misc/cronjobs/cloud-kw.pl

   1 #!/usr/bin/perl
   2
   3 #
   4 # Copyright 2008 Tamil s.a.r.l.
   5 #
   6 # This file is part of Koha.
   7 #
   8 # Koha is free software; you can redistribute it and/or modify it
   9 # under the terms of the GNU General Public License as published by
  10 # the Free Software Foundation; either version 3 of the License, or
  11 # (at your option) any later version.
  12 #
  13 # Koha is distributed in the hope that it will be useful, but
  14 # WITHOUT ANY WARRANTY; without even the implied warranty of
  15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  16 # GNU General Public License for more details.
  17 #
  18 # You should have received a copy of the GNU General Public License
  19 # along with Koha; if not, see <http://www.gnu.org/licenses>.
  20
  21 use strict;
  22 use warnings;
  23 use diagnostics;
  24 use Carp;
  25 use YAML::Syck;
  26 use Pod::Usage;
  27 use Getopt::Long;
  28
  29 use Koha::Script -cron;
  30 use C4::Context;
  31 use C4::Log;
  32
  33 my $verbose     = 0;
  34 my $help        = 0;
  35 my $conf        = '';
  36 GetOptions(
  37     'verbose'   => \$verbose,
  38     'help'      => \$help,
  39     'conf=s'    => \$conf,
  40 );
  41
  42 sub usage {
  43     pod2usage( -verbose => 2 );
  44     exit;
  45 }
  46
  47 usage() if $help || !$conf;
  48
  49 cronlogaction();
  50
  51 my @clouds;
  52 print "Reading configuration file: $conf\n" if $verbose;
  53 eval {
  54     @clouds = LoadFile( $conf );
  55 };
  56 croak "Unable to read configuration file: $conf\n" if $@;
  57
  58 for my $cloud ( @clouds ) {
  59     print "Create a cloud\n",
  60           "  Koha conf file:  ", $cloud->{KohaConf} ? $cloud->{KohaConf} : "default", "\n",
  61           "  Zebra Index:     ", $cloud->{ZebraIndex}, "\n",
  62           "  Koha Keyword:    ", $cloud->{KohaIndex}, "\n",
  63           "  Count:           ", $cloud->{Count}, "\n",
  64           "  Withcss:         ", $cloud->{Withcss}, "\n",
  65           "  Output:          ", $cloud->{Output}, "\n",
  66       if $verbose;
  67
  68     # Set Koha context if KohaConf is present
  69     my $set_new_context = 0;
  70     if ( $cloud->{KohaConf} ) {
  71         if ( -e $cloud->{KohaConf} ) {
  72             my $context = C4::Context->new( $cloud->{KohaConf} );
  73             $context->set_context();
  74             $set_new_context = 1;
  75         }
  76         else {
  77             carp "Koha conf file doesn't exist: ", $cloud->{KohaConf}, " ; use KOHA_CONF\n";
  78         }
  79     }
  80
  81     my $index = new ZebraIndex( $cloud->{ZebraIndex} );
  82     $index->scan( $cloud->{Count} );
  83
  84     open my $fh, ">", $cloud->{Output}
  85         or croak "Unable to create file ", $cloud->{Output};
  86
  87     my $withcss = $cloud->{Withcss} =~ /^yes/i;
  88     print $fh $index->html_cloud( $cloud->{KohaIndex}, $withcss );
  89     close $fh;
  90     $set_new_context && restore_context C4::Context;
  91 }
  92
  93
  94
  95 package ZebraIndex;
  96
  97 use strict;
  98 use warnings;
  99 use diagnostics;
 100 use Carp;
 101
 102 sub new {
 103     my $self = {};
 104     my $class = shift;
 105     $self->{ zebra_index  } = shift;
 106     $self->{ top_terms    } = undef;
 107     $self->{ levels_cloud } = 24;
 108     bless $self, $class;
 109
 110     # Test Zebra index
 111     my $zbiblio = C4::Context->Zconn( "biblioserver" );
 112     eval {
 113         my $ss = $zbiblio->scan_pqf(
 114             '@attr 1=' . $self->{ zebra_index } . ' @attr 4=1 @attr 6=3 "a"'
 115         );
 116     };
 117     croak "Invalid Zebra index: ", $self->{ zebra_index } if $@;
 118
 119     return $self;
 120 }
 121
 122
 123 #
 124 # scan
 125 #   Scan zebra index and populate an array of top terms
 126 #
 127 # PARAMETERS:
 128 #   $max_terms    Max number of top terms
 129 #
 130 # RETURN:
 131 #   A 4-dimensionnal array in $self->{top_terms}
 132 #   [0] term
 133 #   [1] term number of occurrences
 134 #   [2] term proportional relative weight in terms set E[0-1]
