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::XSLT;
  29 use C4::Branch;
  30 use C4::Reserves;    # CheckReserves
  31 use C4::Debug;
  32 use URI::Escape;
  33
  34 use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS $DEBUG);
  35
  36 # set the version for version checking
  37 BEGIN {
  38     $VERSION = 3.01;
  39     $DEBUG = ($ENV{DEBUG}) ? 1 : 0;
  40 }
  41
  42 =head1 NAME
  43
  44 C4::Search - Functions for searching the Koha catalog.
  45
  46 =head1 SYNOPSIS
  47
  48 See opac/opac-search.pl or catalogue/search.pl for example of usage
  49
  50 =head1 DESCRIPTION
  51
  52 This module provides searching functions for Koha's bibliographic databases
  53
  54 =head1 FUNCTIONS
  55
  56 =cut
  57
  58 @ISA    = qw(Exporter);
  59 @EXPORT = qw(
  60   &FindDuplicate
  61   &SimpleSearch
  62   &searchResults
  63   &getRecords
  64   &buildQuery
  65   &NZgetRecords
  66   &AddSearchHistory
  67   &GetDistinctValues
  68   &BiblioAddAuthorities
  69 );
  70 #FIXME: i had to add BiblioAddAuthorities here because in Biblios.pm it caused circular dependencies (C4::Search uses C4::Biblio, and BiblioAddAuthorities uses SimpleSearch from C4::Search)
  71
  72 # make all your functions, whether exported or not;
  73
  74 =head2 FindDuplicate
  75
  76 ($biblionumber,$biblionumber,$title) = FindDuplicate($record);
  77
  78 This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm
  79
  80 =cut
  81
  82 sub FindDuplicate {
  83     my ($record) = @_;
  84     my $dbh = C4::Context->dbh;
  85     my $result = TransformMarcToKoha( $dbh, $record, '' );
  86     my $sth;
  87     my $query;
  88     my $search;
  89     my $type;
  90     my ( $biblionumber, $title );
  91
  92     # search duplicate on ISBN, easy and fast..
  93     # ... normalize first
  94     if ( $result->{isbn} ) {
  95         $result->{isbn} =~ s/\(.*$//;
  96         $result->{isbn} =~ s/\s+$//;
  97         $query = "isbn=$result->{isbn}";
  98     }
  99     else {
 100         $result->{title} =~ s /\\//g;
 101         $result->{title} =~ s /\"//g;
 102         $result->{title} =~ s /\(//g;
 103         $result->{title} =~ s /\)//g;
 104
 105         # FIXME: instead of removing operators, could just do
 106         # quotes around the value
 107         $result->{title} =~ s/(and|or|not)//g;
 108         $query = "ti,ext=$result->{title}";
 109         $query .= " and itemtype=$result->{itemtype}"
 110           if ( $result->{itemtype} );
 111         if   ( $result->{author} ) {
 112             $result->{author} =~ s /\\//g;
 113             $result->{author} =~ s /\"//g;
 114             $result->{author} =~ s /\(//g;
 115             $result->{author} =~ s /\)//g;
 116
 117             # remove valid operators
 118             $result->{author} =~ s/(and|or|not)//g;
 119             $query .= " and au,ext=$result->{author}";
 120         }
 121     }
 122
 123     # FIXME: add error handling
 124     my ( $error, $searchresults ) = SimpleSearch($query); # FIXME :: hardcoded !
 125     my @results;
 126     foreach my $possible_duplicate_record (@$searchresults) {
 127         my $marcrecord =
 128           MARC::Record->new_from_usmarc($possible_duplicate_record);
 129         my $result = TransformMarcToKoha( $dbh, $marcrecord, '' );
 130
 131         # FIXME :: why 2 $biblionumber ?
 132         if ($result) {
 133             push @results, $result->{'biblionumber'};
 134             push @results, $result->{'title'};
 135         }
 136     }
 137     return @results;
 138 }
 139
 140 =head2 SimpleSearch
 141
 142 ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers] );
 143
 144 This function provides a simple search API on the bibliographic catalog
 145
 146 =over 2
 147
 148 =item C<input arg:>
 149
 150     * $query can be a simple keyword or a complete CCL query
 151     * @servers is optional. Defaults to biblioserver as found in koha-conf.xml
 152     * $offset - If present, represents the number of records at the beggining to omit. Defaults to 0
 153     * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef.
 154
 155
 156 =item C<Output:>
 157
 158     * $error is a empty unless an error is detected
 159     * \@results is an array of records.
 160     * $total_hits is the number of hits that would have been returned with no limit
 161
 162 =item C<usage in the script:>
 163
 164 =back
 165
 166 my ( $error, $marcresults, $total_hits ) = SimpleSearch($query);
 167
 168 if (defined $error) {
 169     $template->param(query_error => $error);
 170     warn "error: ".$error;
 171     output_html_with_http_headers $input, $cookie, $template->output;
 172     exit;
 173 }
 174
 175 my $hits = scalar @$marcresults;
 176 my @results;
 177
 178 for my $i (0..$hits) {
 179     my %resultsloop;
 180     my $marcrecord = MARC::File::USMARC::decode($marcresults->[$i]);
 181     my $biblio = TransformMarcToKoha(C4::Context->dbh,$marcrecord,'');
 182
 183     #build the hash for the template.
 184     $resultsloop{title}           = $biblio->{'title'};
 185     $resultsloop{subtitle}        = $biblio->{'subtitle'};
 186     $resultsloop{biblionumber}    = $biblio->{'biblionumber'};
 187     $resultsloop{author}          = $biblio->{'author'};
 188     $resultsloop{publishercode}   = $biblio->{'publishercode'};
 189     $resultsloop{publicationyear} = $biblio->{'publicationyear'};
 190
 191     push @results, \%resultsloop;
 192 }
 193
 194 $template->param(result=>\@results);
 195
 196 =cut
 197
 198 sub SimpleSearch {
 199     my ( $query, $offset, $max_results, $servers )  = @_;
 200
 201     if ( C4::Context->preference('NoZebra') ) {
 202         my $result = NZorder( NZanalyse($query) )->{'biblioserver'};
 203         my $search_result =
 204           (      $result->{hits}
 205               && $result->{hits} > 0 ? $result->{'RECORDS'} : [] );
 206         return ( undef, $search_result, scalar($result->{hits}) );
 207     }
 208     else {
 209         # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too.
 210         my @servers = defined ( $servers ) ? @$servers : ( "biblioserver" );
 211         my @results;
 212         my @zoom_queries;
 213         my @tmpresults;
 214         my @zconns;
 215         my $total_hits;
 216         return ( "No query entered", undef, undef ) unless $query;
 217
 218         # Initialize & Search Zebra
 219         for ( my $i = 0 ; $i < @servers ; $i++ ) {
 220             eval {
 221                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 222                 $zoom_queries[$i] = new ZOOM::Query::CCL2RPN( $query, $zconns[$i]);
 223                 $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] );
 224
 225                 # error handling
 226                 my $error =
 227                     $zconns[$i]->errmsg() . " ("
 228                   . $zconns[$i]->errcode() . ") "
 229                   . $zconns[$i]->addinfo() . " "
 230                   . $zconns[$i]->diagset();
 231
 232                 return ( $error, undef, undef ) if $zconns[$i]->errcode();
 233             };
 234             if ($@) {
 235
 236                 # caught a ZOOM::Exception
 237                 my $error =
 238                     $@->message() . " ("
 239                   . $@->code() . ") "
 240                   . $@->addinfo() . " "
 241                   . $@->diagset();
 242                 warn $error;
 243                 return ( $error, undef, undef );
 244             }
 245         }
 246         while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
 247             my $event = $zconns[ $i - 1 ]->last_event();
 248             if ( $event == ZOOM::Event::ZEND ) {
 249
 250                 my $first_record = defined( $offset ) ? $offset+1 : 1;
 251                 my $hits = $tmpresults[ $i - 1 ]->size();
 252                 $total_hits += $hits;
 253                 my $last_record = $hits;
 254                 if ( defined $max_results && $offset + $max_results < $hits ) {
 255                     $last_record  = $offset + $max_results;
 256                 }
 257
 258                 for my $j ( $first_record..$last_record ) {
 259                     my $record = $tmpresults[ $i - 1 ]->record( $j-1 )->raw(); # 0 indexed
 260                     push @results, $record;
 261                 }
 262             }
 263         }
 264
 265         foreach my $result (@tmpresults) {
 266             $result->destroy();
 267         }
 268         foreach my $zoom_query (@zoom_queries) {
 269             $zoom_query->destroy();
 270         }
 271
 272         return ( undef, \@results, $total_hits );
 273     }
 274 }
 275
 276 =head2 getRecords
 277
 278 ( undef, $results_hashref, \@facets_loop ) = getRecords (
 279
 280         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 281         $results_per_page, $offset,       $expanded_facet, $branches,
 282         $query_type,       $scan
 283     );
 284
 285 The all singing, all dancing, multi-server, asynchronous, scanning,
 286 searching, record nabbing, facet-building
 287
 288 See verbse embedded documentation.
 289
 290 =cut
 291
 292 sub getRecords {
 293     my (
 294         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 295         $results_per_page, $offset,       $expanded_facet, $branches,
 296         $query_type,       $scan
 297     ) = @_;
 298
 299     my @servers = @$servers_ref;
 300     my @sort_by = @$sort_by_ref;
 301
 302     # Initialize variables for the ZOOM connection and results object
 303     my $zconn;
 304     my @zconns;
 305     my @results;
 306     my $results_hashref = ();
 307
 308     # Initialize variables for the faceted results objects
 309     my $facets_counter = ();
 310     my $facets_info    = ();
 311     my $facets         = getFacets();
 312
 313     my @facets_loop;    # stores the ref to array of hashes for template facets loop
 314
 315     ### LOOP THROUGH THE SERVERS
 316     for ( my $i = 0 ; $i < @servers ; $i++ ) {
 317         $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
 318
 319 # perform the search, create the results objects
 320 # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query
 321         my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query;
 322
 323         #$query_to_use = $simple_query if $scan;
 324         warn $simple_query if ( $scan and $DEBUG );
 325
 326         # Check if we've got a query_type defined, if so, use it
 327         eval {
 328             if ($query_type) {
 329                 if ($query_type =~ /^ccl/) {
 330                     $query_to_use =~ s/\:/\=/g;    # change : to = last minute (FIXME)
 331                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 332                 } elsif ($query_type =~ /^cql/) {
 333                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CQL($query_to_use, $zconns[$i]));
 334                 } elsif ($query_type =~ /^pqf/) {
 335                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::PQF($query_to_use, $zconns[$i]));
 336                 } else {
 337                     warn "Unknown query_type '$query_type'.  Results undetermined.";
 338                 }
 339             } elsif ($scan) {
 340                     $results[$i] = $zconns[$i]->scan(  new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 341             } else {
 342                     $results[$i] = $zconns[$i]->search(new ZOOM::Query::CCL2RPN($query_to_use, $zconns[$i]));
 343             }
 344         };
 345         if ($@) {
 346             warn "WARNING: query problem with $query_to_use " . $@;
 347         }
 348
 349         # Concatenate the sort_by limits and pass them to the results object
 350         # Note: sort will override rank
 351         my $sort_by;
 352         foreach my $sort (@sort_by) {
 353             if ( $sort eq "author_az" ) {
 354                 $sort_by .= "1=1003 <i ";
 355             }
 356             elsif ( $sort eq "author_za" ) {
 357                 $sort_by .= "1=1003 >i ";
 358             }
 359             elsif ( $sort eq "popularity_asc" ) {
 360                 $sort_by .= "1=9003 <i ";
 361             }
 362             elsif ( $sort eq "popularity_dsc" ) {
 363                 $sort_by .= "1=9003 >i ";
 364             }
 365             elsif ( $sort eq "call_number_asc" ) {
 366                 $sort_by .= "1=20  <i ";
 367             }
 368             elsif ( $sort eq "call_number_dsc" ) {
 369                 $sort_by .= "1=20 >i ";
 370             }
 371             elsif ( $sort eq "pubdate_asc" ) {
 372                 $sort_by .= "1=31 <i ";
 373             }
 374             elsif ( $sort eq "pubdate_dsc" ) {
 375                 $sort_by .= "1=31 >i ";
 376             }
 377             elsif ( $sort eq "acqdate_asc" ) {
 378                 $sort_by .= "1=32 <i ";
 379             }
 380             elsif ( $sort eq "acqdate_dsc" ) {
 381                 $sort_by .= "1=32 >i ";
 382             }
 383             elsif ( $sort eq "title_az" ) {
 384                 $sort_by .= "1=4 <i ";
 385             }
 386             elsif ( $sort eq "title_za" ) {
 387                 $sort_by .= "1=4 >i ";
 388             }
 389             else {
 390                 warn "Ignoring unrecognized sort '$sort' requested" if $sort_by;
 391             }
 392         }
 393         if ($sort_by) {
 394             if ( $results[$i]->sort( "yaz", $sort_by ) < 0 ) {
 395                 warn "WARNING sort $sort_by failed";
 396             }
 397         }
 398     }    # finished looping through servers
 399
 400     # The big moment: asynchronously retrieve results from all servers
 401     while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
 402         my $ev = $zconns[ $i - 1 ]->last_event();
 403         if ( $ev == ZOOM::Event::ZEND ) {
 404             next unless $results[ $i - 1 ];
 405             my $size = $results[ $i - 1 ]->size();
 406             if ( $size > 0 ) {
 407                 my $results_hash;
 408
 409                 # loop through the results
 410                 $results_hash->{'hits'} = $size;
 411                 my $times;
 412                 if ( $offset + $results_per_page <= $size ) {
