
2126
2127                 # is item in transit?
2128                 my $transfertwhen = '';
2129                 my ($transfertfrom, $transfertto);
2130
2131                 # is item on the reserve shelf?
2132                 my $reservestatus = '';
2133
2134                 unless ($item->{withdrawn}
2135                         || $item->{itemlost}
2136                         || $item->{damaged}
2137                         || $item->{notforloan}
2138                         || $items_count > 20) {
2139
2140                     # A couple heuristics to limit how many times
2141                     # we query the database for item transfer information, sacrificing
2142                     # accuracy in some cases for speed;
2143                     #
2144                     # 1. don't query if item has one of the other statuses
2145                     # 2. don't check transit status if the bib has
2146                     #    more than 20 items
2147                     #
2148                     # FIXME: to avoid having the query the database like this, and to make
2149                     #        the in transit status count as unavailable for search limiting,
2150                     #        should map transit status to record indexed in Zebra.
2151                     #
2152                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
2153                     $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber}, $oldbiblio->{biblionumber} );
2154                 }
2155
2156                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
2157                 if (   $item->{withdrawn}
2158                     || $item->{itemlost}
2159                     || $item->{damaged}
2160                     || $item->{notforloan}
2161                     || $reservestatus eq 'Waiting'
2162                     || ($transfertwhen ne ''))
2163                 {
2164                     $withdrawn_count++        if $item->{withdrawn};
2165                     $itemlost_count++        if $item->{itemlost};
2166                     $itemdamaged_count++     if $item->{damaged};
2167                     $item_in_transit_count++ if $transfertwhen ne '';
2168                     $item_onhold_count++     if $reservestatus eq 'Waiting';
2169                     $item->{status} = $item->{withdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
2170
2171                     # can place a hold on a item if
2172                     # not lost nor withdrawn
2173                     # not damaged unless AllowHoldsOnDamagedItems is true
2174                     # item is either for loan or on order (notforloan < 0)
2175                     $can_place_holds = 1
2176                       if (
2177                            !$item->{itemlost}
2178                         && !$item->{withdrawn}
2179                         && ( !$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems') )
2180                         && ( !$item->{notforloan} || $item->{notforloan} < 0 )
2181                       );
2182
2183                     $other_count++;
2184
2185                     my $key = $prefix . $item->{status};
2186                     foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) {
2187                         $other_items->{$key}->{$_} = $item->{$_};
2188                     }
2189                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
2190                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
2191                     $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan};
2192                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
2193                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
2194                                         $other_items->{$key}->{description} = $item->{description};
2195                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2196                 }
2197                 # item is available
2198                 else {
2199                     $can_place_holds = 1;
2200                     $available_count++;
2201                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
2202                                         foreach (qw(branchname itemcallnumber description)) {
2203                         $available_items->{$prefix}->{$_} = $item->{$_};
2204                                         }
2205                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
2206                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
2207                 }
2208             }
2209         }    # notforloan, item level and biblioitem level
2210
2211         # if all items are hidden, do not show the record
2212         if ($items_count > 0 && $hideatopac_count == $items_count) {
2213             next;
2214         }
2215
2216         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
2217         for my $key ( sort keys %$onloan_items ) {
2218             (++$onloanitemscount > $maxitems) and last;
2219             push @onloan_items_loop, $onloan_items->{$key};
2220         }
2221         for my $key ( sort keys %$other_items ) {
2222             (++$otheritemscount > $maxitems) and last;
2223             push @other_items_loop, $other_items->{$key};
2224         }
2225         for my $key ( sort keys %$available_items ) {
2226             (++$availableitemscount > $maxitems) and last;
2227             push @available_items_loop, $available_items->{$key}
2228         }
2229
2230         # XSLT processing of some stuff
2231         SetUTF8Flag($marcrecord);
2232         warn $marcrecord->as_formatted if $DEBUG;
2233         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
2234         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
2235             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, $interface."XSLTResultsDisplay", 1, \@hiddenitems);
2236         # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
2237         }
2238
2239         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
2240         if (!C4::Context->preference("item-level_itypes")) {
