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;    # CheckReserves
  32 use C4::Debug;
  33 use C4::Items;
  34 use C4::Charset;
  35 use YAML;
  36 use URI::Escape;
  37 use Business::ISBN;
  38
  39 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  40
  41 # set the version for version checking
  42 BEGIN {
  43     $VERSION = 3.01;
  44     $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
  45 }
  46
  47 =head1 NAME
  48
  49 C4::Search - Functions for searching the Koha catalog.
  50
  51 =head1 SYNOPSIS
  52
  53 See opac/opac-search.pl or catalogue/search.pl for example of usage
  54
  55 =head1 DESCRIPTION
  56
  57 This module provides searching functions for Koha's bibliographic databases
  58
  59 =head1 FUNCTIONS
  60
  61 =cut
  62
  63 @ISA    = qw(Exporter);
  64 @EXPORT = qw(
  65   &FindDuplicate
  66   &SimpleSearch
  67   &searchResults
  68   &getRecords
  69   &buildQuery
  70   &NZgetRecords
  71   &AddSearchHistory
  72   &GetDistinctValues
  73   &enabled_staff_search_views
  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         $result->{title} =~ s /\\//g;
 105         $result->{title} =~ s /\"//g;
 106         $result->{title} =~ s /\(//g;
 107         $result->{title} =~ s /\)//g;
 108
 109         # FIXME: instead of removing operators, could just do
 110         # quotes around the value
 111         $result->{title} =~ s/(and|or|not)//g;
 112         $query = "ti,ext=$result->{title}";
 113         $query .= " and itemtype=$result->{itemtype}"
 114           if ( $result->{itemtype} );
 115         if   ( $result->{author} ) {
 116             $result->{author} =~ s /\\//g;
 117             $result->{author} =~ s /\"//g;
 118             $result->{author} =~ s /\(//g;
 119             $result->{author} =~ s /\)//g;
 120
 121             # remove valid operators
 122             $result->{author} =~ s/(and|or|not)//g;
 123             $query .= " and au,ext=$result->{author}";
 124         }
 125     }
 126
 127     my ( $error, $searchresults, undef ) = SimpleSearch($query); # FIXME :: hardcoded !
 128     my @results;
 129     if (!defined $error) {
 130         foreach my $possible_duplicate_record (@{$searchresults}) {
 131             my $marcrecord =
 132             MARC::Record->new_from_usmarc($possible_duplicate_record);
 133             my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
 134
 135             # FIXME :: why 2 $biblionumber ?
 136             if ($result) {
 137                 push @results, $result->{'biblionumber'};
 138                 push @results, $result->{'title'};
 139             }
 140         }
 141     }
 142     return @results;
 143 }
 144
 145 =head2 SimpleSearch
 146
 147 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
 148
 149 This function provides a simple search API on the bibliographic catalog
 150
 151 =over 2
 152
 153 =item C<input arg:>
 154
 155     * $query can be a simple keyword or a complete CCL query
 156     * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
 157     * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
 158     * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
 159
 160
 161 =item C<Return:>
 162
 163     Returns an array consisting of three elements
 164     * $error is undefined unless an error is detected
 165     * $results is a reference to an array of records.
 166     * $total_hits is the number of hits that would have been returned with no limit
 167
 168     If an error is returned the two other return elements are undefined. If error itself is undefined
 169     the other two elements are always defined
 170
 171 =item C<usage in the script:>
 172
 173 =back
 174
 175 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
 176
 177 if (defined $error) {
 178     $template->param(query_error => $error);
 179     warn "error: ".$error;
 180     output_html_with_http_headers $input, $cookie, $template->output;
 181     exit;
 182 }
 183
 184 my $hits = @{$marcresults};
 185 my @results;
 186
 187 for my $r ( @{$marcresults} ) {
 188     my $marcrecord = MARC::File::USMARC::decode($r);
 189     my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,q{});
 190
 191     #build the iarray of hashs for the template.
 192     push @results, {
 193         title           => $biblio->{'title'},
 194         subtitle        => $biblio->{'subtitle'},
 195         biblionumber    => $biblio->{'biblionumber'},
 196         author          => $biblio->{'author'},
 197         publishercode   => $biblio->{'publishercode'},
 198         publicationyear => $biblio->{'publicationyear'},
 199         };
 200
 201 }
 202
 203 $template->param(result=>\@results);
 204
 205 =cut
 206
 207 sub SimpleSearch {
 208     my ( $query, $offset, $max_results, $servers )  = @_;
 209
 210     if ( C4::Context->preference('NoZebra') ) {
 211         my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
 212         my $search_result =
 213           (      $result->{hits}
 214               && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
 215         return ( undef, $search_result, scalar($result->{hits}) );
 216     }
 217     else {
 218         return ( 'No query entered', undef, undef ) unless $query;
 219         # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
 220         my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' );
 221         my @zoom_queries;
 222         my @tmpresults;
 223         my @zconns;
 224         my $results = [];
 225         my $total_hits = 0;
 226
 227         # Initialize & Search Zebra
 228         for ( my $i = 0 ; $i < @servers ; $i++ ) {
 229             eval {
 230                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 231                 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
 232                 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
 233
 234                 # error handling
 235                 my $error =
 236                     $zconns[$i]->errmsg() . " ("
 237                   . $zconns[$i]->errcode() . ") "
 238                   . $zconns[$i]->addinfo() . " "
 239                   . $zconns[$i]->diagset();
 240
 241                 return ( $error, undef, undef ) if $zconns[$i]->errcode();
 242             };
 243             if ($@) {
 244
 245                 # caught a ZOOM::Exception
 246                 my $error =
 247                     $@->message() . " ("
 248                   . $@->code() . ") "
 249                   . $@->addinfo() . " "
 250                   . $@->diagset();
 251                 warn $error;
 252                 return ( $error, undef, undef );
 253             }
 254         }
 255         while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
 256             my $event = $zconns[ $i - 1 ]->last_event();
 257             if ( $event == ZOOM::Event::ZEND ) {
 258
 259                 my $first_record = defined( $offset ) ? $offset+1 : 1;
 260                 my $hits = $tmpresults[ $i - 1 ]->size();
 261                 $total_hits += $hits;
 262                 my $last_record = $hits;
 263                 if ( defined $max_results && $offset + $max_results < $hits ) {
 264                     $last_record  = $offset + $max_results;
 265                 }
 266
 267                 for my $j ( $first_record..$last_record ) {
 268                     my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
 269                     push @{$results}, $record;
 270                 }
 271             }
 272         }
 273
 274         foreach my $result (@tmpresults) {
 275             $result->destroy();
 276         }
 277         foreach my $zoom_query (@zoom_queries) {
 278             $zoom_query->destroy();
 279         }
 280
 281         return ( undef, $results, $total_hits );
 282     }
 283 }
 284
 285 =head2 getRecords
 286
 287 ( undef, $results_hashref, \@facets_loop ) = getRecords (
 288
 289         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 290         $results_per_page, $offset,       $expanded_facet, $branches,
 291         $query_type,       $scan
 292     );
 293
 294 The all singing, all dancing, multi-server, asynchronous, scanning,
 295 searching, record nabbing, facet-building
 296
 297 See verbse embedded documentation.
 298
 299 =cut
 300
 301 sub getRecords {
 302     my (
 303         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 304         $results_per_page, $offset,       $expanded_facet, $branches,
 305         $query_type,       $scan
 306     ) = @_;
 307
 308     my @servers = @$servers_ref;
 309     my @sort_by = @$sort_by_ref;
 310
 311     # Initialize variables for the ZOOM connection and results object
 312     my $zconn;
 313     my @zconns;
 314     my @results;
 315     my $results_hashref = ();
 316
 317     # Initialize variables for the faceted results objects
 318     my $facets_counter = ();
 319     my $facets_info    = ();
 320     my $facets         = getFacets();
 321     my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets')||20;
 322
 323     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 324
 325     ### LOOP THROUGH THE SERVERS
 326     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 327         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 328
 329 # perform the search, create the results objects
 330 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 331         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 332
 333         #$query_to_use = $simple_query if $scan;
 334         warn $simple_query if ( $scan and $DEBUG );
 335
 336         # Check if we've got a query_type defined, if so, use it
 337         eval {
 338             if ($query_type) {
 339                 if ($query_type =~ /^ccl/) {
 340                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 341                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 342                 } elsif ($query_type =~ /^cql/) {
 343                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 344                 } elsif ($query_type =~ /^pqf/) {
 345                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 346                 } else {
 347                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 348                 }
 349             } elsif ($scan) {
 350                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 351             } else {
 352                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 353             }
 354         };
 355         if ($@) {
 356             warn "WARNING: query problem with $query_to_use " . $@;
 357         }
 358
 359         # Concatenate the sort_by limits and pass them to the results object
 360         # Note: sort will override rank
 361         my $sort_by;
 362         foreach my $sort (@sort_by) {
 363             if ( $sort eq "author_az" || $sort eq "author_asc" ) {
 364                 $sort_by .= "1=1003 <i ";
 365             }
 366             elsif ( $sort eq "author_za" || $sort eq "author_dsc" ) {
 367                 $sort_by .= "1=1003 >i ";
 368             }
 369             elsif ( $sort eq "popularity_asc" ) {
 370                 $sort_by .= "1=9003 <i ";
 371             }
 372             elsif ( $sort eq "popularity_dsc" ) {
 373                 $sort_by .= "1=9003 >i ";
 374             }
 375             elsif ( $sort eq "call_number_asc" ) {
 376                 $sort_by .= "1=8007  <i ";
 377             }
 378             elsif ( $sort eq "call_number_dsc" ) {
 379                 $sort_by .= "1=8007 >i ";
 380             }
 381             elsif ( $sort eq "pubdate_asc" ) {
 382                 $sort_by .= "1=31 <i ";
 383             }
 384             elsif ( $sort eq "pubdate_dsc" ) {
 385                 $sort_by .= "1=31 >i ";
 386             }
 387             elsif ( $sort eq "acqdate_asc" ) {
 388                 $sort_by .= "1=32 <i ";
 389             }
 390             elsif ( $sort eq "acqdate_dsc" ) {
 391                 $sort_by .= "1=32 >i ";
 392             }
 393             elsif ( $sort eq "title_az" || $sort eq "title_asc" ) {
 394                 $sort_by .= "1=4 <i ";
 395             }
 396             elsif ( $sort eq "title_za" || $sort eq "title_dsc" ) {
 397                 $sort_by .= "1=4 >i ";
 398             }
 399             else {
 400                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 401             }
 402         }
 403         if ($sort_by && !$scan) {
 404             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 405                 warn "WARNING sort $sort_by failed";
 406             }
 407         }
 408     }    # finished looping through servers
 409
 410     # The big moment: asynchronously retrieve results from all servers
 411     while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
 412         my $ev = $zconns[ $i - 1 ]->last_event();
 413         if ( $ev == ZOOM::Event::ZEND ) {
 414             next unless $results[ $i - 1 ];
 415             my $size = $results[ $i - 1 ]->size();
 416             if ( $size > 0 ) {
 417                 my $results_hash;
 418
 419                 # loop through the results
 420                 $results_hash->{'hits'} = $size;
 421                 my $times;
 422                 if ( $offset + $results_per_page <= $size ) {
 423                     $times = $offset + $results_per_page;
 424                 }
 425                 else {
 426                     $times = $size;
 427                 }
 428                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 429                     my $records_hash;
 430                     my $record;
 431
 432                     ## Check if it's an index scan
 433                     if ($scan) {
 434                         my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
 435
 436                  # here we create a minimal MARC record and hand it off to the
 437                  # template just like a normal result ... perhaps not ideal, but
 438                  # it works for now
 439                         my $tmprecord = MARC::Record->new();
 440                         $tmprecord->encoding('UTF-8');
 441                         my $tmptitle;
 442                         my $tmpauthor;
 443
 444                 # the minimal record in author/title (depending on MARC flavour)