 413                     $times = $offset + $results_per_page;
 414                 }
 415                 else {
 416                     $times = $size;
 417                 }
 418                 for ( my $j = $offset ; $j < $times ; $j++ ) {
 419                     my $records_hash;
 420                     my $record;
 421                     my $facet_record;
 422
 423                     ## Check if it's an index scan
 424                     if ($scan) {
 425                         my ( $term, $occ ) = $results[ $i - 1 ]->term($j);
 426
 427                  # here we create a minimal MARC record and hand it off to the
 428                  # template just like a normal result ... perhaps not ideal, but
 429                  # it works for now
 430                         my $tmprecord = MARC::Record->new();
 431                         $tmprecord->encoding('UTF-8');
 432                         my $tmptitle;
 433                         my $tmpauthor;
 434
 435                 # the minimal record in author/title (depending on MARC flavour)
 436                         if (C4::Context->preference("marcflavour") eq "UNIMARC") {
 437                             $tmptitle = MARC::Field->new('200',' ',' ', a => $term, f => $occ);
 438                             $tmprecord->append_fields($tmptitle);
 439                         } else {
 440                             $tmptitle  = MARC::Field->new('245',' ',' ', a => $term,);
 441                             $tmpauthor = MARC::Field->new('100',' ',' ', a => $occ,);
 442                             $tmprecord->append_fields($tmptitle);
 443                             $tmprecord->append_fields($tmpauthor);
 444                         }
 445                         $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc();
 446                     }
 447
 448                     # not an index scan
 449                     else {
 450                         $record = $results[ $i - 1 ]->record($j)->raw();
 451
 452                         # warn "RECORD $j:".$record;
 453                         $results_hash->{'RECORDS'}[$j] = $record;
 454
 455             # Fill the facets while we're looping, but only for the biblioserver
 456                         $facet_record = MARC::Record->new_from_usmarc($record)
 457                           if $servers[ $i - 1 ] =~ /biblioserver/;
 458
 459                     #warn $servers[$i-1]."\n".$record; #.$facet_record->title();
 460                         if ($facet_record) {
 461                             for ( my $k = 0 ; $k <= @$facets ; $k++ ) {
 462                                 ($facets->[$k]) or next;
 463                                 my @fields = map {$facet_record->field($_)} @{$facets->[$k]->{'tags'}} ;
 464                                 for my $field (@fields) {
 465                                     my @subfields = $field->subfields();
 466                                     for my $subfield (@subfields) {
 467                                         my ( $code, $data ) = @$subfield;
 468                                         ($code eq $facets->[$k]->{'subfield'}) or next;
 469                                         $facets_counter->{ $facets->[$k]->{'link_value'} }->{$data}++;
 470                                     }
 471                                 }
 472                                 $facets_info->{ $facets->[$k]->{'link_value'} }->{'label_value'} =
 473                                     $facets->[$k]->{'label_value'};
 474                                 $facets_info->{ $facets->[$k]->{'link_value'} }->{'expanded'} =
 475                                     $facets->[$k]->{'expanded'};
 476                             }
 477                         }
 478                     }
 479                 }
 480                 $results_hashref->{ $servers[ $i - 1 ] } = $results_hash;
 481             }
 482
 483             # warn "connection ", $i-1, ": $size hits";
 484             # warn $results[$i-1]->record(0)->render() if $size > 0;
 485
 486             # BUILD FACETS
 487             if ( $servers[ $i - 1 ] =~ /biblioserver/ ) {
 488                 for my $link_value (
 489                     sort { $facets_counter->{$b} <=> $facets_counter->{$a} }
 490                         keys %$facets_counter )
 491                 {
 492                     my $expandable;
 493                     my $number_of_facets;
 494                     my @this_facets_array;
 495                     for my $one_facet (
 496                         sort {
 497                              $facets_counter->{$link_value}->{$b}
 498                          <=> $facets_counter->{$link_value}->{$a}
 499                         } keys %{ $facets_counter->{$link_value} }
 500                       )
 501                     {
 502                         $number_of_facets++;
 503                         if (   ( $number_of_facets < 6 )
 504                             || ( $expanded_facet eq $link_value )
 505                             || ( $facets_info->{$link_value}->{'expanded'} ) )
 506                         {
 507
 508                       # Sanitize the link value ), ( will cause errors with CCL,
 509                             my $facet_link_value = $one_facet;
 510                             $facet_link_value =~ s/(\(|\))/ /g;
 511
 512                             # fix the length that will display in the label,
 513                             my $facet_label_value = $one_facet;
 514                             $facet_label_value =
 515                               substr( $one_facet, 0, 20 ) . "..."
 516                               unless length($facet_label_value) <= 20;
 517
 518                             # if it's a branch, label by the name, not the code,
 519                             if ( $link_value =~ /branch/ ) {
 520                                                                 if (defined $branches
 521                                                                         && ref($branches) eq "HASH"
 522                                                                         && defined $branches->{$one_facet}
 523                                                                         && ref ($branches->{$one_facet}) eq "HASH")
 524                                                                 {
 525                                         $facet_label_value =
 526                                                 $branches->{$one_facet}->{'branchname'};
 527                                                                 }
 528                                                                 else {
 529                                                                         $facet_label_value = "*";
 530                                                                 }
 531                             }
 532
 533                             # but we're down with the whole label being in the link's title.
 534                             push @this_facets_array, {
 535                                 facet_count       => $facets_counter->{$link_value}->{$one_facet},
 536                                 facet_label_value => $facet_label_value,
 537                                 facet_title_value => $one_facet,
 538                                 facet_link_value  => $facet_link_value,
 539                                 type_link_value   => $link_value,
 540                             };
 541                         }
 542                     }
 543
 544                     # handle expanded option
 545                     unless ( $facets_info->{$link_value}->{'expanded'} ) {
 546                         $expandable = 1
 547                           if ( ( $number_of_facets > 6 )
 548                             && ( $expanded_facet ne $link_value ) );
 549                     }
 550                     push @facets_loop, {
 551                         type_link_value => $link_value,
 552                         type_id         => $link_value . "_id",
 553                         "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1,
 554                         facets     => \@this_facets_array,
 555                         expandable => $expandable,
 556                         expand     => $link_value,
 557                     } unless ( ($facets_info->{$link_value}->{'label_value'} =~ /Libraries/) and (C4::Context->preference('singleBranchMode')) );
 558                 }
 559             }
 560         }
 561     }
 562     return ( undef, $results_hashref, \@facets_loop );
 563 }
 564
 565 sub pazGetRecords {
 566     my (
 567         $koha_query,       $simple_query, $sort_by_ref,    $servers_ref,
 568         $results_per_page, $offset,       $expanded_facet, $branches,
 569         $query_type,       $scan
 570     ) = @_;
 571
 572     my $paz = C4::Search::PazPar2->new(C4::Context->config('pazpar2url'));
 573     $paz->init();
 574     $paz->search($simple_query);
 575     sleep 1;   # FIXME: WHY?
 576
 577     # do results
 578     my $results_hashref = {};
 579     my $stats = XMLin($paz->stat);
 580     my $results = XMLin($paz->show($offset, $results_per_page, 'work-title:1'), forcearray => 1);
 581
 582     # for a grouped search result, the number of hits
 583     # is the number of groups returned; 'bib_hits' will have
 584     # the total number of bibs.
 585     $results_hashref->{'biblioserver'}->{'hits'} = $results->{'merged'}->[0];
 586     $results_hashref->{'biblioserver'}->{'bib_hits'} = $stats->{'hits'};
 587
 588     HIT: foreach my $hit (@{ $results->{'hit'} }) {
 589         my $recid = $hit->{recid}->[0];
 590
 591         my $work_title = $hit->{'md-work-title'}->[0];
 592         my $work_author;
 593         if (exists $hit->{'md-work-author'}) {
 594             $work_author = $hit->{'md-work-author'}->[0];
 595         }
 596         my $group_label = (defined $work_author) ? "$work_title / $work_author" : $work_title;
 597
 598         my $result_group = {};
 599         $result_group->{'group_label'} = $group_label;
 600         $result_group->{'group_merge_key'} = $recid;
 601
 602         my $count = 1;
 603         if (exists $hit->{count}) {
 604             $count = $hit->{count}->[0];
 605         }
 606         $result_group->{'group_count'} = $count;
 607
 608         for (my $i = 0; $i < $count; $i++) {
 609             # FIXME -- may need to worry about diacritics here
 610             my $rec = $paz->record($recid, $i);
 611             push @{ $result_group->{'RECORDS'} }, $rec;
 612         }
 613
 614         push @{ $results_hashref->{'biblioserver'}->{'GROUPS'} }, $result_group;
 615     }
 616
 617     # pass through facets
 618     my $termlist_xml = $paz->termlist('author,subject');
 619     my $terms = XMLin($termlist_xml, forcearray => 1);
 620     my @facets_loop = ();
 621     #die Dumper($results);
 622 #    foreach my $list (sort keys %{ $terms->{'list'} }) {
 623 #        my @facets = ();
 624 #        foreach my $facet (sort @{ $terms->{'list'}->{$list}->{'term'} } ) {
 625 #            push @facets, {
 626 #                facet_label_value => $facet->{'name'}->[0],
 627 #            };
 628 #        }
 629 #        push @facets_loop, ( {
 630 #            type_label => $list,
 631 #            facets => \@facets,
 632 #        } );
 633 #    }
 634
 635     return ( undef, $results_hashref, \@facets_loop );
 636 }
 637
 638 # STOPWORDS
 639 sub _remove_stopwords {
 640     my ( $operand, $index ) = @_;
 641     my @stopwords_removed;
 642
 643     # phrase and exact-qualified indexes shouldn't have stopwords removed
 644     if ( $index !~ m/phr|ext/ ) {
 645
 646 # remove stopwords from operand : parse all stopwords & remove them (case insensitive)
 647 #       we use IsAlpha unicode definition, to deal correctly with diacritics.
 648 #       otherwise, a French word like "leçon" woudl be split into "le" "çon", "le"
 649 #       is a stopword, we'd get "çon" and wouldn't find anything...
 650 #
 651                 foreach ( keys %{ C4::Context->stopwords } ) {
 652                         next if ( $_ =~ /(and|or|not)/ );    # don't remove operators
 653                         if ( my ($matched) = ($operand =~
 654                                 /([^\X\p{isAlnum}]\Q$_\E[^\X\p{isAlnum}]|[^\X\p{isAlnum}]\Q$_\E$|^\Q$_\E[^\X\p{isAlnum}])/gi))
 655                         {
 656                                 $operand =~ s/\Q$matched\E/ /gi;
 657                                 push @stopwords_removed, $_;
 658                         }
 659                 }
 660         }
 661     return ( $operand, \@stopwords_removed );
 662 }
 663
 664 # TRUNCATION
 665 sub _detect_truncation {
 666     my ( $operand, $index ) = @_;
 667     my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated,
 668         @regexpr );
 669     $operand =~ s/^ //g;
 670     my @wordlist = split( /\s/, $operand );
 671     foreach my $word (@wordlist) {
 672         if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) {
 673             push @rightlefttruncated, $word;
 674         }
 675         elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) {
 676             push @lefttruncated, $word;
 677         }
 678         elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) {
 679             push @righttruncated, $word;
 680         }
 681         elsif ( index( $word, "*" ) < 0 ) {
 682             push @nontruncated, $word;
 683         }
 684         else {
 685             push @regexpr, $word;
 686         }
 687     }
 688     return (
 689         \@nontruncated,       \@righttruncated, \@lefttruncated,
 690         \@rightlefttruncated, \@regexpr
 691     );
 692 }
 693
 694 # STEMMING
 695 sub _build_stemmed_operand {
 696     my ($operand,$lang) = @_;
 697     require Lingua::Stem::Snowball ;
 698     my $stemmed_operand;
 699
 700     # If operand contains a digit, it is almost certainly an identifier, and should
 701     # not be stemmed.  This is particularly relevant for ISBNs and ISSNs, which
 702     # can contain the letter "X" - for example, _build_stemmend_operand would reduce
 703     # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant
 704     # results (e.g., "23 x 29 cm." from the 300$c).  Bug 2098.