2241             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
2242                 $can_place_holds = 0;
2243             }
2244         }
2245         $oldbiblio->{norequests} = 1 unless $can_place_holds;
2246         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
2247         $oldbiblio->{items_count}          = $items_count;
2248         $oldbiblio->{available_items_loop} = \@available_items_loop;
2249         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
2250         $oldbiblio->{other_items_loop}     = \@other_items_loop;
2251         $oldbiblio->{availablecount}       = $available_count;
2252         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
2253         $oldbiblio->{onloancount}          = $onloan_count;
2254         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
2255         $oldbiblio->{othercount}           = $other_count;
2256         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
2257         $oldbiblio->{withdrawncount}        = $withdrawn_count;
2258         $oldbiblio->{itemlostcount}        = $itemlost_count;
2259         $oldbiblio->{damagedcount}         = $itemdamaged_count;
2260         $oldbiblio->{intransitcount}       = $item_in_transit_count;
2261         $oldbiblio->{onholdcount}          = $item_onhold_count;
2262         $oldbiblio->{orderedcount}         = $ordered_count;
2263         $oldbiblio->{notforloancount}      = $notforloan_count;
2264
2265         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
2266             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
2267             my $subfields = substr $fieldspec, 3;
2268             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
2269             my @alternateholdingsinfo = ();
2270             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
2271             my $alternateholdingscount = 0;
2272
2273             for my $field (@holdingsfields) {
2274                 my %holding = ( holding => '' );
2275                 my $havesubfield = 0;
2276                 for my $subfield ($field->subfields()) {
2277                     if ((index $subfields, $$subfield[0]) >= 0) {
2278                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
2279                         $holding{'holding'} .= $$subfield[1];
2280                         $havesubfield++;
2281                     }
2282                 }
2283                 if ($havesubfield) {
2284                     push(@alternateholdingsinfo, \%holding);
2285                     $alternateholdingscount++;
2286                 }
2287             }
2288
2289             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
2290             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
2291         }
2292
2293         push( @newresults, $oldbiblio );
2294     }
2295
2296     return @newresults;
2297 }
2298
2299 =head2 SearchAcquisitions
2300     Search for acquisitions
2301 =cut
2302
2303 sub SearchAcquisitions{
2304     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
2305
2306     my $dbh=C4::Context->dbh;
2307     # Variable initialization
2308     my $str=qq|
2309     SELECT marcxml
2310     FROM biblio
2311     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
2312     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
2313     WHERE dateaccessioned BETWEEN ? AND ?
2314     |;
2315
2316     my (@params,@loopcriteria);
2317
2318     push @params, $datebegin->output("iso");
2319     push @params, $dateend->output("iso");
2320
2321     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
2322         if(C4::Context->preference("item-level_itypes")){
2323             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2324         }else{
2325             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
2326         }
2327         push @params, @$itemtypes;
2328     }
2329
2330     if ($criteria =~/itemtype/){
2331         if(C4::Context->preference("item-level_itypes")){
2332             $str .= "AND items.itype=? ";
2333         }else{
2334             $str .= "AND biblioitems.itemtype=? ";
2335         }
2336
2337         if(scalar(@$itemtypes) == 0){
2338             my $itypes = GetItemTypes();
2339             for my $key (keys %$itypes){
2340                 push @$itemtypes, $key;
2341             }
2342         }
2343
2344         @loopcriteria= @$itemtypes;
2345     }elsif ($criteria=~/itemcallnumber/){
2346         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
2347                  OR items.itemcallnumber is NULL
2348                  OR items.itemcallnumber = '')";
2349
2350         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
2351     }else {
2352         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
2353         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
2354     }
2355
2356     if ($orderby =~ /date_desc/){
2357         $str.=" ORDER BY dateaccessioned DESC";
2358     } else {
2359         $str.=" ORDER BY title";
2360     }
2361
2362     my $qdataacquisitions=$dbh->prepare($str);
2363
2364     my @loopacquisitions;
2365     foreach my $value(@loopcriteria){
2366         push @params,$value;
2367         my %cell;
2368         $cell{"title"}=$value;
2369         $cell{"titlecode"}=$value;
2370
2371         eval{$qdataacquisitions->execute(@params);};
2372
2373         if ($@){ warn "recentacquisitions Error :$@";}
2374         else {
2375             my @loopdata;
2376             while (my $data=$qdataacquisitions->fetchrow_hashref){
2377                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
2378             }
2379             $cell{"loopdata"}=\@loopdata;
2380         }
2381         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
2382         pop @params;
2383     }
2384     $qdataacquisitions->finish;
2385     return \@loopacquisitions;
2386 }
2387
2388 =head2 enabled_staff_search_views
2389
2390 %hash = enabled_staff_search_views()
2391
2392 This function returns a hash that contains three flags obtained from the system
2393 preferences, used to determine whether a particular staff search results view
2394 is enabled.