 445                         if (C4::Context->preference("marcflavour") eq "UNIMARC") {
 446                             $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
 447                             $tmprecord->append_fields($tmptitle);
 448                         } else {
 449                             $tmptitle  = MARC::Field->new('245',' ',' ', a => $term,);
 450                             $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
 451                             $tmprecord->append_fields($tmptitle);
 452                             $tmprecord->append_fields($tmpauthor);
 453                         }
 454                         $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
 455                     }
 456
 457                     # not an index scan
 458                     else {
 459                         $record = $results[ $i - 1 ]->record($j)->raw();
 460
 461                         # warn "RECORD $j:".$record;
 462                         $results_hash->{'RECORDS'}[$j] = $record;
 463                     }
 464
 465                 }
 466                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 467
 468                 # Fill the facets while we're looping, but only for the biblioserver and not for a scan
 469                 if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) {
 470
 471                     my $jmax = $size>$facets_maxrecs? $facets_maxrecs: $size;
 472                     for my $facet ( @$facets ) {
 473                                 for ( my $j = 0 ; $j < $jmax ; $j++ ) {
 474                                     my $render_record = $results[ $i - 1 ]->record($j)->render();
 475                             my @used_datas = ();
 476                             foreach my $tag ( @{$facet->{tags}} ) {
 477                                 # avoid first line
 478                                 my $tag_num = substr($tag, 0, 3);
 479                                 my $letters = substr($tag, 3);
 480                                 my $field_pattern = '\n' . $tag_num . ' ([^\n]+)';
 481                                 my @field_tokens = ( $render_record =~ /$field_pattern/g ) ;
 482                                 foreach my $field_token (@field_tokens) {
 483                                     my @subf = ( $field_token =~ /\$([a-zA-Z0-9]) ([^\$]+)/g );
 484                                     my @values;
 485                                     for (my $i = 0; $i < @subf; $i += 2) {
 486                                         if ( $letters =~ $subf[$i] ) {
 487                                              my $value = $subf[$i+1];
 488                                              $value =~ s/^ *//;
 489                                              $value =~ s/ *$//;
 490                                              push @values, $value;
 491                                         }
 492                                     }
 493                                     my $data = join($facet->{sep}, @values);
 494                                     unless ( $data ~~ @used_datas ) {
 495                                         $facets_counter->{ $facet->{idx} }->{$data}++;
 496                                         push @used_datas, $data;
 497                                     }
 498                                 } # fields
 499                             } # field codes
 500                         } # records
 501                         $facets_info->{ $facet->{idx} }->{label_value} = $facet->{label};
 502                         $facets_info->{ $facet->{idx} }->{expanded} = $facet->{expanded};
 503                     } # facets
 504                 }
 505             }
 506
 507             # warn "connection ", $i-1, ": $size hits";
 508             # warn $results[$i-1]->record(0)->render() if $size > 0;
 509
 510             # BUILD FACETS
 511             if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 512                 for my $link_value (
 513                     sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 514                         keys %$facets_counter )
 515                 {
 516                     my $expandable;
 517                     my $number_of_facets;
 518                     my @this_facets_array;
 519                     for my $one_facet (
 520                         sort {
 521                              $facets_counter->{$link_value}->{$b}
 522                          <=> $facets_counter->{$link_value}->{$a}
 523                         } keys %{ $facets_counter->{$link_value} }
 524                       )
 525                     {
 526                         $number_of_facets++;
 527                         if (   ( $number_of_facets < 6 )
 528                             || ( $expanded_facet eq $link_value )
 529                             || ( $facets_info->{$link_value}->{'expanded'} ) )
 530                         {
 531
 532                       # Sanitize the link value ), ( will cause errors with CCL,
 533                             my $facet_link_value = $one_facet;
 534                             $facet_link_value =~ s/(\(|\))/ /g;
 535
 536                             # fix the length that will display in the label,
 537                             my $facet_label_value = $one_facet;
 538                             my $facet_max_length =
 539                                 C4::Context->preference('FacetLabelTruncationLength') || 20;
 540                             $facet_label_value =
 541                               substr( $one_facet, 0, $facet_max_length ) . "..."
 542                                 if length($facet_label_value) > $facet_max_length;
 543
 544                             # if it's a branch, label by the name, not the code,
 545                             if ( $link_value =~ /branch/ ) {
 546                                                                 if (defined $branches
 547                                                                         && ref($branches) eq "HASH"
 548                                                                         && defined $branches->{$one_facet}
 549                                                                         && ref ($branches->{$one_facet}) eq "HASH")
 550                                                                 {
 551                                         $facet_label_value =
 552                                                 $branches->{$one_facet}->{'branchname'};
 553                                                                 }
 554                                                                 else {
 555                                                                         $facet_label_value = "*";
 556                                                                 }
 557                             }
 558
 559                             # but we're down with the whole label being in the link's title.
 560                             push @this_facets_array, {
 561                                 facet_count       => $facets_counter->{$link_value}->{$one_facet},
 562                                 facet_label_value => $facet_label_value,
 563                                 facet_title_value => $one_facet,
 564                                 facet_link_value  => $facet_link_value,
 565                                 type_link_value   => $link_value,
 566                             };
 567                         }
 568                     }
 569
 570                     # handle expanded option
 571                     unless ( $facets_info->{$link_value}->{'expanded'} ) {
 572                         $expandable = 1
 573                           if ( ( $number_of_facets > 6 )
 574                             && ( $expanded_facet ne $link_value ) );
 575                     }
 576                     push @facets_loop, {
 577                         type_link_value => $link_value,
 578                         type_id         => $link_value . "_id",
 579                         "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
 580                         facets     => \@this_facets_array,
 581                         expandable => $expandable,
 582                         expand     => $link_value,
 583                     } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
 584                 }
 585             }
 586         }
 587     }
 588     return ( undef, $results_hashref, \@facets_loop );
 589 }
 590
 591 sub pazGetRecords {
 592     my (
 593         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 594         $results_per_page, $offset,       $expanded_facet, $branches,
 595         $query_type,       $scan
 596     ) = @_;
 597
 598     my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
 599     $paz->init();
 600     $paz->search($simple_query);
 601     sleep 1;   # FIXME: WHY?
 602
 603     # do results
 604     my $results_hashref = {};
 605     my $stats = XMLin($paz->stat);
 606     my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
 607
 608     # for a grouped search result, the number of hits
 609     # is the number of groups returned; 'bib_hits' will have
 610     # the total number of bibs.
 611     $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
 612     $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
 613
 614     HIT: foreach my $hit (@{ $results->{'hit'} }) {
 615         my $recid = $hit->{recid}->[0];
 616
 617         my $work_title = $hit->{'md-work-title'}->[0];
 618         my $work_author;
 619         if (exists $hit->{'md-work-author'}) {
 620             $work_author = $hit->{'md-work-author'}->[0];
 621         }
 622         my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
 623
 624         my $result_group = {};
 625         $result_group->{'group_label'} = $group_label;
 626         $result_group->{'group_merge_key'} = $recid;
 627
 628         my $count = 1;
 629         if (exists $hit->{count}) {
 630             $count = $hit->{count}->[0];
 631         }
 632         $result_group->{'group_count'} = $count;
 633
 634         for (my $i = 0; $i < $count; $i++) {
 635             # FIXME -- may need to worry about diacritics here
 636             my $rec = $paz->record($recid, $i);
 637             push @{ $result_group->{'RECORDS'} }, $rec;
 638         }
 639
 640         push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
 641     }
 642
 643     # pass through facets
 644     my $termlist_xml = $paz->termlist('author,subject');
 645     my $terms = XMLin($termlist_xml, forcearray => 1);
 646     my @facets_loop = ();
 647     #die Dumper($results);
 648 #    foreach my $list (sort keys %{ $terms->{'list'} }) {
 649 #        my @facets = ();
 650 #        foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
 651 #            push @facets, {
 652 #                facet_label_value => $facet->{'name'}->[0],
 653 #            };
 654 #        }
 655 #        push @facets_loop, ( {
 656 #            type_label => $list,
 657 #            facets => \@facets,
 658 #        } );
 659 #    }
 660
 661     return ( undef, $results_hashref, \@facets_loop );
 662 }
 663
 664 # STOPWORDS
 665 sub _remove_stopwords {
 666     my ( $operand, $index ) = @_;
 667     my @stopwords_removed;
 668
 669     # phrase and exact-qualified indexes shouldn't have stopwords removed
 670     if ( $index !~ m/phr|ext/ ) {
 671
 672 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
 673 #       we use IsAlpha unicode definition, to deal correctly with diacritics.
 674 #       otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
 675 #       is a stopword, we'd get "çon" and wouldn't find anything...
 676 #
 677                 foreach ( keys %{ C4::Context->stopwords } ) {
 678                         next if ( $_ =~ /(and|or|not)/ );    # don't remove operators
 679                         if ( my ($matched) = ($operand =~
 680                                 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
 681                         {
 682                                 $operand =~ s/\Q$matched\E/ /gi;
 683                                 push @stopwords_removed, $_;
 684                         }
 685                 }
 686         }
 687     return ( $operand, \@stopwords_removed );
 688 }
 689
 690 # TRUNCATION
 691 sub _detect_truncation {
 692     my ( $operand, $index ) = @_;
 693     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 694         @regexpr );
 695     $operand =~ s/^ //g;
 696     my @wordlist = split( /\s/, $operand );
 697     foreach my $word (@wordlist) {
 698         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 699             push @rightlefttruncated, $word;
 700         }
 701         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 702             push @lefttruncated, $word;
 703         }
 704         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 705             push @righttruncated, $word;
 706         }
 707         elsif ( index( $word, "*" ) < 0 ) {
 708             push @nontruncated, $word;
 709         }
 710         else {
 711             push @regexpr, $word;
 712         }
 713     }
 714     return (
 715         \@nontruncated,       \@righttruncated, \@lefttruncated,
 716         \@rightlefttruncated, \@regexpr
 717     );
 718 }
 719
 720 # STEMMING
 721 sub _build_stemmed_operand {
 722     my ($operand,$lang) = @_;
 723     require Lingua::Stem::Snowball ;
 724     my $stemmed_operand;
 725
 726     # If operand contains a digit, it is almost certainly an identifier, and should
 727     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 728     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 729     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 730     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 731     return $operand if $operand =~ /\d/;
 732
 733 # FIXME: the locale should be set based on the user's language and/or search choice
 734     #warn "$lang";
 735     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 736                                                encoding => "UTF-8" );
 737
 738     my @words = split( / /, $operand );
 739     my @stems = $stemmer->stem(\@words);
 740     for my $stem (@stems) {
 741         $stemmed_operand .= "$stem";
 742         $stemmed_operand .= "?"