 705     return $operand if $operand =~ /\d/;
 706
 707 # FIXME: the locale should be set based on the user's language and/or search choice
 708     #warn "$lang";
 709     my $stemmer = Lingua::Stem::Snowball->new( lang => $lang,
 710                                                encoding => "UTF-8" );
 711
 712     my @words = split( / /, $operand );
 713     my @stems = $stemmer->stem(\@words);
 714     for my $stem (@stems) {
 715         $stemmed_operand .= "$stem";
 716         $stemmed_operand .= "?"
 717           unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 );
 718         $stemmed_operand .= " ";
 719     }
 720     warn "STEMMED OPERAND: $stemmed_operand" if $DEBUG;
 721     return $stemmed_operand;
 722 }
 723
 724 # FIELD WEIGHTING
 725 sub _build_weighted_query {
 726
 727 # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works
 728 # pretty well but could work much better if we had a smarter query parser
 729     my ( $operand, $stemmed_operand, $index ) = @_;
 730     my $stemming      = C4::Context->preference("QueryStemming")     || 0;
 731     my $weight_fields = C4::Context->preference("QueryWeightFields") || 0;
 732     my $fuzzy_enabled = C4::Context->preference("QueryFuzzy")        || 0;
 733
 734     my $weighted_query .= "(rk=(";    # Specifies that we're applying rank
 735
 736     # Keyword, or, no index specified
 737     if ( ( $index eq 'kw' ) || ( !$index ) ) {
 738         $weighted_query .=
 739           "Title-cover,ext,r1=\"$operand\"";    # exact title-cover
 740         $weighted_query .= " or ti,ext,r2=\"$operand\"";    # exact title
 741         $weighted_query .= " or ti,phr,r3=\"$operand\"";    # phrase title
 742           #$weighted_query .= " or any,ext,r4=$operand";               # exact any
 743           #$weighted_query .=" or kw,wrdl,r5=\"$operand\"";            # word list any
 744         $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\""
 745           if $fuzzy_enabled;    # add fuzzy, word list
 746         $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\""
 747           if ( $stemming and $stemmed_operand )
 748           ;                     # add stemming, right truncation
 749         $weighted_query .= " or wrdl,r9=\"$operand\"";
 750
 751         # embedded sorting: 0 a-z; 1 z-a
 752         # $weighted_query .= ") or (sort1,aut=1";
 753     }
 754
 755     # Barcode searches should skip this process
 756     elsif ( $index eq 'bc' ) {
 757         $weighted_query .= "bc=\"$operand\"";
 758     }
 759
 760     # Authority-number searches should skip this process
 761     elsif ( $index eq 'an' ) {
 762         $weighted_query .= "an=\"$operand\"";
 763     }
 764
 765     # If the index already has more than one qualifier, wrap the operand
 766     # in quotes and pass it back (assumption is that the user knows what they
 767     # are doing and won't appreciate us mucking up their query
 768     elsif ( $index =~ ',' ) {
 769         $weighted_query .= " $index=\"$operand\"";
 770     }
 771
 772     #TODO: build better cases based on specific search indexes
 773     else {
 774         $weighted_query .= " $index,ext,r1=\"$operand\"";    # exact index
 775           #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)";
 776         $weighted_query .= " or $index,phr,r3=\"$operand\"";    # phrase index
 777         $weighted_query .=
 778           " or $index,rt,wrdl,r3=\"$operand\"";    # word list index
 779     }
 780
 781     $weighted_query .= "))";                       # close rank specification
 782     return $weighted_query;
 783 }
 784
 785 =head2 getIndexes
 786
 787 Return an array with available indexes.
 788
 789 =cut
 790
 791 sub getIndexes{
 792     my @indexes = (
 793                     # biblio indexes
 794                     'ab',
 795                     'Abstract',
 796                     'acqdate',
 797                     'allrecords',
 798                     'an',
 799                     'Any',
 800                     'at',
 801                     'au',
 802                     'aub',
 803                     'aud',
 804                     'audience',
 805                     'auo',
 806                     'aut',
 807                     'Author',
 808                     'Author-in-order ',
 809                     'Author-personal-bibliography',
 810                     'Authority-Number',
 811                     'authtype',
 812                     'bc',
 813                     'biblionumber',
 814                     'bio',
 815                     'biography',
 816                     'callnum',
 817                     'cfn',
 818                     'Chronological-subdivision',
 819                     'cn-bib-source',
 820                     'cn-bib-sort',
 821                     'cn-class',
 822                     'cn-item',
 823                     'cn-prefix',
 824                     'cn-suffix',
 825                     'cpn',
 826                     'Code-institution',
 827                     'Conference-name',
 828                     'Conference-name-heading',
 829                     'Conference-name-see',
 830                     'Conference-name-seealso',
 831                     'Content-type',
 832                     'Control-number',
 833                     'copydate',
 834                     'Corporate-name',
 835                     'Corporate-name-heading',
 836                     'Corporate-name-see',
 837                     'Corporate-name-seealso',
 838                     'ctype',
 839                     'date-entered-on-file',
 840                     'Date-of-acquisition',
 841                     'Date-of-publication',
 842                     'Dewey-classification',
 843                     'extent',
 844                     'fic',
 845                     'fiction',
 846                     'Form-subdivision',
 847                     'format',
 848                     'Geographic-subdivision',
 849                     'he',
 850                     'Heading',
 851                     'Heading-use-main-or-added-entry',
 852                     'Heading-use-series-added-entry ',
 853                     'Heading-use-subject-added-entry',
 854                     'Host-item',
 855                     'id-other',
 856                     'Illustration-code',
 857                     'ISBN',
 858                     'ISSN',
 859                     'itemtype',
 860                     'kw',
 861                     'Koha-Auth-Number',
 862                     'l-format',
 863                     'language',
 864                     'lc-card',
 865                     'LC-card-number',
 866                     'lcn',
 867                     'llength',
 868                     'ln',
 869                     'Local-classification',
 870                     'Local-number',
 871                     'Match-heading',
 872                     'Match-heading-see-from',
 873                     'Material-type',
 874                     'mc-itemtype',
 875                     'mc-rtype',
 876                     'mus',
 877                     'Name-geographic',
 878                     'Name-geographic-heading',
 879                     'Name-geographic-see',
 880                     'Name-geographic-seealso',
 881                     'nb',
 882                     'Note',
 883                     'ns',
 884                     'nt',
 885                     'pb',
 886                     'Personal-name',
 887                     'Personal-name-heading',
 888                     'Personal-name-see',
 889                     'Personal-name-seealso',
 890                     'pl',
 891                     'Place-publication',
 892                     'pn',
 893                     'popularity',
 894                     'pubdate',
 895                     'Publisher',
 896                     'Record-type',
 897                     'rtype',
 898                     'se',
 899                     'See',
 900                     'See-also',
 901                     'sn',
 902                     'Stock-number',
 903                     'su',
 904                     'Subject',
 905                     'Subject-heading-thesaurus',
 906                     'Subject-name-personal',
 907                     'Subject-subdivision',
 908                     'Summary',
 909                     'Suppress',
 910                     'su-geo',
 911                     'su-na',
 912                     'su-to',
 913                     'su-ut',
 914                     'ut',
 915                     'Term-genre-form',
 916                     'Term-genre-form-heading',
 917                     'Term-genre-form-see',
 918                     'Term-genre-form-seealso',
 919                     'ti',
 920                     'Title',
 921                     'Title-cover',
 922                     'Title-series',
 923                     'Title-uniform',
 924                     'Title-uniform-heading',
 925                     'Title-uniform-see',
 926                     'Title-uniform-seealso',
 927                     'totalissues',
 928                     'yr',
 929
 930                     # items indexes
 931                     'acqsource',
 932                     'barcode',
 933                     'bc',
 934                     'branch',
 935                     'ccode',
 936                     'classification-source',
 937                     'cn-sort',
 938                     'coded-location-qualifier',
 939                     'copynumber',
 940                     'damaged',
 941                     'datelastborrowed',
 942                     'datelastseen',
 943                     'holdingbranch',
 944                     'homebranch',
 945                     'issues',
 946                     'item',
 947                     'itemnumber',
 948                     'itype',
 949                     'Local-classification',
 950                     'location',
 951                     'lost',
 952                     'materials-specified',
 953                     'mc-ccode',
 954                     'mc-itype',
 955                     'mc-loc',
 956                     'notforloan',
 957                     'onloan',
 958                     'price',
 959                     'renewals',
 960                     'replacementprice',
 961                     'replacementpricedate',
 962                     'reserves',
 963                     'restricted',
 964                     'stack',
 965                     'uri',
 966                     'withdrawn',
 967
 968                     # subject related
 969                   );
 970
 971     return \@indexes;
 972 }
 973
 974 =head2 buildQuery
 975
 976 ( $error, $query,
 977 $simple_query, $query_cgi,
 978 $query_desc, $limit,
 979 $limit_cgi, $limit_desc,
 980 $stopwords_removed, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang);
 981
 982 Build queries and limits in CCL, CGI, Human,
 983 handle truncation, stemming, field weighting, stopwords, fuzziness, etc.
 984
 985 See verbose embedded documentation.
 986
 987
 988 =cut
 989
 990 sub buildQuery {
 991     my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_;
 992
 993     warn "---------\nEnter buildQuery\n---------" if $DEBUG;
 994
 995     # dereference
 996     my @operators = $operators ? @$operators : ();
 997     my @indexes   = $indexes   ? @$indexes   : ();
 998     my @operands  = $operands  ? @$operands  : ();
 999     my @limits    = $limits    ? @$limits    : ();
1000     my @sort_by   = $sort_by   ? @$sort_by   : ();
1001
1002     my $stemming         = C4::Context->preference("QueryStemming")        || 0;
1003     my $auto_truncation  = C4::Context->preference("QueryAutoTruncate")    || 0;
1004     my $weight_fields    = C4::Context->preference("QueryWeightFields")    || 0;
1005     my $fuzzy_enabled    = C4::Context->preference("QueryFuzzy")           || 0;
1006     my $remove_stopwords = C4::Context->preference("QueryRemoveStopwords") || 0;
1007
1008     # no stemming/weight/fuzzy in NoZebra
1009     if ( C4::Context->preference("NoZebra") ) {
1010         $stemming         = 0;
1011         $weight_fields    = 0;
1012         $fuzzy_enabled    = 0;
1013         $auto_truncation  = 0;
1014     }
1015
1016     my $query        = $operands[0];
1017     my $simple_query = $operands[0];
1018
1019     # initialize the variables we're passing back
1020     my $query_cgi;
1021     my $query_desc;
1022     my $query_type;
1023
1024     my $limit;
1025     my $limit_cgi;
1026     my $limit_desc;
1027
1028     my $stopwords_removed;    # flag to determine if stopwords have been removed
1029
1030     my $cclq;
1031     my $cclindexes = getIndexes();
1032     if( $query !~ /\s*ccl=/ ){
1033         for my $index (@$cclindexes){
1034             if($query =~ /($index)(,?\w)*[:=]/){
1035                 $cclq = 1;
1036             }
1037         }
1038         $query = "ccl=$query" if($cclq);
1039     }
1040
1041 # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps
1042 # DIAGNOSTIC ONLY!!
1043     if ( $query =~ /^ccl=/ ) {
1044         return ( undef, $', $', "q=ccl=$'", $', '', '', '', '', 'ccl' );
1045     }
1046     if ( $query =~ /^cql=/ ) {
1047         return ( undef, $', $', "q=cql=$'", $', '', '', '', '', 'cql' );
1048     }
1049     if ( $query =~ /^pqf=/ ) {
1050         return ( undef, $', $', "q=pqf=$'", $', '', '', '', '', 'pqf' );
1051     }
1052
1053     # pass nested queries directly
1054     # FIXME: need better handling of some of these variables in this case
1055     # Nested queries aren't handled well and this implementation is flawed and causes users to be
1056     # unable to search for anything containing () commenting out, will be rewritten for 3.4.0
1057 #    if ( $query =~ /(\(|\))/ ) {
1058 #        return (
1059 #            undef,              $query, $simple_query, $query_cgi,
1060 #            $query,             $limit, $limit_cgi,    $limit_desc,
1061 #            $stopwords_removed, 'ccl'
1062 #        );
1063 #    }
1064
1065 # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming
1066 # query operands and indexes and add stemming, truncation, field weighting, etc.