 135 #   [3] term logarithmic relative weight E [0-levels_cloud]
 136 #
 137 #   This array is sorted alphabetically by terms ([0])
 138 #   It can be easily sorted by occurrences:
 139 #     @t = sort { $a[1] <=> $a[1] } @{$self->{top_terms}};
 140 #
 141 sub scan {
 142     my $self       = shift;
 143     my $index_name = $self->{ zebra_index };
 144     my $max_terms  = shift;
 145
 146     my $MAX_OCCURENCE = 1000000000;
 147
 148     my $zbiblio = C4::Context->Zconn( "biblioserver" );
 149     my $number_of_terms = 0;
 150     my @terms;      # 2 dimensions array
 151     my $min_occurence_index = -1;
 152     my $min_occurence;
 153     my $from = '0';
 154
 155     while (1) {
 156         my $ss;
 157         eval {
 158             print "$from\n" if $verbose;
 159             $from =~ s/\"/\\\"/g;
 160             my $query = '@attr 1=' . $index_name . ' @attr 4=1 @attr 6=3 "'
 161                         . $from . 'a"';
 162             $ss = $zbiblio->scan_pqf( $query );
 163         };
 164         if ($@) {
 165             chop $from;
 166             next;
 167         }
 168         $ss->option( rpnCharset => 'UTF-8' );
 169         last if $ss->size() == 0;
 170         my $term = '';
 171         my $occ = 0;
 172         for my $index ( 0..$ss->size()-1 ) {
 173             ($term, $occ) = $ss->display_term($index);
 174             #print "$term:$occ\n";
 175             if ( $number_of_terms < $max_terms ) {
 176                 push( @terms, [ $term, $occ ] );
 177                 ++$number_of_terms;
 178                 if ( $number_of_terms == $max_terms ) {
 179                     $min_occurence = $MAX_OCCURENCE;
 180                     for (0..$number_of_terms-1) {
 181                         my @term = @{ $terms[$_] };
 182                         if ( $term[1] <= $min_occurence ) {
 183                             $min_occurence       = $term[1];
 184                             $min_occurence_index = $_;
 185                         }
 186                     }
 187                 }
 188             }
 189             else {
 190                 if ( $occ > $min_occurence) {
 191                     @{ $terms[$min_occurence_index] }[0] = $term;
 192                     @{ $terms[$min_occurence_index] }[1] = $occ;
 193                     $min_occurence = $MAX_OCCURENCE;
 194                     for (0..$max_terms-1) {
 195                         my @term = @{ $terms[$_] };
 196                         if ( $term[1] <= $min_occurence ) {
 197                             $min_occurence       = $term[1];
 198                             $min_occurence_index = $_;
 199                         }
 200                     }
 201                 }
 202             }
 203         }
 204         $from = $term;
 205     }
 206
 207     # Sort array of array by terms weight
 208     @terms = sort { @{$a}[1] <=> @{$b}[1] } @terms;
 209
 210     # A relatif weight to other set terms is added to each term
 211     my $min     = $terms[0][1];
 212     my $log_min = log( $min );
 213     my $max     = $terms[$#terms][1];
 214     my $log_max = log( $max );
 215     my $delta   = $max - $min;
 216     $delta = 1 if $delta == 0; # Very unlikely
 217     my $factor;
 218     if ($log_max - $log_min == 0) {
 219         $log_min = $log_min - $self->{levels_cloud};
 220         $factor = 1;
 221     }
 222     else {
 223         $factor = $self->{levels_cloud} / ($log_max - $log_min);
 224     }
 225
 226     foreach (0..$#terms) {
 227         my $count = @{ $terms[$_] }[1];
 228         my $weight = ( $count - $min ) / $delta;
 229         my $log_weight = int( (log($count) - $log_min) * $factor);
 230         push( @{ $terms[$_] }, $weight );
 231         push( @{ $terms[$_] }, $log_weight );
 232     }
 233     $self->{ top_terms } = \@terms;
 234
 235     # Sort array of array by terms alphabetical order
 236     @terms = sort { @{$a}[0] cmp @{$b}[0] } @terms;
 237 }
 238
 239
 240 #
 241 # Returns a HTML version of index top terms formatted
 242 # as a 'tag cloud'.