 743           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
 744         $stemmed_operand .= " ";
 745     }
 746     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
 747     return $stemmed_operand;
 748 }
 749
 750 # FIELD WEIGHTING
 751 sub _build_weighted_query {
 752
 753 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
 754 # pretty well but could work much better if we had a smarter query parser
 755     my ( $operand, $stemmed_operand, $index ) = @_;
 756     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
 757     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
 758     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
 759
 760     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
 761
 762     # Keyword, or, no index specified
 763     if ( ( $index eq 'kw' ) || ( !$index ) ) {
 764         $weighted_query .=
 765           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
 766         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
 767         $weighted_query .= " or ti,phr,r3=\"$operand\"";    # phrase title
 768           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
 769           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
 770         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
 771           if $fuzzy_enabled;    # add fuzzy, word list
 772         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
 773           if ( $stemming and $stemmed_operand )
 774           ;                     # add stemming, right truncation
 775         $weighted_query .= " or wrdl,r9=\"$operand\"";
 776
 777         # embedded sorting: 0 a-z; 1 z-a
 778         # $weighted_query .= ") or (sort1,aut=1";
 779     }
 780
 781     # Barcode searches should skip this process
 782     elsif ( $index eq 'bc' ) {
 783         $weighted_query .= "bc=\"$operand\"";
 784     }
 785
 786     # Authority-number searches should skip this process
 787     elsif ( $index eq 'an' ) {
 788         $weighted_query .= "an=\"$operand\"";
 789     }
 790
 791     # If the index already has more than one qualifier, wrap the operand
 792     # in quotes and pass it back (assumption is that the user knows what they
 793     # are doing and won't appreciate us mucking up their query
 794     elsif ( $index =~ ',' ) {
 795         $weighted_query .= " $index=\"$operand\"";
 796     }
 797
 798     #TODO: build better cases based on specific search indexes
 799     else {
 800         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
 801           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
 802         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
 803         $weighted_query .=
 804           " or $index,rt,wrdl,r3=\"$operand\"";    # word list index
 805     }
 806
 807     $weighted_query .= "))";                       # close rank specification
 808     return $weighted_query;
 809 }
 810
 811 =head2 getIndexes
 812
 813 Return an array with available indexes.
 814
 815 =cut
 816
 817 sub getIndexes{
 818     my @indexes = (
 819                     # biblio indexes
 820                     'ab',
 821                     'Abstract',
 822                     'acqdate',
 823                     'allrecords',
 824                     'an',
 825                     'Any',
 826                     'at',
 827                     'au',
 828                     'aub',
 829                     'aud',
 830                     'audience',
 831                     'auo',
 832                     'aut',
 833                     'Author',
 834                     'Author-in-order ',
 835                     'Author-personal-bibliography',
 836                     'Authority-Number',
 837                     'authtype',
 838                     'bc',
 839                     'Bib-level',
 840                     'biblionumber',
 841                     'bio',
 842                     'biography',
 843                     'callnum',
 844                     'cfn',
 845                     'Chronological-subdivision',
 846                     'cn-bib-source',
 847                     'cn-bib-sort',
 848                     'cn-class',
 849                     'cn-item',
 850                     'cn-prefix',
 851                     'cn-suffix',
 852                     'cpn',
 853                     'Code-institution',
 854                     'Conference-name',
 855                     'Conference-name-heading',
 856                     'Conference-name-see',
 857                     'Conference-name-seealso',
 858                     'Content-type',
 859                     'Control-number',
 860                     'copydate',
 861                     'Corporate-name',
 862                     'Corporate-name-heading',
 863                     'Corporate-name-see',
 864                     'Corporate-name-seealso',
 865                     'ctype',
 866                     'date-entered-on-file',
 867                     'Date-of-acquisition',
 868                     'Date-of-publication',
 869                     'Dewey-classification',
 870                     'EAN',
 871                     'extent',
 872                     'fic',
 873                     'fiction',
 874                     'Form-subdivision',
 875                     'format',
 876                     'Geographic-subdivision',
 877                     'he',
 878                     'Heading',
 879                     'Heading-use-main-or-added-entry',
 880                     'Heading-use-series-added-entry ',
 881                     'Heading-use-subject-added-entry',
 882                     'Host-item',
 883                     'id-other',
 884                     'Illustration-code',
 885                     'ISBN',
 886                     'isbn',
 887                     'ISSN',
 888                     'issn',
 889                     'itemtype',
 890                     'kw',
 891                     'Koha-Auth-Number',
 892                     'l-format',
 893                     'language',
 894                     'lc-card',
 895                     'LC-card-number',
 896                     'lcn',
 897                     'llength',
 898                     'ln',
 899                     'Local-classification',
 900                     'Local-number',
 901                     'Match-heading',
 902                     'Match-heading-see-from',
 903                     'Material-type',
 904                     'mc-itemtype',
 905                     'mc-rtype',
 906                     'mus',
 907                     'name',
 908                     'Music-number',
 909                     'Name-geographic',
 910                     'Name-geographic-heading',
 911                     'Name-geographic-see',
 912                     'Name-geographic-seealso',
 913                     'nb',
 914                     'Note',
 915                     'notes',
 916                     'ns',
 917                     'nt',
 918                     'pb',
 919                     'Personal-name',
 920                     'Personal-name-heading',
 921                     'Personal-name-see',
 922                     'Personal-name-seealso',
 923                     'pl',
 924                     'Place-publication',
 925                     'pn',
 926                     'popularity',
 927                     'pubdate',
 928                     'Publisher',
 929                     'Record-control-number',
 930                     'rcn',
 931                     'Record-type',
 932                     'rtype',
 933                     'se',
 934                     'See',
 935                     'See-also',
 936                     'sn',
 937                     'Stock-number',
 938                     'su',
 939                     'Subject',
 940                     'Subject-heading-thesaurus',
 941                     'Subject-name-personal',
 942                     'Subject-subdivision',
 943                     'Summary',
 944                     'Suppress',
 945                     'su-geo',
 946                     'su-na',
 947                     'su-to',
 948                     'su-ut',
 949                     'ut',
 950                     'UPC',
 951                     'Term-genre-form',
 952                     'Term-genre-form-heading',
 953                     'Term-genre-form-see',
 954                     'Term-genre-form-seealso',
 955                     'ti',
 956                     'Title',
 957                     'Title-cover',
 958                     'Title-series',
 959                     'Title-host',
 960                     'Title-uniform',
 961                     'Title-uniform-heading',
 962                     'Title-uniform-see',
 963                     'Title-uniform-seealso',
 964                     'totalissues',
 965                     'yr',
 966
 967                     # items indexes
 968                     'acqsource',
 969                     'barcode',
 970                     'bc',
 971                     'branch',
 972                     'ccode',
 973                     'classification-source',
 974                     'cn-sort',
 975                     'coded-location-qualifier',
 976                     'copynumber',
 977                     'damaged',
 978                     'datelastborrowed',
 979                     'datelastseen',
 980                     'holdingbranch',
 981                     'homebranch',
 982                     'issues',
 983                     'item',
 984                     'itemnumber',
 985                     'itype',
 986                     'Local-classification',
 987                     'location',
 988                     'lost',
 989                     'materials-specified',
 990                     'mc-ccode',
 991                     'mc-itype',
 992                     'mc-loc',
 993                     'notforloan',
 994                     'onloan',
 995                     'price',
 996                     'renewals',
 997                     'replacementprice',
 998                     'replacementpricedate',
 999                     'reserves',
1000                     'restricted',
1001                     'stack',
1002                     'stocknumber',
1003                     'inv',
1004                     'uri',
1005                     'withdrawn',
1006
1007                     # subject related
1008                   );
1009
1010     return \@indexes;
1011 }
1012
1013 =head2 buildQuery
1014
1015 ( $error, $query,
1016 $simple_query, $query_cgi,
1017 $query_desc, $limit,
1018 $limit_cgi, $limit_desc,
1019 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
1020
1021 Build queries and limits in CCL, CGI, Human,
1022 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
1023
1024 See verbose embedded documentation.
1025
1026
1027 =cut
1028
1029 sub buildQuery {
1030     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
1031
1032     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
1033
1034     # dereference
1035     my @operators = $operators ? @$operators : ();
1036     my @indexes   = $indexes   ? @$indexes   : ();
1037     my @operands  = $operands  ? @$operands  : ();
1038     my @limits    = $limits    ? @$limits    : ();
1039     my @sort_by   = $sort_by   ? @$sort_by   : ();
1040
1041     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1042     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1043     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1044     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1045     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1046
1047     # no stemming/weight/fuzzy in NoZebra
1048     if ( C4::Context->preference("NoZebra") ) {
1049         $stemming         = 0;
1050         $weight_fields    = 0;
1051         $fuzzy_enabled    = 0;
1052         $auto_truncation  = 0;
1053     }
1054
1055     my $query        = $operands[0];
1056     my $simple_query = $operands[0];
1057
1058     # initialize the variables we're passing back
1059     my $query_cgi;
1060     my $query_desc;
1061     my $query_type;
1062
1063     my $limit;
1064     my $limit_cgi;
1065     my $limit_desc;
1066
1067     my $stopwords_removed;    # flag to determine if stopwords have been removed
1068
1069     my $cclq       = 0;
1070     my $cclindexes = getIndexes();
1071     if ( $query !~ /\s*ccl=/ ) {
1072         while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) {
1073             my $dx = lc($1);
1074             $cclq = grep { lc($_) eq $dx } @$cclindexes;
1075         }
1076         $query = "ccl=$query" if $cclq;
1077     }
1078
1079 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1080 # DIAGNOSTIC ONLY!!