1067 # Once we do so, we'll end up with a value in $query, just like if we had an
1068 # incoming $query from the user
1069     else {
1070         $query = ""
1071           ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query
1072         my $previous_operand
1073           ;    # a flag used to keep track if there was a previous query
1074                # if there was, we can apply the current operator
1075                # for every operand
1076         for ( my $i = 0 ; $i <= @operands ; $i++ ) {
1077
1078             # COMBINE OPERANDS, INDEXES AND OPERATORS
1079             if ( $operands[$i] ) {
1080                 $operands[$i]=~s/^\s+//;
1081
1082               # A flag to determine whether or not to add the index to the query
1083                 my $indexes_set;
1084
1085 # If the user is sophisticated enough to specify an index, turn off field weighting, stemming, and stopword handling
1086                 if ( $operands[$i] =~ /(:|=)/ || $scan ) {
1087                     $weight_fields    = 0;
1088                     $stemming         = 0;
1089                     $remove_stopwords = 0;
1090                 }
1091                 my $operand = $operands[$i];
1092                 my $index   = $indexes[$i];
1093
1094                 # Add index-specific attributes
1095                 # Date of Publication
1096                 if ( $index eq 'yr' ) {
1097                     $index .= ",st-numeric";
1098                     $indexes_set++;
1099                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1100                 }
1101
1102                 # Date of Acquisition
1103                 elsif ( $index eq 'acqdate' ) {
1104                     $index .= ",st-date-normalized";
1105                     $indexes_set++;
1106                                         $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = $remove_stopwords = 0;
1107                 }
1108                 # ISBN,ISSN,Standard Number, don't need special treatment
1109                 elsif ( $index eq 'nb' || $index eq 'ns' ) {
1110                     $indexes_set++;
1111                     (
1112                         $stemming,      $auto_truncation,
1113                         $weight_fields, $fuzzy_enabled,
1114                         $remove_stopwords
1115                     ) = ( 0, 0, 0, 0, 0 );
1116
1117                 }
1118
1119                 if(not $index){
1120                     $index = 'kw';
1121                 }
1122
1123                 # Set default structure attribute (word list)
1124                 my $struct_attr = q{};
1125                 unless ( $indexes_set || !$index || $index =~ /(st-|phr|ext|wrdl)/ ) {
1126                     $struct_attr = ",wrdl";
1127                 }
1128
1129                 # Some helpful index variants
1130                 my $index_plus       = $index . $struct_attr . ':';
1131                 my $index_plus_comma = $index . $struct_attr . ',';
1132
1133                 # Remove Stopwords
1134                 if ($remove_stopwords) {
1135                     ( $operand, $stopwords_removed ) =
1136                       _remove_stopwords( $operand, $index );
1137                     warn "OPERAND w/out STOPWORDS: >$operand<" if $DEBUG;
1138                     warn "REMOVED STOPWORDS: @$stopwords_removed"
1139                       if ( $stopwords_removed && $DEBUG );
1140                 }
1141
1142                 if ($auto_truncation){
1143                                         unless ( $index =~ /(st-|phr|ext)/ ) {
1144                                                 #FIXME only valid with LTR scripts
1145                                                 $operand=join(" ",map{
1146                                                                                         (index($_,"*")>0?"$_":"$_*")
1147                                                                                          }split (/\s+/,$operand));
1148                                                 warn $operand if $DEBUG;
1149                                         }
1150                                 }
1151
1152                 # Detect Truncation
1153                 my $truncated_operand;
1154                 my( $nontruncated, $righttruncated, $lefttruncated,
1155                     $rightlefttruncated, $regexpr
1156                 ) = _detect_truncation( $operand, $index );
1157                 warn
1158 "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"
1159                   if $DEBUG;
1160
1161                 # Apply Truncation
1162                 if (
1163                     scalar(@$righttruncated) + scalar(@$lefttruncated) +
1164                     scalar(@$rightlefttruncated) > 0 )
1165                 {
1166
1167                # Don't field weight or add the index to the query, we do it here
1168                     $indexes_set = 1;
1169                     undef $weight_fields;
1170                     my $previous_truncation_operand;
1171                     if (scalar @$nontruncated) {
1172                         $truncated_operand .= "$index_plus @$nontruncated ";
1173                         $previous_truncation_operand = 1;
1174                     }
1175                     if (scalar @$righttruncated) {
1176                         $truncated_operand .= "and " if $previous_truncation_operand;
1177                         $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated ";
1178                         $previous_truncation_operand = 1;
1179                     }
1180                     if (scalar @$lefttruncated) {
1181                         $truncated_operand .= "and " if $previous_truncation_operand;
1182                         $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated ";
1183                         $previous_truncation_operand = 1;
1184                     }
1185                     if (scalar @$rightlefttruncated) {
1186                         $truncated_operand .= "and " if $previous_truncation_operand;
1187                         $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated ";
1188                         $previous_truncation_operand = 1;
1189                     }
1190                 }
1191                 $operand = $truncated_operand if $truncated_operand;
1192                 warn "TRUNCATED OPERAND: >$truncated_operand<" if $DEBUG;
1193
1194                 # Handle Stemming
1195                 my $stemmed_operand;
1196                 $stemmed_operand = _build_stemmed_operand($operand, $lang)
1197                                                                                 if $stemming;
1198
1199                 warn "STEMMED OPERAND: >$stemmed_operand<" if $DEBUG;
1200
1201                 # Handle Field Weighting
1202                 my $weighted_operand;
1203                 if ($weight_fields) {
1204                     $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index );
1205                     $operand = $weighted_operand;
1206                     $indexes_set = 1;
1207                 }
1208
1209                 warn "FIELD WEIGHTED OPERAND: >$weighted_operand<" if $DEBUG;
1210
1211                 # If there's a previous operand, we need to add an operator
1212                 if ($previous_operand) {
1213
1214                     # User-specified operator
1215                     if ( $operators[ $i - 1 ] ) {
1216                         $query     .= " $operators[$i-1] ";
1217                         $query     .= " $index_plus " unless $indexes_set;
1218                         $query     .= " $operand";
1219                         $query_cgi .= "&op=$operators[$i-1]";
1220                         $query_cgi .= "&idx=$index" if $index;
1221                         $query_cgi .= "&q=$operands[$i]" if $operands[$i];
1222                         $query_desc .=
1223                           " $operators[$i-1] $index_plus $operands[$i]";
1224                     }
1225
1226                     # Default operator is and
1227                     else {
1228                         $query      .= " and ";
1229                         $query      .= "$index_plus " unless $indexes_set;
1230                         $query      .= "$operand";
1231                         $query_cgi  .= "&op=and&idx=$index" if $index;
1232                         $query_cgi  .= "&q=$operands[$i]" if $operands[$i];
1233                         $query_desc .= " and $index_plus $operands[$i]";
1234                     }
1235                 }
1236
1237                 # There isn't a pervious operand, don't need an operator
1238                 else {
1239
1240                     # Field-weighted queries already have indexes set
1241                     $query .= " $index_plus " unless $indexes_set;
1242                     $query .= $operand;
1243                     $query_desc .= " $index_plus $operands[$i]";
1244                     $query_cgi  .= "&idx=$index" if $index;
1245                     $query_cgi  .= "&q=$operands[$i]" if $operands[$i];
1246                     $previous_operand = 1;
1247                 }
1248             }    #/if $operands
1249         }    # /for
1250     }
1251     warn "QUERY BEFORE LIMITS: >$query<" if $DEBUG;
1252
1253     # add limits
1254     my $group_OR_limits;
1255     my $availability_limit;
1256     foreach my $this_limit (@limits) {
1257 #        if ( $this_limit =~ /available/ ) {
1258 #
1259 ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0)
1260 ## In English:
1261 ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0
1262 #            $availability_limit .=
1263 #"( ( allrecords,AlwaysMatches='' not onloan,AlwaysMatches='') and (lost,st-numeric=0) )"; #or ( allrecords,AlwaysMatches='' not lost,AlwaysMatches='')) )";
1264 #            $limit_cgi  .= "&limit=available";
1265 #            $limit_desc .= "";
1266 #        }
1267 #
1268         # group_OR_limits, prefixed by mc-
1269         # OR every member of the group
1270 #        elsif ( $this_limit =~ /mc/ ) {
1271         if ( $this_limit =~ /mc/ ) {
1272             $group_OR_limits .= " or " if $group_OR_limits;
1273             $limit_desc      .= " or " if $group_OR_limits;
1274             $group_OR_limits .= "$this_limit";
1275             $limit_cgi       .= "&limit=$this_limit";
1276             $limit_desc      .= " $this_limit";
1277         }
1278
1279         # Regular old limits
1280         else {
1281             $limit .= " and " if $limit || $query;
1282             $limit      .= "$this_limit";
1283             $limit_cgi  .= "&limit=$this_limit";
1284             if ($this_limit =~ /^branch:(.+)/) {
1285                 my $branchcode = $1;
1286                 my $branchname = GetBranchName($branchcode);
1287                 if (defined $branchname) {
1288                     $limit_desc .= " branch:$branchname";
1289                 } else {
1290                     $limit_desc .= " $this_limit";
1291                 }
1292             } else {
1293                 $limit_desc .= " $this_limit";
1294             }
1295         }
1296     }
1297     if ($group_OR_limits) {
1298         $limit .= " and " if ( $query || $limit );
1299         $limit .= "($group_OR_limits)";
1300     }
1301     if ($availability_limit) {
1302         $limit .= " and " if ( $query || $limit );
1303         $limit .= "($availability_limit)";
1304     }
1305
1306     # Normalize the query and limit strings
1307     # This is flawed , means we can't search anything with : in it
1308     # if user wants to do ccl or cql, start the query with that
1309 #    $query =~ s/:/=/g;
1310     $query =~ s/(?<=(ti|au|pb|su|an|kw|mc)):/=/g;
1311     $query =~ s/(?<=rtrn):/=/g;
1312     $limit =~ s/:/=/g;
1313     for ( $query, $query_desc, $limit, $limit_desc ) {
1314         s/  / /g;    # remove extra spaces
1315         s/^ //g;     # remove any beginning spaces
1316         s/ $//g;     # remove any ending spaces
1317         s/==/=/g;    # remove double == from query
1318     }
1319     $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi
1320
1321     for ($query_cgi,$simple_query) {
1322         s/"//g;
1323     }
1324     # append the limit to the query
1325     $query .= " " . $limit;
1326
1327     # Warnings if DEBUG
1328     if ($DEBUG) {
1329         warn "QUERY:" . $query;
1330         warn "QUERY CGI:" . $query_cgi;
1331         warn "QUERY DESC:" . $query_desc;
1332         warn "LIMIT:" . $limit;
1333         warn "LIMIT CGI:" . $limit_cgi;
1334         warn "LIMIT DESC:" . $limit_desc;
1335         warn "---------\nLeave buildQuery\n---------";
1336     }
1337     return (
1338         undef,              $query, $simple_query, $query_cgi,
1339         $query_desc,        $limit, $limit_cgi,    $limit_desc,
1340         $stopwords_removed, $query_type
1341     );
1342 }
1343
1344 =head2 searchResults
1345
1346   my @search_results = searchResults($search_context, $searchdesc, $hits,
1347                                      $results_per_page, $offset, $scan,
1348                                      @marcresults, $hidelostitems);
1349
1350 Format results in a form suitable for passing to the template
1351
1352 =cut
1353
1354 # IMO this subroutine is pretty messy still -- it's responsible for
1355 # building the HTML output for the template
1356 sub searchResults {
1357     my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults, $hidelostitems ) = @_;
1358     my $dbh = C4::Context->dbh;
1359     my @newresults;
1360
1361     $search_context = 'opac' unless $search_context eq 'opac' or $search_context eq 'intranet';
1362
1363     #Build branchnames hash
1364     #find branchname
1365     #get branch information.....
1366     my %branches;
1367     my $bsth =$dbh->prepare("SELECT branchcode,branchname FROM branches"); # FIXME : use C4::Branch::GetBranches
1368     $bsth->execute();
1369     while ( my $bdata = $bsth->fetchrow_hashref ) {
1370         $branches{ $bdata->{'branchcode'} } = $bdata->{'branchname'};
1371     }
1372 # FIXME - We build an authorised values hash here, using the default framework
1373 # though it is possible to have different authvals for different fws.
1374
1375     my $shelflocations =GetKohaAuthorisedValues('items.location','');
1376
1377     # get notforloan authorised value list (see $shelflocations  FIXME)
1378     my $notforloan_authorised_value = GetAuthValCode('items.notforloan','');
1379
1380     #Build itemtype hash
1381     #find itemtype & itemtype image
1382     my %itemtypes;
1383     $bsth =
1384       $dbh->prepare(
1385         "SELECT itemtype,description,imageurl,summary,notforloan FROM itemtypes"
1386       );
1387     $bsth->execute();
1388     while ( my $bdata = $bsth->fetchrow_hashref ) {
1389                 foreach (qw(description imageurl summary notforloan)) {
1390                 $itemtypes{ $bdata->{'itemtype'} }->{$_} = $bdata->{$_};
1391                 }
1392     }
1393
1394     #search item field code
1395     my $sth =
1396       $dbh->prepare(
1397 "SELECT tagfield FROM marc_subfield_structure WHERE kohafield LIKE 'items.itemnumber'"
1398       );
1399     $sth->execute;
1400     my ($itemtag) = $sth->fetchrow;
1401
1402     ## find column names of items related to MARC
1403     my $sth2 = $dbh->prepare("SHOW COLUMNS FROM items");
1404     $sth2->execute;
1405     my %subfieldstosearch;
1406     while ( ( my $column ) = $sth2->fetchrow ) {
1407         my ( $tagfield, $tagsubfield ) =
1408           &GetMarcFromKohaField( "items." . $column, "" );
1409         $subfieldstosearch{$column} = $tagsubfield;
1410     }
1411
1412     # handle which records to actually retrieve
1413     my $times;
1414     if ( $hits && $offset + $results_per_page <= $hits ) {
1415         $times = $offset + $results_per_page;
1416     }
1417     else {
1418         $times = $hits;  # FIXME: if $hits is undefined, why do we want to equal it?