2395
2396 =over 2
2397
2398 =item C<Output arg:>
2399
2400     * $hash{can_view_MARC} is true only if the MARC view is enabled
2401     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2402     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2403
2404 =item C<usage in the script:>
2405
2406 =back
2407
2408 $template->param ( C4::Search::enabled_staff_search_views );
2409
2410 =cut
2411
2412 sub enabled_staff_search_views
2413 {
2414         return (
2415                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2416                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2417                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2418         );
2419 }
2420
2421 sub PurgeSearchHistory{
2422     my ($pSearchhistory)=@_;
2423     my $dbh = C4::Context->dbh;
2424     my $sth = $dbh->prepare("DELETE FROM search_history WHERE time < DATE_SUB( NOW(), INTERVAL ? DAY )");
2425     $sth->execute($pSearchhistory) or die $dbh->errstr;
2426 }
2427
2428 =head2 z3950_search_args
2429
2430 $arrayref = z3950_search_args($matchpoints)
2431
2432 This function returns an array reference that contains the search parameters to be
2433 passed to the Z39.50 search script (z3950_search.pl). The array elements
2434 are hash refs whose keys are name and value, and whose values are the
2435 name of a search parameter, the value of that search parameter and the URL encoded
2436 value of that parameter.
2437
2438 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2439
2440 The search parameter values are obtained from the bibliographic record whose
2441 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2442
2443 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2444 a general purpose search argument. In this case, the returned array contains only
2445 entry: the key is 'title' and the value is derived from $matchpoints.
2446
2447 If a search parameter value is undefined or empty, it is not included in the returned
2448 array.
2449
2450 The returned array reference may be passed directly to the template parameters.
2451
2452 =over 2
2453
2454 =item C<Output arg:>
2455
2456     * $array containing hash refs as described above
2457
2458 =item C<usage in the script:>
2459
2460 =back
2461
2462 $data = Biblio::GetBiblioData($bibno);
2463 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2464
2465 *OR*
2466
2467 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2468
2469 =cut
2470
2471 sub z3950_search_args {
2472     my $bibrec = shift;
2473
2474     my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec;
2475     my $isbn = Business::ISBN->new( $isbn_string );
2476
2477     if (defined $isbn && $isbn->is_valid)
2478     {
2479         if ( ref($bibrec) ) {
2480             $bibrec->{isbn} = $isbn_string;
2481             $bibrec->{title} = undef;
2482         } else {
2483             $bibrec = { isbn => $isbn_string };
2484         }
2485     }
2486     else {
2487         $bibrec = { title => $bibrec } if !ref $bibrec;
2488     }
2489     my $array = [];
2490     for my $field (qw/ lccn isbn issn title author dewey subject /)
2491     {
2492         push @$array, { name => $field, value => $bibrec->{$field} }
2493           if defined $bibrec->{$field};
2494     }
2495     return $array;
2496 }
2497
2498 =head2 GetDistinctValues($field);
2499
2500 C<$field> is a reference to the fields array
2501
2502 =cut
2503
2504 sub GetDistinctValues {
2505     my ($fieldname,$string)=@_;
2506     # returns a reference to a hash of references to branches...