1081     if ( $query =~ /^ccl=/ ) {
1082         my $q=$';
1083         # This is needed otherwise ccl= and &limit won't work together, and
1084         # this happens when selecting a subject on the opac-detail page
1085         if (@limits) {
1086             $q .= ' and '.join(' and ', @limits);
1087         }
1088         return ( undef, $q, $q, "q=ccl=$q", $q, '', '', '', '', 'ccl' );
1089     }
1090     if ( $query =~ /^cql=/ ) {
1091         return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1092     }
1093     if ( $query =~ /^pqf=/ ) {
1094         return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1095     }
1096
1097     # pass nested queries directly
1098     # FIXME: need better handling of some of these variables in this case
1099     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1100     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1101 #    if ( $query =~ /(\(|\))/ ) {
1102 #        return (
1103 #            undef,              $query, $simple_query, $query_cgi,
1104 #            $query,             $limit, $limit_cgi,    $limit_desc,
1105 #            $stopwords_removed, 'ccl'
1106 #        );
1107 #    }
1108
1109 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1110 # query operands and indexes and add stemming, truncation, field weighting, etc.
1111 # Once we do so, we'll end up with a value in $query, just like if we had an
1112 # incoming $query from the user
1113     else {
1114         $query = ""
1115           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1116         my $previous_operand
1117           ;    # a flag used to keep track if there was a previous query
1118                # if there was, we can apply the current operator
1119                # for every operand
1120         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1121
1122             # COMBINE OPERANDS, INDEXES AND OPERATORS
1123             if ( $operands[$i] ) {
1124                 $operands[$i]=~s/^\s+//;
1125
1126               # A flag to determine whether or not to add the index to the query
1127                 my $indexes_set;
1128
1129 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1130                 if ( $operands[$i] =~ /\w(:|=)/ || $scan ) {
1131                     $weight_fields    = 0;
1132                     $stemming         = 0;
1133                     $remove_stopwords = 0;
1134                 } else {
1135                     $operands[$i] =~ s/\?/{?}/g; # need to escape question marks
1136                 }
1137                 my $operand = $operands[$i];
1138                 my $index   = $indexes[$i];
1139
1140                 # Add index-specific attributes
1141                 # Date of Publication
1142                 if ( $index eq 'yr' ) {
1143                     $index .= ",st-numeric";
1144                     $indexes_set++;
1145                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1146                 }
1147
1148                 # Date of Acquisition
1149                 elsif ( $index eq 'acqdate' ) {
1150                     $index .= ",st-date-normalized";
1151                     $indexes_set++;
1152                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1153                 }
1154                 # ISBN,ISSN,Standard Number, don't need special treatment
1155                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1156                     (
1157                         $stemming,      $auto_truncation,
1158                         $weight_fields, $fuzzy_enabled,
1159                         $remove_stopwords
1160                     ) = ( 0, 0, 0, 0, 0 );
1161
1162                 }
1163
1164                 if(not $index){
1165                     $index = 'kw';
1166                 }
1167
1168                 # Set default structure attribute (word list)
1169                 my $struct_attr = q{};
1170                 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl|nb|ns)/ ) {
1171                     $struct_attr = ",wrdl";
1172                 }
1173
1174                 # Some helpful index variants
1175                 my $index_plus       = $index . $struct_attr . ':';
1176                 my $index_plus_comma = $index . $struct_attr . ',';
1177
1178                 # Remove Stopwords
1179                 if ($remove_stopwords) {
1180                     ( $operand, $stopwords_removed ) =
1181                       _remove_stopwords( $operand, $index );
1182                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1183                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1184                       if ( $stopwords_removed && $DEBUG );
1185                 }
1186
1187                 if ($auto_truncation){
1188                                         unless ( $index =~ /(st-|phr|ext)/ ) {
1189                                                 #FIXME only valid with LTR scripts
1190                                                 $operand=join(" ",map{
1191                                                                                         (index($_,"*")>0?"$_":"$_*")
1192                                                                                          }split (/\s+/,$operand));
1193                                                 warn $operand if $DEBUG;
1194                                         }
1195                                 }
1196
1197                 # Detect Truncation
1198                 my $truncated_operand;
1199                 my( $nontruncated, $righttruncated, $lefttruncated,
1200                     $rightlefttruncated, $regexpr
1201                 ) = _detect_truncation( $operand, $index );
1202                 warn
1203 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1204                   if $DEBUG;
1205
1206                 # Apply Truncation
1207                 if (
1208                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1209                     scalar(@$rightlefttruncated) > 0 )
1210                 {
1211
1212                # Don't field weight or add the index to the query, we do it here
1213                     $indexes_set = 1;
1214                     undef $weight_fields;
1215                     my $previous_truncation_operand;
1216                     if (scalar @$nontruncated) {
1217                         $truncated_operand .= "$index_plus @$nontruncated ";
1218                         $previous_truncation_operand = 1;
1219                     }
1220                     if (scalar @$righttruncated) {
1221                         $truncated_operand .= "and " if $previous_truncation_operand;
1222                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1223                         $previous_truncation_operand = 1;
1224                     }
1225                     if (scalar @$lefttruncated) {
1226                         $truncated_operand .= "and " if $previous_truncation_operand;
1227                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1228                         $previous_truncation_operand = 1;
1229                     }
1230                     if (scalar @$rightlefttruncated) {
1231                         $truncated_operand .= "and " if $previous_truncation_operand;
1232                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1233                         $previous_truncation_operand = 1;
1234                     }
1235                 }
1236                 $operand = $truncated_operand if $truncated_operand;
1237                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1238
1239                 # Handle Stemming
1240                 my $stemmed_operand;
1241                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1242                                                                                 if $stemming;
1243
1244                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1245
1246                 # Handle Field Weighting
1247                 my $weighted_operand;
1248                 if ($weight_fields) {
1249                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1250                     $operand = $weighted_operand;
1251                     $indexes_set = 1;
1252                 }
1253
1254                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1255
1256                 # If there's a previous operand, we need to add an operator
1257                 if ($previous_operand) {
1258
1259                     # User-specified operator
1260                     if ( $operators[ $i - 1 ] ) {
1261                         $query     .= " $operators[$i-1] ";
1262                         $query     .= " $index_plus " unless $indexes_set;
1263                         $query     .= " $operand";
1264                         $query_cgi .= "&op=$operators[$i-1]";
1265                         $query_cgi .= "&idx=$index" if $index;
1266                         $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1267                         $query_desc .=
1268                           " $operators[$i-1] $index_plus $operands[$i]";
1269                     }
1270
1271                     # Default operator is and
1272                     else {
1273                         $query      .= " and ";
1274                         $query      .= "$index_plus " unless $indexes_set;
1275                         $query      .= "$operand";
1276                         $query_cgi  .= "&op=and&idx=$index" if $index;
1277                         $query_cgi  .= "&q=$operands[$i]" if $operands[$i];
1278                         $query_desc .= " and $index_plus $operands[$i]";
1279                     }
1280                 }
1281
1282                 # There isn't a pervious operand, don't need an operator
1283                 else {
1284
1285                     # Field-weighted queries already have indexes set
1286                     $query .= " $index_plus " unless $indexes_set;
1287                     $query .= $operand;
1288                     $query_desc .= " $index_plus $operands[$i]";
1289                     $query_cgi  .= "&idx=$index" if $index;
1290                     $query_cgi  .= "&q=$operands[$i]" if $operands[$i];
1291                     $previous_operand = 1;
1292                 }
1293             }    #/if $operands
1294         }    # /for
1295     }
1296     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1297
1298     # add limits
1299     my %group_OR_limits;
1300     my $availability_limit;
1301     foreach my $this_limit (@limits) {
1302         if ( $this_limit =~ /available/ ) {
1303 #
1304 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1305 ## In English:
1306 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1307             $availability_limit .=
1308 "( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1309             $limit_cgi  .= "&limit=available";
1310             $limit_desc .= "";
1311         }
1312
1313         # group_OR_limits, prefixed by mc-
1314         # OR every member of the group
1315         elsif ( $this_limit =~ /mc/ ) {
1316             my ($k,$v) = split(/:/, $this_limit,2);
1317             if ( $k !~ /mc-i(tem)?type/ ) {
1318                 # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes
1319                 $this_limit =~ tr/"//d;
1320                 $this_limit = $k.":\"".$v."\"";
1321             }
1322
1323             $group_OR_limits{$k} .= " or " if $group_OR_limits{$k};
1324             $limit_desc      .= " or " if $group_OR_limits{$k};
1325             $group_OR_limits{$k} .= "$this_limit";
1326             $limit_cgi       .= "&limit=$this_limit";
1327             $limit_desc      .= " $this_limit";
1328         }
1329
1330         # Regular old limits
1331         else {
1332             $limit .= " and " if $limit || $query;
1333             $limit      .= "$this_limit";
1334             $limit_cgi  .= "&limit=$this_limit";
1335             if ($this_limit =~ /^branch:(.+)/) {
1336                 my $branchcode = $1;
1337                 my $branchname = GetBranchName($branchcode);
1338                 if (defined $branchname) {
1339                     $limit_desc .= " branch:$branchname";
1340                 } else {
1341                     $limit_desc .= " $this_limit";
1342                 }
1343             } else {
1344                 $limit_desc .= " $this_limit";
1345             }
1346         }
1347     }
1348     foreach my $k (keys (%group_OR_limits)) {
1349         $limit .= " and " if ( $query || $limit );
1350         $limit .= "($group_OR_limits{$k})";
1351     }
1352     if ($availability_limit) {
1353         $limit .= " and " if ( $query || $limit );
1354         $limit .= "($availability_limit)";
1355     }
1356
1357     # Normalize the query and limit strings
1358     # This is flawed , means we can't search anything with : in it
1359     # if user wants to do ccl or cql, start the query with that
1360 #    $query =~ s/:/=/g;
1361     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g;
1362     $query =~ s/(?<=(wrdl)):/=/g;
1363     $query =~ s/(?<=(trn|phr)):/=/g;
1364     $limit =~ s/:/=/g;
1365     for ( $query, $query_desc, $limit, $limit_desc ) {
1366         s/  +/ /g;    # remove extra spaces
1367         s/^ //g;     # remove any beginning spaces
1368         s/ $//g;     # remove any ending spaces
1369         s/==/=/g;    # remove double == from query
1370     }
1371     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1372
1373     for ($query_cgi,$simple_query) {
1374         s/"//g;
1375     }
1376     # append the limit to the query
1377     $query .= " " . $limit;
1378
1379     # Warnings if DEBUG
1380     if ($DEBUG) {
1381         warn "QUERY:" . $query;
1382         warn "QUERY CGI:" . $query_cgi;
1383         warn "QUERY DESC:" . $query_desc;
1384         warn "LIMIT:" . $limit;
1385         warn "LIMIT CGI:" . $limit_cgi;
1386         warn "LIMIT DESC:" . $limit_desc;
1387         warn "---------\nLeave buildQuery\n---------";
1388     }
1389     return (
1390         undef,              $query, $simple_query, $query_cgi,
1391         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1392         $stopwords_removed, $query_type
1393     );
1394 }
1395
1396 =head2 searchResults
1397
1398   my @search_results = searchResults($search_context, $searchdesc, $hits,
1399                                      $results_per_page, $offset, $scan,
1400                                      @marcresults, $hidelostitems);
1401
1402 Format results in a form suitable for passing to the template
1403
1404 =cut
1405
1406 # IMO this subroutine is pretty messy still -- it's responsible for
1407 # building the HTML output for the template
1408 sub searchResults {
1409     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults ) = @_;
1410     my $dbh = C4::Context->dbh;
1411     my @newresults;
1412
1413     $search_context = 'opac' if !$search_context || $search_context ne 'intranet';
1414     my ($is_opac, $hidelostitems);
1415     if ($search_context eq 'opac') {
1416         $hidelostitems = C4::Context->preference('hidelostitems');
1417         $is_opac       = 1;
1418     }
1419
1420     #Build branchnames hash
1421     #find branchname
1422     #get branch information.....