1419     }
1420
1421         my $marcflavour = C4::Context->preference("marcflavour");
1422     # We get the biblionumber position in MARC
1423     my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber','');
1424     my $fw;
1425
1426     # loop through all of the records we've retrieved
1427     for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) {
1428         my $marcrecord = MARC::File::USMARC::decode( $marcresults[$i] );
1429         if ($bibliotag<10){
1430             $fw = GetFrameworkCode($marcrecord->field($bibliotag)->data);
1431         }else{
1432             $fw = GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf));
1433         }
1434
1435         my $oldbiblio = TransformMarcToKoha( $dbh, $marcrecord, $fw );
1436         $oldbiblio->{subtitle} = GetRecordValue('subtitle', $marcrecord, $fw);
1437         $oldbiblio->{result_number} = $i + 1;
1438
1439         # add imageurl to itemtype if there is one
1440         $oldbiblio->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $oldbiblio->{itemtype} }->{imageurl} );
1441
1442         $oldbiblio->{'authorised_value_images'}  = C4::Items::get_authorised_value_images( C4::Biblio::get_biblio_authorised_values( $oldbiblio->{'biblionumber'}, $marcrecord ) );
1443                 $oldbiblio->{normalized_upc}  = GetNormalizedUPC(       $marcrecord,$marcflavour);
1444                 $oldbiblio->{normalized_ean}  = GetNormalizedEAN(       $marcrecord,$marcflavour);
1445                 $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour);
1446                 $oldbiblio->{normalized_isbn} = GetNormalizedISBN(undef,$marcrecord,$marcflavour);
1447                 $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc});
1448
1449                 # edition information, if any
1450         $oldbiblio->{edition} = $oldbiblio->{editionstatement};
1451                 $oldbiblio->{description} = $itemtypes{ $oldbiblio->{itemtype} }->{description};
1452  # Build summary if there is one (the summary is defined in the itemtypes table)
1453  # FIXME: is this used anywhere, I think it can be commented out? -- JF
1454         if ( $itemtypes{ $oldbiblio->{itemtype} }->{summary} ) {
1455             my $summary = $itemtypes{ $oldbiblio->{itemtype} }->{summary};
1456             my @fields  = $marcrecord->fields();
1457
1458             my $newsummary;
1459             foreach my $line ( "$summary\n" =~ /(.*)\n/g ){
1460                 my $tags = {};
1461                 foreach my $tag ( $line =~ /\[(\d{3}[\w|\d])\]/ ) {
1462                     $tag =~ /(.{3})(.)/;
1463                     if($marcrecord->field($1)){
1464                         my @abc = $marcrecord->field($1)->subfield($2);
1465                         $tags->{$tag} = $#abc + 1 ;
1466                     }
1467                 }
1468
1469                 # We catch how many times to repeat this line
1470                 my $max = 0;
1471                 foreach my $tag (keys(%$tags)){
1472                     $max = $tags->{$tag} if($tags->{$tag} > $max);
1473                  }
1474
1475                 # we replace, and repeat each line
1476                 for (my $i = 0 ; $i < $max ; $i++){
1477                     my $newline = $line;
1478
1479                     foreach my $tag ( $newline =~ /\[(\d{3}[\w|\d])\]/g ) {
1480                         $tag =~ /(.{3})(.)/;
1481
1482                         if($marcrecord->field($1)){
1483                             my @repl = $marcrecord->field($1)->subfield($2);
1484                             my $subfieldvalue = $repl[$i];
1485
1486                             if (! utf8::is_utf8($subfieldvalue)) {
1487                                 utf8::decode($subfieldvalue);
1488                             }
1489
1490                              $newline =~ s/\[$tag\]/$subfieldvalue/g;
1491                         }
1492                     }
1493                     $newsummary .= "$newline\n";
1494                 }
1495             }
1496
1497             $newsummary =~ s/\[(.*?)]//g;
1498             $newsummary =~ s/\n/<br\/>/g;
1499             $oldbiblio->{summary} = $newsummary;
1500         }
1501
1502         # Pull out the items fields
1503         my @fields = $marcrecord->field($itemtag);
1504
1505         # Setting item statuses for display
1506         my @available_items_loop;
1507         my @onloan_items_loop;
1508         my @other_items_loop;
1509
1510         my $available_items;
1511         my $onloan_items;
1512         my $other_items;
1513
1514         my $ordered_count         = 0;
1515         my $available_count       = 0;
1516         my $onloan_count          = 0;
1517         my $longoverdue_count     = 0;
1518         my $other_count           = 0;
1519         my $wthdrawn_count        = 0;
1520         my $itemlost_count        = 0;
1521         my $itembinding_count     = 0;
1522         my $itemdamaged_count     = 0;
1523         my $item_in_transit_count = 0;
1524         my $can_place_holds       = 0;
1525         my $item_onhold_count     = 0;
1526         my $items_count           = scalar(@fields);
1527         my $maxitems =
1528           ( C4::Context->preference('maxItemsinSearchResults') )
1529           ? C4::Context->preference('maxItemsinSearchResults') - 1
1530           : 1;
1531
1532         # loop through every item
1533         foreach my $field (@fields) {
1534             my $item;
1535
1536             # populate the items hash
1537             foreach my $code ( keys %subfieldstosearch ) {
1538                 $item->{$code} = $field->subfield( $subfieldstosearch{$code} );
1539             }
1540
1541                         my $hbranch     = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'homebranch'    : 'holdingbranch';
1542                         my $otherbranch = C4::Context->preference('HomeOrHoldingBranch') eq 'homebranch' ? 'holdingbranch' : 'homebranch';
1543             # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one
1544             if ($item->{$hbranch}) {
1545                 $item->{'branchname'} = $branches{$item->{$hbranch}};
1546             }
1547             elsif ($item->{$otherbranch}) {     # Last resort
1548                 $item->{'branchname'} = $branches{$item->{$otherbranch}};
1549             }
1550
1551                         my $prefix = $item->{$hbranch} . '--' . $item->{location} . $item->{itype} . $item->{itemcallnumber};
1552 # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item
1553             if ( $item->{onloan} ) {
1554                 $onloan_count++;
1555                                 my $key = $prefix . $item->{onloan} . $item->{barcode};
1556                                 $onloan_items->{$key}->{due_date} = format_date($item->{onloan});
1557                                 $onloan_items->{$key}->{count}++ if $item->{$hbranch};
1558                                 $onloan_items->{$key}->{branchname} = $item->{branchname};
1559                                 $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1560                                 $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber};
1561                                 $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1562                 # if something's checked out and lost, mark it as 'long overdue'
1563                 if ( $item->{itemlost} ) {
1564                     $onloan_items->{$prefix}->{longoverdue}++;
1565                     $longoverdue_count++;
1566                 } else {        # can place holds as long as item isn't lost
1567                     $can_place_holds = 1;
1568                 }
1569             }
1570
1571          # items not on loan, but still unavailable ( lost, withdrawn, damaged )
1572             else {
1573
1574                 # item is on order
1575                 if ( $item->{notforloan} == -1 ) {
1576                     $ordered_count++;
1577                 }
1578
1579                 # is item in transit?
1580                 my $transfertwhen = '';
1581                 my ($transfertfrom, $transfertto);
1582
1583                 # is item on the reserve shelf?
1584                 my $reservestatus = 0;
1585                 my $reserveitem;
1586
1587                 unless ($item->{wthdrawn}
1588                         || $item->{itemlost}
1589                         || $item->{damaged}
1590                         || $item->{notforloan}
1591                         || $items_count > 20) {
1592
1593                     # A couple heuristics to limit how many times
1594                     # we query the database for item transfer information, sacrificing
1595                     # accuracy in some cases for speed;
1596                     #
1597                     # 1. don't query if item has one of the other statuses
1598                     # 2. don't check transit status if the bib has
1599                     #    more than 20 items
1600                     #
1601                     # FIXME: to avoid having the query the database like this, and to make
1602                     #        the in transit status count as unavailable for search limiting,
1603                     #        should map transit status to record indexed in Zebra.
1604                     #
1605                     ($transfertwhen, $transfertfrom, $transfertto) = C4::Circulation::GetTransfers($item->{itemnumber});
1606                     ($reservestatus, $reserveitem) = C4::Reserves::CheckReserves($item->{itemnumber});
1607                 }
1608
1609                 # item is withdrawn, lost or damaged
1610                 if (   $item->{wthdrawn}
1611                     || $item->{itemlost}
1612                     || $item->{damaged}
1613                     || $item->{notforloan}
1614                     || $reservestatus eq 'Waiting'
1615                     || ($transfertwhen ne ''))
1616                 {
1617                     $wthdrawn_count++        if $item->{wthdrawn};
1618                     $itemlost_count++        if $item->{itemlost};
1619                     $itemdamaged_count++     if $item->{damaged};
1620                     $item_in_transit_count++ if $transfertwhen ne '';
1621                     $item_onhold_count++     if $reservestatus eq 'Waiting';
1622                     $item->{status} = $item->{wthdrawn} . "-" . $item->{itemlost} . "-" . $item->{damaged} . "-" . $item->{notforloan};
1623                     $other_count++;
1624
1625                                         my $key = $prefix . $item->{status};
1626                                         foreach (qw(wthdrawn itemlost damaged branchname itemcallnumber)) {
1627                         $other_items->{$key}->{$_} = $item->{$_};
1628                                         }
1629                     $other_items->{$key}->{intransit} = ($transfertwhen ne '') ? 1 : 0;
1630                     $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0;
1631                                         $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value;
1632                                         $other_items->{$key}->{count}++ if $item->{$hbranch};
1633                                         $other_items->{$key}->{location} = $shelflocations->{ $item->{location} };
1634                                         $other_items->{$key}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1635                 }
1636                 # item is available
1637                 else {
1638                     $can_place_holds = 1;
1639                     $available_count++;
1640                                         $available_items->{$prefix}->{count}++ if $item->{$hbranch};
1641                                         foreach (qw(branchname itemcallnumber)) {
1642                         $available_items->{$prefix}->{$_} = $item->{$_};
1643                                         }
1644                                         $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} };
1645                                         $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( 'opac', $itemtypes{ $item->{itype} }->{imageurl} );
1646                 }
1647             }
1648         }    # notforloan, item level and biblioitem level
1649         my ( $availableitemscount, $onloanitemscount, $otheritemscount );
1650         $maxitems =
1651           ( C4::Context->preference('maxItemsinSearchResults') )
1652           ? C4::Context->preference('maxItemsinSearchResults') - 1
1653           : 1;
1654         for my $key ( sort keys %$onloan_items ) {
1655             (++$onloanitemscount > $maxitems) and last;
1656             push @onloan_items_loop, $onloan_items->{$key};
1657         }
1658         for my $key ( sort keys %$other_items ) {
1659             (++$otheritemscount > $maxitems) and last;
1660             push @other_items_loop, $other_items->{$key};
1661         }
1662         for my $key ( sort keys %$available_items ) {
1663             (++$availableitemscount > $maxitems) and last;
1664             push @available_items_loop, $available_items->{$key}
1665         }
1666
1667         # XSLT processing of some stuff
1668         use C4::Charset;
1669         SetUTF8Flag($marcrecord);
1670         $debug && warn $marcrecord->as_formatted;
1671         if (!$scan && $search_context eq 'opac' && C4::Context->preference("OPACXSLTResultsDisplay")) {
1672             # FIXME note that XSLTResultsDisplay (use of XSLT to format staff interface bib search results)
1673             # is not implemented yet
1674             $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display($oldbiblio->{biblionumber}, $marcrecord, 'Results',
1675                                                                 $search_context);
1676         }
1677
1678         # last check for norequest : if itemtype is notforloan, it can't be reserved either, whatever the items
1679         $can_place_holds = 0
1680           if $itemtypes{ $oldbiblio->{itemtype} }->{notforloan};
1681         $oldbiblio->{norequests} = 1 unless $can_place_holds;
1682         $oldbiblio->{itemsplural}          = 1 if $items_count > 1;
1683         $oldbiblio->{items_count}          = $items_count;
1684         $oldbiblio->{available_items_loop} = \@available_items_loop;
1685         $oldbiblio->{onloan_items_loop}    = \@onloan_items_loop;
1686         $oldbiblio->{other_items_loop}     = \@other_items_loop;
1687         $oldbiblio->{availablecount}       = $available_count;
1688         $oldbiblio->{availableplural}      = 1 if $available_count > 1;
1689         $oldbiblio->{onloancount}          = $onloan_count;
1690         $oldbiblio->{onloanplural}         = 1 if $onloan_count > 1;
1691         $oldbiblio->{othercount}           = $other_count;
1692         $oldbiblio->{otherplural}          = 1 if $other_count > 1;
1693         $oldbiblio->{wthdrawncount}        = $wthdrawn_count;
1694         $oldbiblio->{itemlostcount}        = $itemlost_count;
1695         $oldbiblio->{damagedcount}         = $itemdamaged_count;
1696         $oldbiblio->{intransitcount}       = $item_in_transit_count;
1697         $oldbiblio->{onholdcount}          = $item_onhold_count;
1698         $oldbiblio->{orderedcount}         = $ordered_count;
1699         $oldbiblio->{isbn} =~
1700           s/-//g;    # deleting - in isbn to enable amazon content
1701         push( @newresults, $oldbiblio )
1702             if(not $hidelostitems
1703                or (($items_count > $itemlost_count )
1704                     && $hidelostitems));
1705     }
1706
1707     return @newresults;
1708 }
1709
1710 =head2 SearchAcquisitions
1711     Search for acquisitions
1712 =cut
1713
1714 sub SearchAcquisitions{
1715     my ($datebegin, $dateend, $itemtypes,$criteria, $orderby) = @_;
1716
1717     my $dbh=C4::Context->dbh;