2507     if ($fieldname=~/\./){
2508                         my ($table,$column)=split /\./, $fieldname;
2509                         my $dbh = C4::Context->dbh;
2510                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2511                         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 ");
2512                         $sth->execute;
2513                         my $elements=$sth->fetchall_arrayref({});
2514                         return $elements;
2515    }
2516    else {
2517                 $string||= qq("");
2518                 my @servers=qw<biblioserver authorityserver>;
2519                 my (@zconns,@results);
2520         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2521                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2522                         $results[$i] =
2523                       $zconns[$i]->scan(
2524                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2525                       );
2526                 }
2527                 # The big moment: asynchronously retrieve results from all servers
2528                 my @elements;
2529         _ZOOM_event_loop(
2530             \@zconns,
2531             \@results,
2532             sub {
2533                 my ( $i, $size ) = @_;
2534                 for ( my $j = 0 ; $j < $size ; $j++ ) {
2535                     my %hashscan;
2536                     @hashscan{qw(value cnt)} =
2537                       $results[ $i - 1 ]->display_term($j);
2538                     push @elements, \%hashscan;
2539                 }
2540             }
2541         );
2542                 return \@elements;
2543    }
2544 }
2545
2546 =head2 _ZOOM_event_loop
2547
2548     _ZOOM_event_loop(\@zconns, \@results, sub {
2549         my ( $i, $size ) = @_;
2550         ....
2551     } );
2552
2553 Processes a ZOOM event loop and passes control to a closure for
2554 processing the results, and destroying the resultsets.
2555
2556 =cut
2557
2558 sub _ZOOM_event_loop {
2559     my ($zconns, $results, $callback) = @_;
2560     while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) {
2561         my $ev = $zconns->[ $i - 1 ]->last_event();
2562         if ( $ev == ZOOM::Event::ZEND ) {
2563             next unless $results->[ $i - 1 ];
2564             my $size = $results->[ $i - 1 ]->size();
2565             if ( $size > 0 ) {
2566                 $callback->($i, $size);
2567             }
2568         }
2569     }
2570
2571     foreach my $result (@$results) {
2572         $result->destroy();
2573     }
2574 }
2575
2576 =head2 new_record_from_zebra
2577
2578 Given raw data from a Zebra result set, return a MARC::Record object
2579
2580 This helper function is needed to take into account all the involved
2581 system preferences and configuration variables to properly create the
2582 MARC::Record object.
2583
2584 If we are using GRS-1, then the raw data we get from Zebra should be USMARC
2585 data. If we are using DOM, then it has to be MARCXML.
2586
2587 =cut
2588
2589 sub new_record_from_zebra {
2590
2591     my $server   = shift;
2592     my $raw_data = shift;
2593     # Set the default indexing modes
2594     my $index_mode = ( $server eq 'biblioserver' )
2595                         ? C4::Context->config('zebra_bib_index_mode') // 'grs1'
2596                         : C4::Context->config('zebra_auth_index_mode') // 'dom';
2597
2598     my $marc_record =  eval {
2599         if ( $index_mode eq 'dom' ) {
2600             MARC::Record->new_from_xml( $raw_data, 'UTF-8' );
2601         } else {
2602             MARC::Record->new_from_usmarc( $raw_data );
2603         }
2604     };
2605
2606     if ($@) {
2607         return;
2608     } else {
2609         return $marc_record;
2610     }
2611
2612 }
2613
2614 END { }    # module clean-up code here (global destructor)
2615
2616 1;
2617 __END__
2618
2619 =head1 AUTHOR
2620
2621 Koha Development Team <http://koha-community.org/>
2622
2623 =cut