1423     my %branches;
1424     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1425     $bsth->execute();
1426     while ( my $bdata = $bsth->fetchrow_hashref ) {
1427         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1428     }
1429 # FIXME - We build an authorised values hash here, using the default framework
1430 # though it is possible to have different authvals for different fws.
1431
1432     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1433
1434     # get notforloan authorised value list (see $shelflocations  FIXME)
1435     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1436
1437     #Build itemtype hash
1438     #find itemtype & itemtype image
1439     my %itemtypes;
1440     $bsth =
1441       $dbh->prepare(
1442         "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1443       );
1444     $bsth->execute();
1445     while ( my $bdata = $bsth->fetchrow_hashref ) {
1446                 foreach (qw(description imageurl summary notforloan)) {
1447                 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1448                 }
1449     }
1450
1451     #search item field code
1452     my $sth =
1453       $dbh->prepare(
1454 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1455       );
1456     $sth->execute;
1457     my ($itemtag) = $sth->fetchrow;
1458
1459     ## find column names of items related to MARC
1460     my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1461     $sth2->execute;
1462     my %subfieldstosearch;
1463     while ( ( my $column ) = $sth2->fetchrow ) {
1464         my ( $tagfield, $tagsubfield ) =
1465           &GetMarcFromKohaField( "items." . $column, "" );
1466         $subfieldstosearch{$column} = $tagsubfield;
1467     }
1468
1469     # handle which records to actually retrieve
1470     my $times;
1471     if ( $hits && $offset + $results_per_page <= $hits ) {
1472         $times = $offset + $results_per_page;
1473     }
1474     else {
1475         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1476     }
1477
1478         my $marcflavour = C4::Context->preference("marcflavour");
1479     # We get the biblionumber position in MARC
1480     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1481
1482     # loop through all of the records we've retrieved
1483     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1484         my $marcrecord = MARC::File::USMARC::decode( $marcresults->[$i] );
1485         my $fw = $scan
1486              ? undef
1487              : $bibliotag < 10
1488                ? GetFrameworkCode($marcrecord->field($bibliotag)->data)
1489                : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1490         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1491         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1492         $oldbiblio->{result_number} = $i + 1;
1493
1494         # add imageurl to itemtype if there is one
1495         $oldbiblio->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1496
1497         $oldbiblio->{'authorised_value_images'}  = ($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 ) ) : [];
1498                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1499                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1500                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1501                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1502                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1503
1504                 # edition information, if any
1505         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1506                 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1507  # Build summary if there is one (the summary is defined in the itemtypes table)
1508  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1509         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1510             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1511             my @fields  = $marcrecord->fields();
1512
1513             my $newsummary;
1514             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1515                 my $tags = {};
1516                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1517                     $tag =~ /(.{3})(.)/;
1518                     if($marcrecord->field($1)){
1519                         my @abc = $marcrecord->field($1)->subfield($2);
1520                         $tags->{$tag} = $#abc + 1 ;
1521                     }
1522                 }
1523
1524                 # We catch how many times to repeat this line
1525                 my $max = 0;
1526                 foreach my $tag (keys(%$tags)){
1527                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1528                  }
1529
1530                 # we replace, and repeat each line
1531                 for (my $i = 0 ; $i < $max ; $i++){
1532                     my $newline = $line;
1533
1534                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1535                         $tag =~ /(.{3})(.)/;
1536
1537                         if($marcrecord->field($1)){
1538                             my @repl = $marcrecord->field($1)->subfield($2);
1539                             my $subfieldvalue = $repl[$i];
1540
1541                             if (! utf8::is_utf8($subfieldvalue)) {
1542                                 utf8::decode($subfieldvalue);
1543                             }
1544
1545                              $newline =~ s/\[$tag\]/$subfieldvalue/g;
1546                         }
1547                     }
1548                     $newsummary .= "$newline\n";
1549                 }
1550             }
1551
1552             $newsummary =~ s/\[(.*?)]//g;
1553             $newsummary =~ s/\n/<br\/>/g;
1554             $oldbiblio->{summary} = $newsummary;
1555         }
1556
1557         # Pull out the items fields
1558         my @fields = $marcrecord->field($itemtag);
1559         my $marcflavor = C4::Context->preference("marcflavour");
1560         # adding linked items that belong to host records
1561         my $analyticsfield = '773';
1562         if ($marcflavor eq 'MARC21' || $marcflavor eq 'NORMARC') {
1563             $analyticsfield = '773';
1564         } elsif ($marcflavor eq 'UNIMARC') {
1565             $analyticsfield = '461';
1566         }
1567         foreach my $hostfield ( $marcrecord->field($analyticsfield)) {
1568             my $hostbiblionumber = $hostfield->subfield("0");
1569             my $linkeditemnumber = $hostfield->subfield("9");
1570             if(!$hostbiblionumber eq undef){
1571                 my $hostbiblio = GetMarcBiblio($hostbiblionumber, 1);
1572                 my ($itemfield, undef) = GetMarcFromKohaField( 'items.itemnumber', GetFrameworkCode($hostbiblionumber) );
1573                 if(!$hostbiblio eq undef){
1574                     my @hostitems = $hostbiblio->field($itemfield);
1575                     foreach my $hostitem (@hostitems){
1576                         if ($hostitem->subfield("9") eq $linkeditemnumber){
1577                             my $linkeditem =$hostitem;
1578                             # append linked items if they exist
1579                             if (!$linkeditem eq undef){
1580                                 push (@fields, $linkeditem);}
1581                         }
1582                     }
1583                 }
1584             }
1585         }
1586
1587         # Setting item statuses for display
1588         my @available_items_loop;
1589         my @onloan_items_loop;
1590         my @other_items_loop;
1591
1592         my $available_items;
1593         my $onloan_items;
1594         my $other_items;
1595
1596         my $ordered_count         = 0;
1597         my $available_count       = 0;
1598         my $onloan_count          = 0;
1599         my $longoverdue_count     = 0;
1600         my $other_count           = 0;
1601         my $wthdrawn_count        = 0;
1602         my $itemlost_count        = 0;
1603         my $hideatopac_count      = 0;
1604         my $itembinding_count     = 0;
1605         my $itemdamaged_count     = 0;
1606         my $item_in_transit_count = 0;
1607         my $can_place_holds       = 0;
1608         my $item_onhold_count     = 0;
1609         my $items_count           = scalar(@fields);
1610         my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults');
1611         my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1;
1612
1613         # loop through every item
1614               my @hiddenitems;
1615         foreach my $field (@fields) {
1616             my $item;
1617
1618             # populate the items hash
1619             foreach my $code ( keys %subfieldstosearch ) {
1620                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1621             }
1622             $item->{description} = $itemtypes{ $item->{itype} }{description};
1623
1624                 # Hidden items
1625             if ($is_opac) {
1626                     my @hi = GetHiddenItemnumbers($item);
1627                 $item->{'hideatopac'} = @hi;
1628               push @hiddenitems, @hi;
1629             }
1630
1631             my $hbranch     = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch'    : 'holdingbranch';
1632             my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1633
1634             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1635             if ($item->{$hbranch}) {
1636                 $item->{'branchname'} = $branches{$item->{$hbranch}};
1637             }
1638             elsif ($item->{$otherbranch}) {     # Last resort
1639                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1640             }
1641
1642                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1643 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1644             my $userenv = C4::Context->userenv;
1645             if ( $item->{onloan} && !(C4::Members::GetHideLostItemsPreference($userenv->{'number'}) && $item->{itemlost}) ) {
1646                 $onloan_count++;
1647                                 my $key = $prefix . $item->{onloan} . $item->{barcode};
1648                                 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1649                                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1650                                 $onloan_items->{$key}->{branchname} = $item->{branchname};
1651                                 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1652                                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1653                                 $onloan_items->{$key}->{description} = $item->{description};
1654                                 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1655                 # if something's checked out and lost, mark it as 'long overdue'
1656                 if ( $item->{itemlost} ) {
1657                     $onloan_items->{$prefix}->{longoverdue}++;
1658                     $longoverdue_count++;
1659                 } else {        # can place holds as long as item isn't lost
1660                     $can_place_holds = 1;
1661                 }
1662             }
1663
1664          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1665             else {
1666
1667                 # item is on order
1668                 if ( $item->{notforloan} == -1 ) {
1669                     $ordered_count++;
1670                 }
1671
1672                 # is item in transit?
1673                 my $transfertwhen = '';
1674                 my ($transfertfrom, $transfertto);
1675
1676                 # is item on the reserve shelf?
1677                 my $reservestatus = '';
1678                 my $reserveitem;
1679
1680                 unless ($item->{wthdrawn}
1681                         || $item->{itemlost}
1682                         || $item->{damaged}
1683                         || $item->{notforloan}
1684                         || $items_count > 20) {
1685
1686                     # A couple heuristics to limit how many times
1687                     # we query the database for item transfer information, sacrificing
1688                     # accuracy in some cases for speed;
1689                     #
1690                     # 1. don't query if item has one of the other statuses
1691                     # 2. don't check transit status if the bib has
1692                     #    more than 20 items
1693                     #
1694                     # FIXME: to avoid having the query the database like this, and to make
1695                     #        the in transit status count as unavailable for search limiting,
1696                     #        should map transit status to record indexed in Zebra.