1718     # Variable initialization
1719     my $str=qq|
1720     SELECT marcxml
1721     FROM biblio
1722     LEFT JOIN biblioitems ON biblioitems.biblionumber=biblio.biblionumber
1723     LEFT JOIN items ON items.biblionumber=biblio.biblionumber
1724     WHERE dateaccessioned BETWEEN ? AND ?
1725     |;
1726
1727     my (@params,@loopcriteria);
1728
1729     push @params, $datebegin->output("iso");
1730     push @params, $dateend->output("iso");
1731
1732     if (scalar(@$itemtypes)>0 and $criteria ne "itemtype" ){
1733         if(C4::Context->preference("item-level_itypes")){
1734             $str .= "AND items.itype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1735         }else{
1736             $str .= "AND biblioitems.itemtype IN (?".( ',?' x scalar @$itemtypes - 1 ).") ";
1737         }
1738         push @params, @$itemtypes;
1739     }
1740
1741     if ($criteria =~/itemtype/){
1742         if(C4::Context->preference("item-level_itypes")){
1743             $str .= "AND items.itype=? ";
1744         }else{
1745             $str .= "AND biblioitems.itemtype=? ";
1746         }
1747
1748         if(scalar(@$itemtypes) == 0){
1749             my $itypes = GetItemTypes();
1750             for my $key (keys %$itypes){
1751                 push @$itemtypes, $key;
1752             }
1753         }
1754
1755         @loopcriteria= @$itemtypes;
1756     }elsif ($criteria=~/itemcallnumber/){
1757         $str .= "AND (items.itemcallnumber LIKE CONCAT(?,'%')
1758                  OR items.itemcallnumber is NULL
1759                  OR items.itemcallnumber = '')";
1760
1761         @loopcriteria = ("AA".."ZZ", "") unless (scalar(@loopcriteria)>0);
1762     }else {
1763         $str .= "AND biblio.title LIKE CONCAT(?,'%') ";
1764         @loopcriteria = ("A".."z") unless (scalar(@loopcriteria)>0);
1765     }
1766
1767     if ($orderby =~ /date_desc/){
1768         $str.=" ORDER BY dateaccessioned DESC";
1769     } else {
1770         $str.=" ORDER BY title";
1771     }
1772
1773     my $qdataacquisitions=$dbh->prepare($str);
1774
1775     my @loopacquisitions;
1776     foreach my $value(@loopcriteria){
1777         push @params,$value;
1778         my %cell;
1779         $cell{"title"}=$value;
1780         $cell{"titlecode"}=$value;
1781
1782         eval{$qdataacquisitions->execute(@params);};
1783
1784         if ($@){ warn "recentacquisitions Error :$@";}
1785         else {
1786             my @loopdata;
1787             while (my $data=$qdataacquisitions->fetchrow_hashref){
1788                 push @loopdata, {"summary"=>GetBiblioSummary( $data->{'marcxml'} ) };
1789             }
1790             $cell{"loopdata"}=\@loopdata;
1791         }
1792         push @loopacquisitions,\%cell if (scalar(@{$cell{loopdata}})>0);
1793         pop @params;
1794     }
1795     $qdataacquisitions->finish;
1796     return \@loopacquisitions;
1797 }
1798 #----------------------------------------------------------------------
1799 #
1800 # Non-Zebra GetRecords#
1801 #----------------------------------------------------------------------
1802
1803 =head2 NZgetRecords
1804
1805   NZgetRecords has the same API as zera getRecords, even if some parameters are not managed
1806
1807 =cut
1808
1809 sub NZgetRecords {
1810     my (
1811         $query,            $simple_query, $sort_by_ref,    $servers_ref,
1812         $results_per_page, $offset,       $expanded_facet, $branches,
1813         $query_type,       $scan
1814     ) = @_;
1815     warn "query =$query" if $DEBUG;
1816     my $result = NZanalyse($query);
1817     warn "results =$result" if $DEBUG;
1818     return ( undef,
1819         NZorder( $result, @$sort_by_ref[0], $results_per_page, $offset ),
1820         undef );
1821 }
1822
1823 =head2 NZanalyse
1824
1825   NZanalyse : get a CQL string as parameter, and returns a list of biblionumber;title,biblionumber;title,...
1826   the list is built from an inverted index in the nozebra SQL table
1827   note that title is here only for convenience : the sorting will be very fast when requested on title
1828   if the sorting is requested on something else, we will have to reread all results, and that may be longer.
1829
1830 =cut
1831
1832 sub NZanalyse {
1833     my ( $string, $server ) = @_;
1834 #     warn "---------"       if $DEBUG;
1835     warn " NZanalyse" if $DEBUG;
1836 #     warn "---------"       if $DEBUG;
1837
1838  # $server contains biblioserver or authorities, depending on what we search on.
1839  #warn "querying : $string on $server";
1840     $server = 'biblioserver' unless $server;
1841
1842 # if we have a ", replace the content to discard temporarily any and/or/not inside
1843     my $commacontent;
1844     if ( $string =~ /"/ ) {
1845         $string =~ s/"(.*?)"/__X__/;
1846         $commacontent = $1;
1847         warn "commacontent : $commacontent" if $DEBUG;
1848     }
1849
1850 # split the query string in 3 parts : X AND Y means : $left="X", $operand="AND" and $right="Y"
1851 # then, call again NZanalyse with $left and $right
1852 # (recursive until we find a leaf (=> something without and/or/not)
1853 # delete repeated operator... Would then go in infinite loop
1854     while ( $string =~ s/( and| or| not| AND| OR| NOT)\1/$1/g ) {
1855     }
1856
1857     #process parenthesis before.
1858     if ( $string =~ /^\s*\((.*)\)(( and | or | not | AND | OR | NOT )(.*))?/ ) {
1859         my $left     = $1;
1860         my $right    = $4;
1861         my $operator = lc($3);   # FIXME: and/or/not are operators, not operands
1862         warn
1863 "dealing w/parenthesis before recursive sub call. left :$left operator:$operator right:$right"
1864           if $DEBUG;
1865         my $leftresult = NZanalyse( $left, $server );
1866         if ($operator) {
1867             my $rightresult = NZanalyse( $right, $server );
1868
1869             # OK, we have the results for right and left part of the query
1870             # depending of operand, intersect, union or exclude both lists
1871             # to get a result list
1872             if ( $operator eq ' and ' ) {
1873                 return NZoperatorAND($leftresult,$rightresult);
1874             }
1875             elsif ( $operator eq ' or ' ) {
1876
1877                 # just merge the 2 strings
1878                 return $leftresult . $rightresult;
1879             }
1880             elsif ( $operator eq ' not ' ) {
1881                 return NZoperatorNOT($leftresult,$rightresult);
1882             }
1883         }
1884         else {
1885 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1886             return $leftresult;
1887         }
1888     }
1889     warn "string :" . $string if $DEBUG;
1890     my $left = "";
1891     my $right = "";
1892     my $operator = "";
1893     if ($string =~ /(.*?)( and | or | not | AND | OR | NOT )(.*)/) {
1894         $left     = $1;
1895         $right    = $3;
1896         $operator = lc($2);    # FIXME: and/or/not are operators, not operands
1897     }
1898     warn "no parenthesis. left : $left operator: $operator right: $right"
1899       if $DEBUG;
1900
1901     # it's not a leaf, we have a and/or/not
1902     if ($operator) {
1903
1904         # reintroduce comma content if needed
1905         $right =~ s/__X__/"$commacontent"/ if $commacontent;
1906         $left  =~ s/__X__/"$commacontent"/ if $commacontent;
1907         warn "node : $left / $operator / $right\n" if $DEBUG;
1908         my $leftresult  = NZanalyse( $left,  $server );
1909         my $rightresult = NZanalyse( $right, $server );
1910         warn " leftresult : $leftresult" if $DEBUG;
1911         warn " rightresult : $rightresult" if $DEBUG;
1912         # OK, we have the results for right and left part of the query
1913         # depending of operand, intersect, union or exclude both lists
1914         # to get a result list
1915         if ( $operator eq ' and ' ) {
1916             return NZoperatorAND($leftresult,$rightresult);
1917         }
1918         elsif ( $operator eq ' or ' ) {
1919
1920             # just merge the 2 strings
1921             return $leftresult . $rightresult;
1922         }
1923         elsif ( $operator eq ' not ' ) {
1924             return NZoperatorNOT($leftresult,$rightresult);
1925         }
1926         else {
1927
1928 # this error is impossible, because of the regexp that isolate the operand, but just in case...
1929             die "error : operand unknown : $operator for $string";
1930         }
1931
1932         # it's a leaf, do the real SQL query and return the result
1933     }
1934     else {
1935         $string =~ s/__X__/"$commacontent"/ if $commacontent;
1936         $string =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|&|\+|\*|\// /g;
1937         #remove trailing blank at the beginning
1938         $string =~ s/^ //g;
1939         warn "leaf:$string" if $DEBUG;
1940
1941         # parse the string in in operator/operand/value again
1942         my $left = "";
1943         my $operator = "";
1944         my $right = "";
1945         if ($string =~ /(.*)(>=|<=)(.*)/) {
1946             $left     = $1;
1947             $operator = $2;
1948             $right    = $3;
1949         } else {
1950             $left = $string;
1951         }
1952 #         warn "handling leaf... left:$left operator:$operator right:$right"
1953 #           if $DEBUG;
1954         unless ($operator) {
1955             if ($string =~ /(.*)(>|<|=)(.*)/) {
1956                 $left     = $1;
1957                 $operator = $2;
1958                 $right    = $3;
1959                 warn
1960     "handling unless (operator)... left:$left operator:$operator right:$right"
1961                 if $DEBUG;
1962             } else {
1963                 $left = $string;
1964             }
1965         }
1966         my $results;
1967
1968 # strip adv, zebra keywords, currently not handled in nozebra: wrdl, ext, phr...
1969         $left =~ s/ .*$//;
1970
1971         # automatic replace for short operators
1972         $left = 'title'            if $left =~ '^ti$';
1973         $left = 'author'           if $left =~ '^au$';
1974         $left = 'publisher'        if $left =~ '^pb$';
1975         $left = 'subject'          if $left =~ '^su$';
1976         $left = 'koha-Auth-Number' if $left =~ '^an$';
1977         $left = 'keyword'          if $left =~ '^kw$';
1978         $left = 'itemtype'         if $left =~ '^mc$'; # Fix for Bug 2599 - Search limits not working for NoZebra
1979         warn "handling leaf... left:$left operator:$operator right:$right" if $DEBUG;
1980         my $dbh = C4::Context->dbh;
1981         if ( $operator && $left ne 'keyword' ) {
1982             #do a specific search
1983             $operator = 'LIKE' if $operator eq '=' and $right =~ /%/;
1984             my $sth = $dbh->prepare(
1985 "SELECT biblionumbers,value FROM nozebra WHERE server=? AND indexname=? AND value $operator ?"
1986             );
1987             warn "$left / $operator / $right\n" if $DEBUG;
1988
1989             # split each word, query the DB and build the biblionumbers result
1990             #sanitizing leftpart
1991             $left =~ s/^\s+|\s+$//;
1992             foreach ( split / /, $right ) {
1993                 my $biblionumbers;
1994                 $_ =~ s/^\s+|\s+$//;
1995                 next unless $_;
1996                 warn "EXECUTE : $server, $left, $_" if $DEBUG;
1997                 $sth->execute( $server, $left, $_ )
1998                   or warn "execute failed: $!";
1999                 while ( my ( $line, $value ) = $sth->fetchrow ) {
2000
2001 # if we are dealing with a numeric value, use only numeric results (in case of >=, <=, > or <)
2002 # otherwise, fill the result
2003                     $biblionumbers .= $line
2004                       unless ( $right =~ /^\d+$/ && $value =~ /\D/ );
2005                     warn "result : $value "
2006                       . ( $right  =~ /\d/ ) . "=="
2007                       . ( $value =~ /\D/?$line:"" ) if $DEBUG;         #= $line";
2008                 }
2009
2010 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2011                 if ($results) {
2012                     warn "NZAND" if $DEBUG;
2013                     $results = NZoperatorAND($biblionumbers,$results);
2014                 } else {
2015                     $results = $biblionumbers;
2016                 }
2017             }
2018         }
2019         else {
2020       #do a complete search (all indexes), if index='kw' do complete search too.