 243 #
 244 sub html_cloud {
 245     my $self = shift;
 246     my $koha_index = shift;
 247     my $withcss = shift;
 248     my @terms = @{ $self->{top_terms} };
 249     my $html = '';
 250     if ( $withcss ) {
 251         $html = <<EOS;
 252 <style>
 253 .subjectcloud {
 254     text-align:  center;
 255     line-height: 16px;
 256     margin: 20px;
 257     background: #f0f0f0;
 258     padding: 3%;
 259 }
 260 .subjectcloud a {
 261     font-weight: lighter;
 262     text-decoration: none;
 263 }
 264 span.tagcloud0  { font-size: 12px;}
 265 span.tagcloud1  { font-size: 13px;}
 266 span.tagcloud2  { font-size: 14px;}
 267 span.tagcloud3  { font-size: 15px;}
 268 span.tagcloud4  { font-size: 16px;}
 269 span.tagcloud5  { font-size: 17px;}
 270 span.tagcloud6  { font-size: 18px;}
 271 span.tagcloud7  { font-size: 19px;}
 272 span.tagcloud8  { font-size: 20px;}
 273 span.tagcloud9  { font-size: 21px;}
 274 span.tagcloud10 { font-size: 22px;}
 275 span.tagcloud11 { font-size: 23px;}
 276 span.tagcloud12 { font-size: 24px;}
 277 span.tagcloud13 { font-size: 25px;}
 278 span.tagcloud14 { font-size: 26px;}
 279 span.tagcloud15 { font-size: 27px;}
 280 span.tagcloud16 { font-size: 28px;}
 281 span.tagcloud17 { font-size: 29px;}
 282 span.tagcloud18 { font-size: 30px;}
 283 span.tagcloud19 { font-size: 31px;}
 284 span.tagcloud20 { font-size: 32px;}
 285 span.tagcloud21 { font-size: 33px;}
 286 span.tagcloud22 { font-size: 34px;}
 287 span.tagcloud23 { font-size: 35px;}
 288 span.tagcloud24 { font-size: 36px;}
 289 </style>
 290 <div class="subjectcloud">
 291 EOS
 292     }
 293     for (0..$#terms) {
 294         my @term = @{ $terms[$_] };
 295         my $uri = $term[0];
 296         $uri =~ s/\(//g;
 297         #print "  0=", $term[0]," - 1=", $term[1], " - 2=", $term[2], " - 3=", $term[3],"\n";
 298         $html = $html
 299             . '<span class="tagcloud'
 300             . $term[3]
 301             . '">'
 302             . '<a href="/cgi-bin/koha/opac-search.pl?q='
 303             . $koha_index
 304             . '%3A'
 305             . $uri
 306             . '">'
 307             . $term[0]
 308             . "</a></span>\n";
 309     }
 310     $html .= "</div>\n";
 311     return $html;
 312 }
 313
 314
 315 =head1 NAME
 316
 317 cloud-kw.pl - Creates HTML keywords clouds from Koha Zebra Indexes
 318
 319 =head1 USAGE
 320
 321 =over
 322
 323 =item cloud-kw.pl [--verbose|--help] --conf=F<cloud.conf>
 324
 325 Creates multiple HTML files containing kewords cloud with top terms sorted
 326 by their logarithmic weight.
 327 F<cloud.conf> is a YAML configuration file driving cloud generation
 328 process.
 329
 330 =back
 331
 332 =head1 PARAMETERS
 333
 334 =over
 335
 336 =item B<--conf=configuration file>
 337
 338 Specify configuration file name
 339
 340 =item B<--verbose|-v>
 341
 342 Enable script verbose mode.
 343
 344 =item B<--help|-h>
 345
 346 Print this help page.
 347
 348 =back
 349
 350 =head1 CONFIGURATION
 351
 352 Configuration file looks like that:
 353
 354  ---
 355   # Koha configuration file for a specific installation
 356   # If not present, defaults to KOHA_CONF
 357   KohaConf: /home/koha/mylibray/etc/koha-conf.xml
 358   # Zebra index to scan
 359   ZebraIndex: Author
 360   # Koha index used to link found kewords with an opac search URL
 361   KohaIndex: au
 362   # Number of top keyword to use for the cloud
 363   Count: 50
 364   # Include CSS style directives with the cloud
 365   # This could be used as a model and then CSS directives are
 366   # put in the appropriate CSS file directly.
 367   Withcss: Yes
 368   # HTML file where to output the cloud
 369   Output: /home/koha/mylibrary/koharoot/koha-tmpl/cloud-author.html
 370  ---
 371   KohaConf: /home/koha/yourlibray/etc/koha-conf.xml
 372   ZebraIndex: Subject
 373   KohaIndex: su
 374   Count: 200
 375   Withcss: no
 376   Output: /home/koha/yourlibrary/koharoot/koha-tmpl/cloud-subject.html
 377
 378 =head1 IMPROVEMENTS
 379
 380 Generated top terms have more informations than those outputted from
 381 the time being. Some parameters could be easily added to improve
 382 this script:
 383
 384 =over
 385
 386 =item B<WithCount>
 387
 388 In order to output terms with the number of occurrences they
 389 have been found in Koha Catalogue by Zebra.
 390
 391 =item B<CloudLevels>
 392
 393 Number of levels in the cloud. Now 24 levels are hardcoded.
 394
 395 =item B<Weithing>
 396
 397 Weighting method used to distribute terms in the cloud. We could have two
 398 values: Logarithmic and Linear. Now it's Logarithmic by default.
 399
 400 =item B<Order>
 401
 402 Now terms are outputted in the lexical order. They could be sorted
 403 by their weight.
 404
 405 =back
 406
 407 =cut
 408
 409