1697                     #
1698                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1699                     ($reservestatus, $reserveitem, undef) = C4::Reserves::CheckReserves($item->{itemnumber});
1700                 }
1701
1702                 # item is withdrawn, lost, damaged, not for loan, reserved or in transit
1703                 if (   $item->{wthdrawn}
1704                     || $item->{itemlost}
1705                     || $item->{damaged}
1706                     || $item->{notforloan} > 0
1707                     || $item->{hideatopac}
1708                     || $reservestatus eq 'Waiting'
1709                     || ($transfertwhen ne ''))
1710                 {
1711                     $wthdrawn_count++        if $item->{wthdrawn};
1712                     $itemlost_count++        if $item->{itemlost};
1713                     $itemdamaged_count++     if $item->{damaged};
1714                     $hideatopac_count++      if $item->{hideatopac};
1715                     $item_in_transit_count++ if $transfertwhen ne '';
1716                     $item_onhold_count++     if $reservestatus eq 'Waiting';
1717                     $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1718
1719                     # can place hold on item ?
1720                     if ((!$item->{damaged} || C4::Context->preference('AllowHoldsOnDamagedItems'))
1721                       && !$item->{itemlost}
1722                       && !$item->{withdrawn}
1723                     ) {
1724                         $can_place_holds = 1;
1725                     }
1726
1727                     $other_count++;
1728
1729                     my $key = $prefix . $item->{status};
1730                     foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber hideatopac)) {
1731                         $other_items->{$key}->{$_} = $item->{$_};
1732                     }
1733                     $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0;
1734                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1735                                         $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1736                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
1737                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1738                                         $other_items->{$key}->{description} = $item->{description};
1739                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1740                 }
1741                 # item is available
1742                 else {
1743                     $can_place_holds = 1;
1744                     $available_count++;
1745                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1746                                         foreach (qw(branchname itemcallnumber hideatopac description)) {
1747                         $available_items->{$prefix}->{$_} = $item->{$_};
1748                                         }
1749                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1750                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context, $itemtypes{ $item->{itype} }->{imageurl} );
1751                 }
1752             }
1753         }    # notforloan, item level and biblioitem level
1754         if ($items_count > 0) {
1755         next if $is_opac       && $hideatopac_count >= $items_count;
1756         next if $hidelostitems && $itemlost_count   >= $items_count;
1757         }
1758         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1759         for my $key ( sort keys %$onloan_items ) {
1760             (++$onloanitemscount > $maxitems) and last;
1761             push @onloan_items_loop, $onloan_items->{$key};
1762         }
1763         for my $key ( sort keys %$other_items ) {
1764             (++$otheritemscount > $maxitems) and last;
1765             push @other_items_loop, $other_items->{$key};
1766         }
1767         for my $key ( sort keys %$available_items ) {
1768             (++$availableitemscount > $maxitems) and last;
1769             push @available_items_loop, $available_items->{$key}
1770         }
1771
1772         # XSLT processing of some stuff
1773         use C4::Charset;
1774         SetUTF8Flag($marcrecord);
1775         $debug && warn $marcrecord->as_formatted;
1776         my $interface = $search_context eq 'opac' ? 'OPAC' : '';
1777         if (!$scan && C4::Context->preference($interface . "XSLTResultsDisplay")) {
1778             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1779                                                                 $search_context, 1, \@hiddenitems);
1780             # the last parameter tells Koha to clean up the problematic ampersand entities that Zebra outputs
1781         }
1782
1783         # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either
1784         if (!C4::Context->preference("item-level_itypes")) {
1785             if ($itemtypes{ $oldbiblio->{itemtype} }->{notforloan}) {
1786                 $can_place_holds = 0;
1787             }
1788         }
1789         $oldbiblio->{norequests} = 1 unless $can_place_holds;
1790         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
1791         $oldbiblio->{items_count}          = $items_count;
1792         $oldbiblio->{available_items_loop} = \@available_items_loop;
1793         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
1794         $oldbiblio->{other_items_loop}     = \@other_items_loop;
1795         $oldbiblio->{availablecount}       = $available_count;
1796         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
1797         $oldbiblio->{onloancount}          = $onloan_count;
1798         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
1799         $oldbiblio->{othercount}           = $other_count;
1800         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
1801         $oldbiblio->{wthdrawncount}        = $wthdrawn_count;
1802         $oldbiblio->{itemlostcount}        = $itemlost_count;
1803         $oldbiblio->{damagedcount}         = $itemdamaged_count;
1804         $oldbiblio->{intransitcount}       = $item_in_transit_count;
1805         $oldbiblio->{onholdcount}          = $item_onhold_count;
1806         $oldbiblio->{orderedcount}         = $ordered_count;
1807         # deleting - in isbn to enable amazon content
1808         $oldbiblio->{isbn} =~ s/-//g;
1809
1810         if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) {
1811             my $fieldspec = C4::Context->preference("AlternateHoldingsField");
1812             my $subfields = substr $fieldspec, 3;
1813             my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' ';
1814             my @alternateholdingsinfo = ();
1815             my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3);
1816             my $alternateholdingscount = 0;
1817
1818             for my $field (@holdingsfields) {
1819                 my %holding = ( holding => '' );
1820                 my $havesubfield = 0;
1821                 for my $subfield ($field->subfields()) {
1822                     if ((index $subfields, $$subfield[0]) >= 0) {
1823                         $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0);
1824                         $holding{'holding'} .= $$subfield[1];
1825                         $havesubfield++;
1826                     }
1827                 }
1828                 if ($havesubfield) {
1829                     push(@alternateholdingsinfo, \%holding);
1830                     $alternateholdingscount++;
1831                 }
1832             }
1833
1834             $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo;
1835             $oldbiblio->{'alternateholdings_count'} = $alternateholdingscount;
1836         }
1837
1838         push( @newresults, $oldbiblio );
1839     }
1840
1841     return @newresults;
1842 }
1843
1844 =head2 SearchAcquisitions
1845     Search for acquisitions
1846 =cut
1847
1848 sub SearchAcquisitions{
1849     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1850
1851     my $dbh=C4::Context->dbh;
1852     # Variable initialization
1853     my $str=qq|
1854     SELECT marcxml
1855     FROM biblio
1856     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1857     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1858     WHERE dateaccessioned BETWEEN ? AND ?
1859     |;
1860
1861     my (@params,@loopcriteria);
1862
1863     push @params, $datebegin->output("iso");
1864     push @params, $dateend->output("iso");
1865
1866     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1867         if(C4::Context->preference("item-level_itypes")){
1868             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1869         }else{
1870             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1871         }
1872         push @params, @$itemtypes;
1873     }
1874
1875     if ($criteria =~/itemtype/){
1876         if(C4::Context->preference("item-level_itypes")){
1877             $str .= "AND items.itype=? ";
1878         }else{
1879             $str .= "AND biblioitems.itemtype=? ";
1880         }
1881
1882         if(scalar(@$itemtypes) == 0){
1883             my $itypes = GetItemTypes();
1884             for my $key (keys %$itypes){
1885                 push @$itemtypes, $key;
1886             }
1887         }
1888
1889         @loopcriteria= @$itemtypes;
1890     }elsif ($criteria=~/itemcallnumber/){
1891         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1892                  OR items.itemcallnumber is NULL
1893                  OR items.itemcallnumber = '')";
1894
1895         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1896     }else {
1897         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1898         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1899     }
1900
1901     if ($orderby =~ /date_desc/){
1902         $str.=" ORDER BY dateaccessioned DESC";
1903     } else {
1904         $str.=" ORDER BY title";
1905     }
1906
1907     my $qdataacquisitions=$dbh->prepare($str);
1908
1909     my @loopacquisitions;
1910     foreach my $value(@loopcriteria){
1911         push @params,$value;
1912         my %cell;
1913         $cell{"title"}=$value;
1914         $cell{"titlecode"}=$value;
1915
1916         eval{$qdataacquisitions->execute(@params);};
1917
1918         if ($@){ warn "recentacquisitions Error :$@";}
1919         else {
1920             my @loopdata;
1921             while (my $data=$qdataacquisitions->fetchrow_hashref){
1922                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1923             }
1924             $cell{"loopdata"}=\@loopdata;
1925         }
1926         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1927         pop @params;
1928     }
1929     $qdataacquisitions->finish;
1930     return \@loopacquisitions;
1931 }
1932 #----------------------------------------------------------------------
1933 #
1934 # Non-Zebra GetRecords#
1935 #----------------------------------------------------------------------
1936
1937 =head2 NZgetRecords
1938
1939   NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1940
1941 =cut
1942
1943 sub NZgetRecords {
1944     my (
1945         $query,            $simple_query, $sort_by_ref,    $servers_ref,
1946         $results_per_page, $offset,       $expanded_facet, $branches,
1947         $query_type,       $scan
1948     ) = @_;
1949     warn "query =$query" if $DEBUG;
1950     my $result = NZanalyse($query);
1951     warn "results =$result" if $DEBUG;
1952     return ( undef,
1953         NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1954         undef );
1955 }
1956
1957 =head2 NZanalyse
1958
1959   NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1960   the list is built from an inverted index in the nozebra SQL table
1961   note that title is here only for convenience : the sorting will be very fast when requested on title
1962   if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1963
1964 =cut
1965
1966 sub NZanalyse {
1967     my ( $string, $server ) = @_;
1968 #     warn "---------"       if $DEBUG;
1969     warn " NZanalyse" if $DEBUG;
1970 #     warn "---------"       if $DEBUG;
1971
1972  # $server contains biblioserver or authorities, depending on what we search on.
1973  #warn "querying : $string on $server";
1974     $server = 'biblioserver' unless $server;
1975
1976 # if we have a ", replace the content to discard temporarily any and/or/not inside
1977     my $commacontent;
1978     if ( $string =~ /"/ ) {
1979         $string =~ s/"(.*?)"/__X__/;
1980         $commacontent = $1;
1981         warn "commacontent : $commacontent" if $DEBUG;
1982     }
1983
1984 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1985 # then, call again NZanalyse with $left and $right
1986 # (recursive until we find a leaf (=> something without and/or/not)
1987 # delete repeated operator... Would then go in infinite loop
1988     while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1989     }
1990
1991     #process parenthesis before.