2021             my $sth = $dbh->prepare(
2022 "SELECT biblionumbers FROM nozebra WHERE server=? AND value LIKE ?"
2023             );
2024
2025             # split each word, query the DB and build the biblionumbers result
2026             foreach ( split / /, $string ) {
2027                 next if C4::Context->stopwords->{ uc($_) };   # skip if stopword
2028                 warn "search on all indexes on $_" if $DEBUG;
2029                 my $biblionumbers;
2030                 next unless $_;
2031                 $sth->execute( $server, $_ );
2032                 while ( my $line = $sth->fetchrow ) {
2033                     $biblionumbers .= $line;
2034                 }
2035
2036 # do a AND with existing list if there is one, otherwise, use the biblionumbers list as 1st result list
2037                 if ($results) {
2038                     $results = NZoperatorAND($biblionumbers,$results);
2039                 }
2040                 else {
2041                     warn "NEW RES for $_ = $biblionumbers" if $DEBUG;
2042                     $results = $biblionumbers;
2043                 }
2044             }
2045         }
2046         warn "return : $results for LEAF : $string" if $DEBUG;
2047         return $results;
2048     }
2049     warn "---------\nLeave NZanalyse\n---------" if $DEBUG;
2050 }
2051
2052 sub NZoperatorAND{
2053     my ($rightresult, $leftresult)=@_;
2054
2055     my @leftresult = split /;/, $leftresult;
2056     warn " @leftresult / $rightresult \n" if $DEBUG;
2057
2058     #             my @rightresult = split /;/,$leftresult;
2059     my $finalresult;
2060
2061 # parse the left results, and if the biblionumber exist in the right result, save it in finalresult
2062 # the result is stored twice, to have the same weight for AND than OR.
2063 # example : TWO : 61,61,64,121 (two is twice in the biblio #61) / TOWER : 61,64,130
2064 # result : 61,61,61,61,64,64 for two AND tower : 61 has more weight than 64
2065     foreach (@leftresult) {
2066         my $value = $_;
2067         my $countvalue;
2068         ( $value, $countvalue ) = ( $1, $2 ) if ($value=~/(.*)-(\d+)$/);
2069         if ( $rightresult =~ /\Q$value\E-(\d+);/ ) {
2070             $countvalue = ( $1 > $countvalue ? $countvalue : $1 );
2071             $finalresult .=
2072                 "$value-$countvalue;$value-$countvalue;";
2073         }
2074     }
2075     warn "NZAND DONE : $finalresult \n" if $DEBUG;
2076     return $finalresult;
2077 }
2078
2079 sub NZoperatorOR{
2080     my ($rightresult, $leftresult)=@_;
2081     return $rightresult.$leftresult;
2082 }
2083
2084 sub NZoperatorNOT{
2085     my ($leftresult, $rightresult)=@_;
2086
2087     my @leftresult = split /;/, $leftresult;
2088
2089     #             my @rightresult = split /;/,$leftresult;
2090     my $finalresult;
2091     foreach (@leftresult) {
2092         my $value=$_;
2093         $value=$1 if $value=~m/(.*)-\d+$/;
2094         unless ($rightresult =~ "$value-") {
2095             $finalresult .= "$_;";
2096         }
2097     }
2098     return $finalresult;
2099 }
2100
2101 =head2 NZorder
2102
2103   $finalresult = NZorder($biblionumbers, $ordering,$results_per_page,$offset);
2104
2105   TODO :: Description
2106
2107 =cut
2108
2109 sub NZorder {
2110     my ( $biblionumbers, $ordering, $results_per_page, $offset ) = @_;
2111     warn "biblionumbers = $biblionumbers and ordering = $ordering\n" if $DEBUG;
2112
2113     # order title asc by default
2114     #     $ordering = '1=36 <i' unless $ordering;
2115     $results_per_page = 20 unless $results_per_page;
2116     $offset           = 0  unless $offset;
2117     my $dbh = C4::Context->dbh;
2118
2119     #
2120     # order by POPULARITY
2121     #
2122     if ( $ordering =~ /popularity/ ) {
2123         my %result;
2124         my %popularity;
2125
2126         # popularity is not in MARC record, it's builded from a specific query
2127         my $sth =
2128           $dbh->prepare("select sum(issues) from items where biblionumber=?");
2129         foreach ( split /;/, $biblionumbers ) {
2130             my ( $biblionumber, $title ) = split /,/, $_;
2131             $result{$biblionumber} = GetMarcBiblio($biblionumber);
2132             $sth->execute($biblionumber);
2133             my $popularity = $sth->fetchrow || 0;
2134
2135 # hint : the key is popularity.title because we can have
2136 # many results with the same popularity. In this case, sub-ordering is done by title
2137 # we also have biblionumber to avoid bug for 2 biblios with the same title & popularity
2138 # (un-frequent, I agree, but we won't forget anything that way ;-)
2139             $popularity{ sprintf( "%10d", $popularity ) . $title
2140                   . $biblionumber } = $biblionumber;
2141         }
2142
2143     # sort the hash and return the same structure as GetRecords (Zebra querying)
2144         my $result_hash;
2145         my $numbers = 0;
2146         if ( $ordering eq 'popularity_dsc' ) {    # sort popularity DESC
2147             foreach my $key ( sort { $b cmp $a } ( keys %popularity ) ) {
2148                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2149                   $result{ $popularity{$key} }->as_usmarc();
2150             }
2151         }
2152         else {                                    # sort popularity ASC
2153             foreach my $key ( sort ( keys %popularity ) ) {
2154                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2155                   $result{ $popularity{$key} }->as_usmarc();
2156             }
2157         }
2158         my $finalresult = ();
2159         $result_hash->{'hits'}         = $numbers;
2160         $finalresult->{'biblioserver'} = $result_hash;
2161         return $finalresult;
2162
2163         #
2164         # ORDER BY author
2165         #
2166     }
2167     elsif ( $ordering =~ /author/ ) {
2168         my %result;
2169         foreach ( split /;/, $biblionumbers ) {
2170             my ( $biblionumber, $title ) = split /,/, $_;
2171             my $record = GetMarcBiblio($biblionumber);
2172             my $author;
2173             if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2174                 $author = $record->subfield( '200', 'f' );
2175                 $author = $record->subfield( '700', 'a' ) unless $author;
2176             }
2177             else {
2178                 $author = $record->subfield( '100', 'a' );
2179             }
2180
2181 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2182 # and we don't want to get only 1 result for each of them !!!
2183             $result{ $author . $biblionumber } = $record;
2184         }
2185
2186     # sort the hash and return the same structure as GetRecords (Zebra querying)
2187         my $result_hash;
2188         my $numbers = 0;
2189         if ( $ordering eq 'author_za' ) {    # sort by author desc
2190             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2191                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2192                   $result{$key}->as_usmarc();
2193             }
2194         }
2195         else {                               # sort by author ASC
2196             foreach my $key ( sort ( keys %result ) ) {
2197                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2198                   $result{$key}->as_usmarc();
2199             }
2200         }
2201         my $finalresult = ();
2202         $result_hash->{'hits'}         = $numbers;
2203         $finalresult->{'biblioserver'} = $result_hash;
2204         return $finalresult;
2205
2206         #
2207         # ORDER BY callnumber
2208         #
2209     }
2210     elsif ( $ordering =~ /callnumber/ ) {
2211         my %result;
2212         foreach ( split /;/, $biblionumbers ) {
2213             my ( $biblionumber, $title ) = split /,/, $_;
2214             my $record = GetMarcBiblio($biblionumber);
2215             my $callnumber;
2216             my $frameworkcode = GetFrameworkCode($biblionumber);
2217             my ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField(  'items.itemcallnumber', $frameworkcode);
2218                ( $callnumber_tag, $callnumber_subfield ) = GetMarcFromKohaField('biblioitems.callnumber', $frameworkcode)
2219                 unless $callnumber_tag;
2220             if ( C4::Context->preference('marcflavour') eq 'UNIMARC' ) {
2221                 $callnumber = $record->subfield( '200', 'f' );
2222             } else {
2223                 $callnumber = $record->subfield( '100', 'a' );
2224             }
2225
2226 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2227 # and we don't want to get only 1 result for each of them !!!
2228             $result{ $callnumber . $biblionumber } = $record;
2229         }
2230
2231     # sort the hash and return the same structure as GetRecords (Zebra querying)
2232         my $result_hash;
2233         my $numbers = 0;
2234         if ( $ordering eq 'call_number_dsc' ) {    # sort by title desc
2235             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2236                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2237                   $result{$key}->as_usmarc();
2238             }
2239         }
2240         else {                                     # sort by title ASC
2241             foreach my $key ( sort { $a cmp $b } ( keys %result ) ) {
2242                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2243                   $result{$key}->as_usmarc();
2244             }
2245         }
2246         my $finalresult = ();
2247         $result_hash->{'hits'}         = $numbers;
2248         $finalresult->{'biblioserver'} = $result_hash;
2249         return $finalresult;
2250     }
2251     elsif ( $ordering =~ /pubdate/ ) {             #pub year
2252         my %result;
2253         foreach ( split /;/, $biblionumbers ) {
2254             my ( $biblionumber, $title ) = split /,/, $_;
2255             my $record = GetMarcBiblio($biblionumber);
2256             my ( $publicationyear_tag, $publicationyear_subfield ) =
2257               GetMarcFromKohaField( 'biblioitems.publicationyear', '' );
2258             my $publicationyear =
2259               $record->subfield( $publicationyear_tag,
2260                 $publicationyear_subfield );
2261
2262 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2263 # and we don't want to get only 1 result for each of them !!!
2264             $result{ $publicationyear . $biblionumber } = $record;
2265         }
2266
2267     # sort the hash and return the same structure as GetRecords (Zebra querying)
2268         my $result_hash;
2269         my $numbers = 0;
2270         if ( $ordering eq 'pubdate_dsc' ) {    # sort by pubyear desc
2271             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2272                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2273                   $result{$key}->as_usmarc();
2274             }
2275         }
2276         else {                                 # sort by pub year ASC
2277             foreach my $key ( sort ( keys %result ) ) {
2278                 $result_hash->{'RECORDS'}[ $numbers++ ] =
2279                   $result{$key}->as_usmarc();
2280             }
2281         }
2282         my $finalresult = ();
2283         $result_hash->{'hits'}         = $numbers;
2284         $finalresult->{'biblioserver'} = $result_hash;
2285         return $finalresult;
2286
2287         #
2288         # ORDER BY title
2289         #
2290     }
2291     elsif ( $ordering =~ /title/ ) {
2292
2293 # the title is in the biblionumbers string, so we just need to build a hash, sort it and return
2294         my %result;
2295         foreach ( split /;/, $biblionumbers ) {
2296             my ( $biblionumber, $title ) = split /,/, $_;
2297
2298 # hint : the result is sorted by title.biblionumber because we can have X biblios with the same title
2299 # and we don't want to get only 1 result for each of them !!!
2300 # hint & speed improvement : we can order without reading the record
2301 # so order, and read records only for the requested page !
2302             $result{ $title . $biblionumber } = $biblionumber;
2303         }
2304
2305     # sort the hash and return the same structure as GetRecords (Zebra querying)
2306         my $result_hash;
2307         my $numbers = 0;
2308         if ( $ordering eq 'title_az' ) {    # sort by title desc
2309             foreach my $key ( sort ( keys %result ) ) {
2310                 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2311             }
2312         }
2313         else {                              # sort by title ASC
2314             foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2315                 $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2316             }
2317         }
2318
2319         # limit the $results_per_page to result size if it's more
2320         $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2321
2322         # for the requested page, replace biblionumber by the complete record
2323         # speed improvement : avoid reading too much things
2324         for (
2325             my $counter = $offset ;
2326             $counter <= $offset + $results_per_page ;
2327             $counter++
2328           )
2329         {
2330             $result_hash->{'RECORDS'}[$counter] =
2331               GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc;
2332         }
2333         my $finalresult = ();
2334         $result_hash->{'hits'}         = $numbers;
2335         $finalresult->{'biblioserver'} = $result_hash;
2336         return $finalresult;
2337     }
2338     else {
2339
2340 #
2341 # order by ranking
2342 #
2343 # we need 2 hashes to order by ranking : the 1st one to count the ranking, the 2nd to order by ranking
2344         my %result;
2345         my %count_ranking;
2346         foreach ( split /;/, $biblionumbers ) {
2347             my ( $biblionumber, $title ) = split /,/, $_;
2348             $title =~ /(.*)-(\d)/;
2349
2350             # get weight
2351             my $ranking = $2;
2352
2353 # note that we + the ranking because ranking is calculated on weight of EACH term requested.
2354 # if we ask for "two towers", and "two" has weight 2 in biblio N, and "towers" has weight 4 in biblio N
2355 # biblio N has ranking = 6
2356             $count_ranking{$biblionumber} += $ranking;
2357         }
2358
2359 # build the result by "inverting" the count_ranking hash
2360 # 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
2361 #         warn "counting";
2362         foreach ( keys %count_ranking ) {
2363             $result{ sprintf( "%10d", $count_ranking{$_} ) . '-' . $_ } = $_;
2364         }
2365
2366     # sort the hash and return the same structure as GetRecords (Zebra querying)
2367         my $result_hash;
2368         my $numbers = 0;
2369         foreach my $key ( sort { $b cmp $a } ( keys %result ) ) {
2370             $result_hash->{'RECORDS'}[ $numbers++ ] = $result{$key};
2371         }
2372
2373         # limit the $results_per_page to result size if it's more
2374         $results_per_page = $numbers - 1 if $numbers < $results_per_page;
2375
2376         # for the requested page, replace biblionumber by the complete record
2377         # speed improvement : avoid reading too much things
2378         for (
2379             my $counter = $offset ;
2380             $counter <= $offset + $results_per_page ;
2381             $counter++
2382           )
2383         {
2384             $result_hash->{'RECORDS'}[$counter] =
2385               GetMarcBiblio( $result_hash->{'RECORDS'}[$counter] )->as_usmarc
2386               if $result_hash->{'RECORDS'}[$counter];
2387         }
2388         my $finalresult = ();
2389         $result_hash->{'hits'}         = $numbers;
2390         $finalresult->{'biblioserver'} = $result_hash;
2391         return $finalresult;
2392     }
2393 }
2394
2395 =head2 enabled_staff_search_views
2396
2397 %hash = enabled_staff_search_views()
2398
2399 This function returns a hash that contains three flags obtained from the system
2400 preferences, used to determine whether a particular staff search results view
2401 is enabled.
2402
2403 =over 2
2404
2405 =item C<Output arg:>
2406
2407     * $hash{can_view_MARC} is true only if the MARC view is enabled
2408     * $hash{can_view_ISBD} is true only if the ISBD view is enabled
2409     * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled
2410
2411 =item C<usage in the script:>
2412
2413 =back
2414
2415 $template->param ( C4::Search::enabled_staff_search_views );
2416
2417 =cut
2418
2419 sub enabled_staff_search_views
2420 {
2421         return (
2422                 can_view_MARC                   => C4::Context->preference('viewMARC'),                 # 1 if the staff search allows the MARC view
2423                 can_view_ISBD                   => C4::Context->preference('viewISBD'),                 # 1 if the staff search allows the ISBD view
2424                 can_view_labeledMARC    => C4::Context->preference('viewLabeledMARC'),  # 1 if the staff search allows the Labeled MARC view
2425         );
2426 }
2427
2428 sub AddSearchHistory{
2429         my ($borrowernumber,$session,$query_desc,$query_cgi, $total)=@_;
2430     my $dbh = C4::Context->dbh;
2431
2432     # Add the request the user just made
2433     my $sql = "INSERT INTO search_history(userid, sessionid, query_desc, query_cgi, total, time) VALUES(?, ?, ?, ?, ?, NOW())";
2434     my $sth   = $dbh->prepare($sql);
2435     $sth->execute($borrowernumber, $session, $query_desc, $query_cgi, $total);
2436         return $dbh->last_insert_id(undef, 'search_history', undef,undef,undef);
2437 }
2438
2439 sub GetSearchHistory{
2440         my ($borrowernumber,$session)=@_;
2441     my $dbh = C4::Context->dbh;
2442
2443     # Add the request the user just made
2444     my $query = "SELECT FROM search_history WHERE (userid=? OR sessionid=?)";
2445     my $sth   = $dbh->prepare($query);
2446         $sth->execute($borrowernumber, $session);
2447     return  $sth->fetchall_hashref({});
2448 }
2449
2450 =head2 z3950_search_args
2451
2452 $arrayref = z3950_search_args($matchpoints)
2453
2454 This function returns an array reference that contains the search parameters to be
2455 passed to the Z39.50 search script (z3950_search.pl). The array elements
2456 are hash refs whose keys are name, value and encvalue, and whose values are the
2457 name of a search parameter, the value of that search parameter and the URL encoded
2458 value of that parameter.
2459
2460 The search parameter names are lccn, isbn, issn, title, author, dewey and subject.
2461
2462 The search parameter values are obtained from the bibliographic record whose
2463 data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData().
2464
2465 If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g.
2466 a general purpose search argument. In this case, the returned array contains only
2467 entry: the key is 'title' and the value and encvalue are derived from $matchpoints.
2468
2469 If a search parameter value is undefined or empty, it is not included in the returned
2470 array.
2471
2472 The returned array reference may be passed directly to the template parameters.
2473
2474 =over 2
2475
2476 =item C<Output arg:>
2477
2478     * $array containing hash refs as described above
2479
2480 =item C<usage in the script:>
2481
2482 =back
2483
2484 $data = Biblio::GetBiblioData($bibno);
2485 $template->param ( MYLOOP => C4::Search::z3950_search_args($data) )
2486
2487 *OR*
2488
2489 $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) )
2490
2491 =cut
2492
2493 sub z3950_search_args {
2494     my $bibrec = shift;
2495     $bibrec = { title => $bibrec } if !ref $bibrec;
2496     my $array = [];
2497     for my $field (qw/ lccn isbn issn title author dewey subject /)
2498     {
2499         my $encvalue = URI::Escape::uri_escape_utf8($bibrec->{$field});
2500         push @$array, { name=>$field, value=>$bibrec->{$field}, encvalue=>$encvalue } if defined $bibrec->{$field};
2501     }
2502     return $array;
2503 }
2504
2505 =head2 BiblioAddAuthorities
2506
2507 ( $countlinked, $countcreated ) = BiblioAddAuthorities($record, $frameworkcode);
2508
2509 this function finds the authorities linked to the biblio
2510     * search in the authority DB for the same authid (in $9 of the biblio)
2511     * search in the authority DB for the same 001 (in $3 of the biblio in UNIMARC)
2512     * search in the authority DB for the same values (exactly) (in all subfields of the biblio)
2513 OR adds a new authority record
2514
2515 =over 2
2516
2517 =item C<input arg:>
2518
2519     * $record is the MARC record in question (marc blob)
2520     * $frameworkcode is the bibliographic framework to use (if it is "" it uses the default framework)
2521
2522 =item C<Output arg:>
2523
2524     * $countlinked is the number of authorities records that are linked to this authority
2525     * $countcreated
2526
2527 =item C<BUGS>
2528     * I had to add this to Search.pm (instead of the logical Biblio.pm) because of a circular dependency (this sub uses SimpleSearch, and Search.pm uses Biblio.pm)
2529
2530 =back
2531
2532 =cut
2533
2534
2535 sub BiblioAddAuthorities{
2536   my ( $record, $frameworkcode ) = @_;
2537   my $dbh=C4::Context->dbh;
2538   my $query=$dbh->prepare(qq|
2539 SELECT authtypecode,tagfield
2540 FROM marc_subfield_structure
2541 WHERE frameworkcode=?
2542 AND (authtypecode IS NOT NULL AND authtypecode<>\"\")|);
2543 # SELECT authtypecode,tagfield
2544 # FROM marc_subfield_structure
2545 # WHERE frameworkcode=?
2546 # AND (authtypecode IS NOT NULL OR authtypecode<>\"\")|);
2547   $query->execute($frameworkcode);
2548   my ($countcreated,$countlinked);
2549   while (my $data=$query->fetchrow_hashref){
2550     foreach my $field ($record->field($data->{tagfield})){
2551       next if ($field->subfield('3')||$field->subfield('9'));
2552       # No authorities id in the tag.
2553       # Search if there is any authorities to link to.
2554       my $query='at='.$data->{authtypecode}.' ';
2555       map {$query.= ' and he,ext="'.$_->[1].'"' if ($_->[0]=~/[A-z]/)}  $field->subfields();
2556       my ($error, $results, $total_hits)=SimpleSearch( $query, undef, undef, [ "authorityserver" ] );
2557     # there is only 1 result
2558           if ( $error ) {
2559         warn "BIBLIOADDSAUTHORITIES: $error";
2560             return (0,0) ;
2561           }
2562       if ($results && scalar(@$results)==1) {
2563         my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2564         $field->add_subfields('9'=>$marcrecord->field('001')->data);
2565         $countlinked++;
2566       } elsif (scalar(@$results)>1) {
2567    #More than One result
2568    #This can comes out of a lack of a subfield.
2569 #         my $marcrecord = MARC::File::USMARC::decode($results->[0]);
2570 #         $record->field($data->{tagfield})->add_subfields('9'=>$marcrecord->field('001')->data);
2571   $countlinked++;
2572       } else {
2573   #There are no results, build authority record, add it to Authorities, get authid and add it to 9
2574   ###NOTICE : This is only valid if a subfield is linked to one and only one authtypecode
2575   ###NOTICE : This can be a problem. We should also look into other types and rejected forms.
2576          my $authtypedata=C4::AuthoritiesMarc->GetAuthType($data->{authtypecode});
2577          next unless $authtypedata;
2578          my $marcrecordauth=MARC::Record->new();
2579          my $authfield=MARC::Field->new($authtypedata->{auth_tag_to_report},'','',"a"=>"".$field->subfield('a'));
2580          map { $authfield->add_subfields($_->[0]=>$_->[1]) if ($_->[0]=~/[A-z]/ && $_->[0] ne "a" )}  $field->subfields();
2581          $marcrecordauth->insert_fields_ordered($authfield);
2582
2583          # bug 2317: ensure new authority knows it's using UTF-8; currently
2584          # only need to do this for MARC21, as MARC::Record->as_xml_record() handles
2585          # automatically for UNIMARC (by not transcoding)
2586          # FIXME: AddAuthority() instead should simply explicitly require that the MARC::Record
2587          # use UTF-8, but as of 2008-08-05, did not want to introduce that kind
2588          # of change to a core API just before the 3.0 release.
2589          if (C4::Context->preference('marcflavour') eq 'MARC21') {
2590             SetMarcUnicodeFlag($marcrecordauth, 'MARC21');
2591          }
2592
2593 #          warn "AUTH RECORD ADDED : ".$marcrecordauth->as_formatted;
2594
2595          my $authid=AddAuthority($marcrecordauth,'',$data->{authtypecode});
2596          $countcreated++;
2597          $field->add_subfields('9'=>$authid);
2598       }
2599     }
2600   }
2601   return ($countlinked,$countcreated);
2602 }
2603
2604 =head2 GetDistinctValues($field);
2605
2606 C<$field> is a reference to the fields array
2607
2608 =cut
2609
2610 sub GetDistinctValues {
2611     my ($fieldname,$string)=@_;
2612     # returns a reference to a hash of references to branches...
2613     if ($fieldname=~/\./){
2614                         my ($table,$column)=split /\./, $fieldname;
2615                         my $dbh = C4::Context->dbh;
2616                         warn "select DISTINCT($column) as value, count(*) as cnt from $table group by lib order by $column " if $DEBUG;
2617                         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 ");
2618                         $sth->execute;
2619                         my $elements=$sth->fetchall_arrayref({});
2620                         return $elements;
2621    }
2622    else {
2623                 $string||= qq("");
2624                 my @servers=qw<biblioserver authorityserver>;
2625                 my (@zconns,@results);
2626         for ( my $i = 0 ; $i < @servers ; $i++ ) {
2627                 $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 );
2628                         $results[$i] =
2629                       $zconns[$i]->scan(
2630                         ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i])
2631                       );
2632                 }
2633                 # The big moment: asynchronously retrieve results from all servers
2634                 my @elements;
2635                 while ( ( my $i = ZOOM::event( \@zconns ) ) != 0 ) {
2636                         my $ev = $zconns[ $i - 1 ]->last_event();
2637                         if ( $ev == ZOOM::Event::ZEND ) {
2638                                 next unless $results[ $i - 1 ];
2639                                 my $size = $results[ $i - 1 ]->size();
2640                                 if ( $size > 0 ) {
2641                       for (my $j=0;$j<$size;$j++){
2642                                                 my %hashscan;
2643                                                 @hashscan{qw(value cnt)}=$results[ $i - 1 ]->display_term($j);
2644                                                 push @elements, \%hashscan;
2645                                           }
2646                                 }
2647                         }
2648                 }
2649                 return \@elements;
2650    }
2651 }
2652
2653
2654 END { }    # module clean-up code here (global destructor)
2655
2656 1;
2657 __END__
2658
2659 =head1 AUTHOR
2660
2661 Koha Development Team <http://koha-community.org/>
2662
2663 =cut