1992     if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1993         my $left     = $1;
1994         my $right    = $4;
1995         my $operator = lc($3);   # FIXME: and/or/not are operators, not operands
1996         warn
1997 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1998           if $DEBUG;
1999         my $leftresult = NZanalyse( $left, $server );
2000         if ($operator) {
2001             my $rightresult = NZanalyse( $right, $server );
2002
2003             # OK, we have the results for right and left part of the query
2004             # depending of operand, intersect, union or exclude both lists
2005             # to get a result list
2006             if ( $operator eq ' and ' ) {
2007                 return NZoperatorAND($leftresult,$rightresult);
2008             }
2009             elsif ( $operator eq ' or ' ) {
2010
2011                 # just merge the 2 strings
2012                 return $leftresult . $rightresult;
2013             }
2014             elsif ( $operator eq ' not ' ) {
2015                 return NZoperatorNOT($leftresult,$rightresult);
2016             }
2017         }
2018         else {
2019 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2020             return $leftresult;
2021         }
2022     }
2023     warn "string :" . $string if $DEBUG;
2024     my $left = "";
2025     my $right = "";
2026     my $operator = "";
2027     if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
2028         $left     = $1;
2029         $right    = $3;
2030         $operator = lc($2);    # FIXME: and/or/not are operators, not operands
2031     }
2032     warn "no parenthesis. left : $left operator: $operator right: $right"
2033       if $DEBUG;
2034
2035     # it's not a leaf, we have a and/or/not
2036     if ($operator) {
2037
2038         # reintroduce comma content if needed
2039         $right =~ s/__X__/"$commacontent"/ if $commacontent;
2040         $left  =~ s/__X__/"$commacontent"/ if $commacontent;
2041         warn "node : $left / $operator / $right\n" if $DEBUG;
2042         my $leftresult  = NZanalyse( $left,  $server );
2043         my $rightresult = NZanalyse( $right, $server );
2044         warn " leftresult : $leftresult" if $DEBUG;
2045         warn " rightresult : $rightresult" if $DEBUG;
2046         # OK, we have the results for right and left part of the query
2047         # depending of operand, intersect, union or exclude both lists
2048         # to get a result list
2049         if ( $operator eq ' and ' ) {
2050             return NZoperatorAND($leftresult,$rightresult);
2051         }
2052         elsif ( $operator eq ' or ' ) {
2053
2054             # just merge the 2 strings
2055             return $leftresult . $rightresult;
2056         }
2057         elsif ( $operator eq ' not ' ) {
2058             return NZoperatorNOT($leftresult,$rightresult);
2059         }
2060         else {
2061
2062 # this error is impossible, because of the regexp that isolate the operand, but just in case...
2063             die "error : operand unknown : $operator for $string";
2064         }
2065
2066         # it's a leaf, do the real SQL query and return the result
2067     }
2068     else {
2069         $string =~ s/__X__/"$commacontent"/ if $commacontent;
2070         $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
2071         #remove trailing blank at the beginning
2072         $string =~ s/^ //g;
2073         warn "leaf:$string" if $DEBUG;
2074
2075         # parse the string in in operator/operand/value again
2076         my $left = "";
2077         my $operator = "";
2078         my $right = "";
2079         if ($string =~ /(.*)(>=|<=)(.*)/) {
2080             $left     = $1;
2081             $operator = $2;
2082             $right    = $3;
2083         } else {
2084             $left = $string;
2085         }
2086 #         warn "handling leaf... left:$left operator:$operator right:$right"
2087 #           if $DEBUG;
2088         unless ($operator) {
2089             if ($string =~ /(.*)(>|<|=)(.*)/) {
2090                 $left     = $1;
2091                 $operator = $2;
2092                 $right    = $3;
2093                 warn
2094     "handling unless (operator)... left:$left operator:$operator right:$right"
2095                 if $DEBUG;
2096             } else {
2097                 $left = $string;
2098             }
2099         }
2100         my $results;
2101
2102 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
2103         $left =~ s/ .*$//;
2104
2105         # automatic replace for short operators
2106         $left = 'title'            if $left =~ '^ti$';
2107         $left = 'author'           if $left =~ '^au$';
2108         $left = 'publisher'        if $left =~ '^pb$';
2109         $left = 'subject'          if $left =~ '^su$';
2110         $left = 'koha-Auth-Number' if $left =~ '^an$';
2111         $left = 'keyword'          if $left =~ '^kw$';
2112         $left = 'itemtype'         if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
2113         warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
2114         my $dbh = C4::Context->dbh;
2115         if ( $operator && $left ne 'keyword' ) {
2116             #do a specific search
2117             $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
2118             my $sth = $dbh->prepare(
2119 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
2120             );
2121             warn "$left / $operator / $right\n" if $DEBUG;
2122
2123             # split each word, query the DB and build the biblionumbers result
2124             #sanitizing leftpart
2125             $left =~ s/^\s+|\s+$//;
2126             foreach ( split / /, $right ) {
2127                 my $biblionumbers;
2128                 $_ =~ s/^\s+|\s+$//;
2129                 next unless $_;
2130                 warn "EXECUTE : $server, $left, $_" if $DEBUG;
2131                 $sth->execute( $server, $left, $_ )
2132                   or warn "execute failed: $!";
2133                 while ( my ( $line, $value ) = $sth->fetchrow ) {
2134
2135 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2136 # otherwise, fill the result
2137                     $biblionumbers .= $line
2138                       unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2139                     warn "result : $value "
2140                       . ( $right  =~ /\d/ ) . "=="
2141                       . ( $value =~ /\D/?$line:"" ) if $DEBUG;         #= $line";
2142                 }
2143
2144 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2145                 if ($results) {
2146                     warn "NZAND" if $DEBUG;
2147                     $results = NZoperatorAND($biblionumbers,$results);
2148                 } else {
2149                     $results = $biblionumbers;
2150                 }
2151             }
2152         }
2153         else {
2154       #do a complete search (all indexes), if index='kw' do complete search too.
2155             my $sth = $dbh->prepare(
2156 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2157             );
2158
2159             # split each word, query the DB and build the biblionumbers result
2160             foreach ( split / /, $string ) {
2161                 next if C4::Context->stopwords->{ uc($_) };   # skip if stopword
2162                 warn "search on all indexes on $_" if $DEBUG;
2163                 my $biblionumbers;
2164                 next unless $_;
2165                 $sth->execute( $server, $_ );
2166                 while ( my $line = $sth->fetchrow ) {
2167                     $biblionumbers .= $line;
2168                 }
2169
2170 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2171                 if ($results) {
2172                     $results = NZoperatorAND($biblionumbers,$results);
2173                 }
2174                 else {
2175                     warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2176                     $results = $biblionumbers;
2177                 }
2178             }
2179         }
2180         warn "return : $results for LEAF : $string" if $DEBUG;
2181         return $results;
2182     }
2183     warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2184 }
2185
2186 sub NZoperatorAND{
2187     my ($rightresult, $leftresult)=@_;
2188
2189     my @leftresult = split /;/, $leftresult;
2190     warn " @leftresult / $rightresult \n" if $DEBUG;
2191
2192     #             my @rightresult = split /;/,$leftresult;
2193     my $finalresult;
2194
2195 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2196 # the result is stored twice, to have the same weight for AND than OR.
2197 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2198 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2199     foreach (@leftresult) {
2200         my $value = $_;
2201         my $countvalue;
2202         ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2203         if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2204             $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2205             $finalresult .=
2206                 "$value-$countvalue;$value-$countvalue;";
2207         }
2208     }
2209     warn "NZAND DONE : $finalresult \n" if $DEBUG;
2210     return $finalresult;
2211 }
2212
2213 sub NZoperatorOR{
2214     my ($rightresult, $leftresult)=@_;
2215     return $rightresult.$leftresult;
2216 }
2217
2218 sub NZoperatorNOT{
2219     my ($leftresult, $rightresult)=@_;
2220
2221     my @leftresult = split /;/, $leftresult;
2222
2223     #             my @rightresult = split /;/,$leftresult;
2224     my $finalresult;
2225     foreach (@leftresult) {
2226         my $value=$_;
2227         $value=$1 if $value=~m/(.*)-\d+$/;
2228         unless ($rightresult =~ "$value-") {
2229             $finalresult .= "$_;";
2230         }
2231     }
2232     return $finalresult;
2233 }
2234
2235 =head2 NZorder
2236
2237   $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2238
2239   TODO :: Description
2240
2241 =cut
2242
2243 sub NZorder {
2244     my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2245     warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2246
2247     # order title asc by default
2248     #     $ordering = '1=36 <i' unless $ordering;
2249     $results_per_page = 20 unless $results_per_page;
2250     $offset           = 0  unless $offset;
2251     my $dbh = C4::Context->dbh;
2252
2253     #
2254     # order by POPULARITY
2255     #
2256     if ( $ordering =~ /popularity/ ) {
2257         my %result;
2258         my %popularity;
2259
2260         # popularity is not in MARC record, it's builded from a specific query
2261         my $sth =
2262           $dbh->prepare("select sum(issues) from items where biblionumber=?");
2263         foreach ( split /;/, $biblionumbers ) {
2264             my ( $biblionumber, $title ) = split /,/, $_;
2265             $result{$biblionumber} = GetMarcBiblio($biblionumber);
2266             $sth->execute($biblionumber);
2267             my $popularity = $sth->fetchrow || 0;
2268
2269 # hint : the key is popularity.title because we can have
2270 # many results with the same popularity. In this case, sub-ordering is done by title
2271 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2272 # (un-frequent, I agree, but we won't forget anything that way ;-)
2273             $popularity{ sprintf( "%10d", $popularity ) . $title
2274                   . $biblionumber } = $biblionumber;
2275         }
2276
2277     # sort the hash and return the same structure as GetRecords (Zebra querying)
2278         my $result_hash;
2279         my $numbers = 0;
2280         if ( $ordering eq 'popularity_dsc' ) {    # sort popularity DESC
2281             foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2282                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2283                   $result{ $popularity{$key} }->as_usmarc();
2284             }
2285         }
2286         else {                                    # sort popularity ASC
2287             foreach my $key ( sort ( keys %popularity ) ) {
2288                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2289                   $result{ $popularity{$key} }->as_usmarc();
2290             }
2291         }
2292         my $finalresult = ();
2293         $result_hash->{'hits'}         = $numbers;
2294         $finalresult->{'biblioserver'} = $result_hash;
2295         return $finalresult;
2296
2297         #
2298         # ORDER BY author
2299         #
2300     }
2301     elsif ( $ordering =~ /author/ ) {
2302         my %result;
2303         foreach ( split /;/, $biblionumbers ) {
2304             my ( $biblionumber, $title ) = split /,/, $_;
2305             my $record = GetMarcBiblio($biblionumber);
2306             my $author;
2307             if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2308                 $author = $record->subfield( '200', 'f' );
2309                 $author = $record->subfield( '700', 'a' ) unless $author;
2310             }
2311             else {
2312                 $author = $record->subfield( '100', 'a' );
2313             }
2314
2315 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2316 # and we don't want to get only 1 result for each of them !!!
2317             $result{ $author . $biblionumber } = $record;
2318         }
2319
2320     # sort the hash and return the same structure as GetRecords (Zebra querying)
2321         my $result_hash;
2322         my $numbers = 0;
2323         if ( $ordering eq 'author_za' || $ordering eq 'author_dsc' ) {    # sort by author desc
2324             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2325                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2326                   $result{$key}->as_usmarc();
2327             }
2328         }
2329         else {                               # sort by author ASC
2330             foreach my $key ( sort ( keys %result ) ) {
2331                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2332                   $result{$key}->as_usmarc();
2333             }
2334         }
2335         my $finalresult = ();
2336         $result_hash->{'hits'}         = $numbers;
2337         $finalresult->{'biblioserver'} = $result_hash;
2338         return $finalresult;
2339
2340         #
2341         # ORDER BY callnumber
2342         #
2343     }
2344     elsif ( $ordering =~ /callnumber/ ) {
2345         my %result;
2346         foreach ( split /;/, $biblionumbers ) {
2347             my ( $biblionumber, $title ) = split /,/, $_;
2348             my $record = GetMarcBiblio($biblionumber);
2349             my $callnumber;
2350             my $frameworkcode = GetFrameworkCode($biblionumber);
2351             my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField(  'items.itemcallnumber', $frameworkcode);
2352                ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2353                 unless $callnumber_tag;
2354             if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2355                 $callnumber = $record->subfield( '200', 'f' );
2356             } else {
2357                 $callnumber = $record->subfield( '100', 'a' );
2358             }
2359
2360 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2361 # and we don't want to get only 1 result for each of them !!!
2362             $result{ $callnumber . $biblionumber } = $record;
2363         }
2364
2365     # sort the hash and return the same structure as GetRecords (Zebra querying)
2366         my $result_hash;
2367         my $numbers = 0;
2368         if ( $ordering eq 'call_number_dsc' ) {    # sort by title desc
2369             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2370                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2371                   $result{$key}->as_usmarc();
2372             }
2373         }
2374         else {                                     # sort by title ASC
2375             foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2376                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2377                   $result{$key}->as_usmarc();
2378             }
2379         }
2380         my $finalresult = ();
2381         $result_hash->{'hits'}         = $numbers;
2382         $finalresult->{'biblioserver'} = $result_hash;
2383         return $finalresult;
2384     }
2385     elsif ( $ordering =~ /pubdate/ ) {             #pub year
2386         my %result;
2387         foreach ( split /;/, $biblionumbers ) {
2388             my ( $biblionumber, $title ) = split /,/, $_;
2389             my $record = GetMarcBiblio($biblionumber);
2390             my ( $publicationyear_tag, $publicationyear_subfield ) =
2391               GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2392             my $publicationyear =
2393               $record->subfield( $publicationyear_tag,
2394                 $publicationyear_subfield );
2395
2396 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2397 # and we don't want to get only 1 result for each of them !!!
2398             $result{ $publicationyear . $biblionumber } = $record;
2399         }
2400
2401     # sort the hash and return the same structure as GetRecords (Zebra querying)
2402         my $result_hash;
2403         my $numbers = 0;
2404         if ( $ordering eq 'pubdate_dsc' ) {    # sort by pubyear desc
2405             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2406                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2407                   $result{$key}->as_usmarc();
2408             }
2409         }
2410         else {                                 # sort by pub year ASC
2411             foreach my $key ( sort ( keys %result ) ) {
2412                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2413                   $result{$key}->as_usmarc();
2414             }
2415         }
2416         my $finalresult = ();
2417         $result_hash->{'hits'}         = $numbers;
2418         $finalresult->{'biblioserver'} = $result_hash;
2419         return $finalresult;
2420
2421         #
2422         # ORDER BY title
2423         #
2424     }
2425     elsif ( $ordering =~ /title/ ) {
2426
2427 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2428         my %result;
2429         foreach ( split /;/, $biblionumbers ) {
2430             my ( $biblionumber, $title ) = split /,/, $_;
2431
2432 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2433 # and we don't want to get only 1 result for each of them !!!
2434 # hint & speed improvement : we can order without reading the record
2435 # so order, and read records only for the requested page !
2436             $result{ $title . $biblionumber } = $biblionumber;
2437         }
2438
2439     # sort the hash and return the same structure as GetRecords (Zebra querying)
2440         my $result_hash;
2441         my $numbers = 0;
2442         if ( $ordering eq 'title_az' ) {    # sort by title desc
2443             foreach my $key ( sort ( keys %result ) ) {
2444                 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2445             }
2446         }
2447         else {                              # sort by title ASC
2448             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2449                 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2450             }
2451         }
2452
2453         # limit the $results_per_page to result size if it's more
2454         $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2455
2456         # for the requested page, replace biblionumber by the complete record
2457         # speed improvement : avoid reading too much things
2458         for (
2459             my $counter = $offset ;
2460             $counter <= $offset + $results_per_page ;
2461             $counter++
2462           )
2463         {
2464             $result_hash->{'RECORDS'}[$counter] =
2465               GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2466         }
2467         my $finalresult = ();
2468         $result_hash->{'hits'}         = $numbers;
2469         $finalresult->{'biblioserver'} = $result_hash;
2470         return $finalresult;
2471     }
2472     else {
2473
2474 #
2475 # order by ranking
2476 #
2477 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2478         my %result;
2479         my %count_ranking;
2480         foreach ( split /;/, $biblionumbers ) {
2481             my ( $biblionumber, $title ) = split /,/, $_;
2482             $title =~ /(.*)-(\d)/;
2483
2484             # get weight
2485             my $ranking = $2;
2486
2487 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2488 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2489 # biblio N has ranking = 6
2490             $count_ranking{$biblionumber} += $ranking;
2491         }
2492
2493 # build the result by "inverting" the count_ranking hash
2494 # hing : as usual, we don't order by ranking only, to avoid having only 1 result for each rank. We build an hash on concat(ranking,biblionumber) instead
2495 #         warn "counting";
2496         foreach ( keys %count_ranking ) {
2497             $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2498         }
2499
2500     # sort the hash and return the same structure as GetRecords (Zebra querying)
2501         my $result_hash;
2502         my $numbers = 0;
2503         foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2504             $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2505         }
2506
2507         # limit the $results_per_page to result size if it's more
2508         $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2509
2510         # for the requested page, replace biblionumber by the complete record
2511         # speed improvement : avoid reading too much things
2512         for (
2513             my $counter = $offset ;
2514             $counter <= $offset + $results_per_page ;
2515             $counter++
2516           )
2517         {
2518             $result_hash->{'RECORDS'}[$counter] =
2519               GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2520               if $result_hash->{'RECORDS'}[$counter];
2521         }
2522         my $finalresult = ();
2523         $result_hash->{'hits'}         = $numbers;
2524         $finalresult->{'biblioserver'} = $result_hash;
2525         return $finalresult;
2526     }
2527 }
2528
2529 =head2 enabled_staff_search_views
2530
2531 %hash = enabled_staff_search_views()
2532
2533 This function returns a hash that contains three flags obtained from the system
2534 preferences, used to determine whether a particular staff search results view
2535 is enabled.
2536
2537 =over 2
2538
2539 =item C<Output arg:>
2540
2541     * $hash{can_view_MARC} is true only if the MARC view is enabled
2542     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2543     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2544
2545 =item C<usage in the script:>
2546
2547 =back
2548
2549 $template->param ( C4::Search::enabled_staff_search_views );
2550
2551 =cut
2552
2553 sub enabled_staff_search_views
2554 {
2555         return (
2556                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2557                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2558                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2559         );
2560 }
2561
2562 sub AddSearchHistory{
2563         my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2564     my $dbh = C4::Context->dbh;
2565
2566     # Add the request the user just made
2567     my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2568     my $sth   = $dbh->prepare($sql);
2569     $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2570         return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2571 }
2572
2573 sub GetSearchHistory{
2574         my ($borrowernumber,$session)=@_;
2575     my $dbh = C4::Context->dbh;
2576
2577     # Add the request the user just made
2578     my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2579     my $sth   = $dbh->prepare($query);
2580         $sth->execute($borrowernumber, $session);
2581     return  $sth->fetchall_hashref({});
2582 }
2583
2584 =head2 z3950_search_args
2585
2586 $arrayref = z3950_search_args($matchpoints)
2587
2588 This function returns an array reference that contains the search parameters to be
2589 passed to the Z39.50 search script (z3950_search.pl). The array elements
2590 are hash refs whose keys are name, value and encvalue, and whose values are the
2591 name of a search parameter, the value of that search parameter and the URL encoded
2592 value of that parameter.
2593
2594 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2595
2596 The search parameter values are obtained from the bibliographic record whose
2597 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2598
2599 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2600 a general purpose search argument. In this case, the returned array contains only
2601 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2602
2603 If a search parameter value is undefined or empty, it is not included in the returned
2604 array.
2605
2606 The returned array reference may be passed directly to the template parameters.
2607
2608 =over 2
2609
2610 =item C<Output arg:>
2611
2612     * $array containing hash refs as described above
2613
2614 =item C<usage in the script:>
2615
2616 =back
2617
2618 $data = Biblio::GetBiblioData($bibno);
2619 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2620
2621 *OR*
2622
2623 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2624
2625 =cut
2626
2627 sub z3950_search_args {
2628     my $bibrec = shift;
2629     my $isbn = Business::ISBN->new($bibrec);
2630
2631     if (defined $isbn && $isbn->is_valid)
2632     {
2633         $bibrec = { isbn => $bibrec } if !ref $bibrec;
2634     }
2635     else {
2636         $bibrec = { title => $bibrec } if !ref $bibrec;
2637     }
2638     my $array = [];
2639     for my $field (qw/ lccn isbn issn title author dewey subject /)
2640     {
2641         my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2642         push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2643     }
2644     return $array;
2645 }
2646
2647 =head2 GetDistinctValues($field);
2648
2649 C<$field> is a reference to the fields array
2650
2651 =cut
2652
2653 sub GetDistinctValues {
2654     my ($fieldname,$string)=@_;
2655     # returns a reference to a hash of references to branches...
2656     if ($fieldname=~/\./){
2657                         my ($table,$column)=split /\./, $fieldname;
2658                         my $dbh = C4::Context->dbh;
2659                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2660                         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 ");
2661                         $sth->execute;
2662                         my $elements=$sth->fetchall_arrayref({});
2663                         return $elements;
2664    }
2665    else {
2666                 $string||= qq("");
2667                 my @servers=qw<biblioserver authorityserver>;
2668                 my (@zconns,@results);
2669         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2670                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2671                         $results[$i] =
2672                       $zconns[$i]->scan(
2673                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2674                       );
2675                 }
2676                 # The big moment: asynchronously retrieve results from all servers
2677                 my @elements;
2678                 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2679                         my $ev = $zconns[ $i - 1 ]->last_event();
2680                         if ( $ev == ZOOM::Event::ZEND ) {
2681                                 next unless $results[ $i - 1 ];
2682                                 my $size = $results[ $i - 1 ]->size();
2683                                 if ( $size > 0 ) {
2684                       for (my $j=0;$j<$size;$j++){
2685                                                 my %hashscan;
2686                                                 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2687                                                 push @elements, \%hashscan;
2688                                           }
2689                                 }
2690                         }
2691                 }
2692                 return \@elements;
2693    }
2694 }
2695
2696
2697 END { }    # module clean-up code here (global destructor)
2698
2699 1;
2700 __END__
2701
2702 =head1 AUTHOR
2703
2704 Koha Development Team <http://koha-community.org/>
2705
2706 =cut