misc/migration_tools/rebuild_nozebra.pl

   1 #!/usr/bin/perl
   2
   3 use C4::Context;
   4 use Getopt::Long;
   5 use C4::Biblio;
   6 use C4::AuthoritiesMarc;
   7
   8 use strict;
   9 #
  10 # script that fills the nozebra table
  11 #
  12 #
  13
  14 $|=1; # flushes output
  15
  16 # limit for database dumping
  17 my $limit;# = "LIMIT 100";
  18 my $directory;
  19 my $skip_export;
  20 my $keep_export;
  21 my $reset;
  22 my $biblios;
  23 my $authorities;
  24 my $sysprefs;
  25 my $commit;
  26
  27 GetOptions(
  28     'd:s'      => \$directory,
  29     'reset'      => \$reset,
  30     's'        => \$skip_export,
  31     'k'        => \$keep_export,
  32     'b'        => \$biblios,
  33     'a'        => \$authorities,
  34     's'        => \$sysprefs,  # rebuild 'NoZebraIndexes' syspref
  35     'commit:f'    => \$commit,
  36     );
  37
  38 my $commitnum = 1000;
  39 $commitnum = $commit if ($commit) ;
  40
  41 $directory = "export" unless $directory;
  42 my $dbh=C4::Context->dbh;
  43 $dbh->do("update systempreferences set value=1 where variable='NoZebra'");
  44
  45 $dbh->do("truncate nozebra");
  46
  47 my %index = GetNoZebraIndexes();
  48
  49 if  (!%index || $sysprefs ) {
  50     if (C4::Context->preference('marcflavour') eq 'UNIMARC') {
  51         $dbh->do("UPDATE systempreferences SET value=\"'title' => '200a,200c,200d,200e,225a,225d,225e,225f,225h,225i,225v,500*,501*,503*,510*,512*,513*,514*,515*,516*,517*,518*,519*,520*,530*,531*,532*,540*,541*,545*,604t,610t,605a',
  52         'author' =>'200f,600a,601a,604a,700a,700b,700c,700d,700a,701b,701c,701d,702a,702b,702c,702d,710a,710b,710c,710d,711a,711b,711c,711d,712a,712b,712c,712d',
  53         'isbn' => '010a',
  54         'issn' => '011a',
  55         'biblionumber' =>'0909',
  56         'itemtype' => '200b',
  57         'language' => '101a',
  58         'publisher' => '210c',
  59         'date' => '210d',
  60         'note' => '300a,301a,302a,303a,304a,305a,306az,307a,308a,309a,310a,311a,312a,313a,314a,315a,316a,317a,318a,319a,320a,321a,322a,323a,324a,325a,326a,327a,328a,330a,332a,333a,336a,337a,345a',
  61         'Koha-Auth-Number' => '6009,6019,6029,6039,6049,6059,6069,6109,7009,7019,7029,7109,7119,7129',
  62         'subject' => '600*,601*,606*,610*',
  63         'dewey' => '676a',
  64         'host-item' => '995a,995c',\" where variable='NoZebraIndexes'");
  65         %index = GetNoZebraIndexes();
  66     } elsif (C4::Context->preference('marcflavour') eq 'MARC21') {
  67         $dbh->do("UPDATE systempreferences SET value=\"
  68 'title' => '130a,210a,222a,240a,243a,245a,245b,246a,246b,247a,247b,250a,250b,440a,830a',
  69 'author' => '100a,100b,100c,100d,110a,111a,111b,111c,111d,245c,700a,710a,711a,800a,810a,811a',
  70 'isbn' => '020a',
  71 'issn' => '022a',
  72 'lccn' => '010a',
  73 'biblionumber => '999c',
  74 'itemtype' => '942c',
  75 'publisher' => '260b',
  76 'date' => '260c',
  77 'note' => '500a, 501a,504a,505a,508a,511a,518a,520a,521a,522a,524a,526a,530a,533a,538a,541a,546a,555a,556a,562a,563a,583a,585a,582a',
  78 'subject' => '600*,610*,611*,630*,650*,651*,653*,654*,655*,662*,690*',
  79 'dewey' => '082',
  80 'bc' => '952p',
  81 'callnum' => '952o',
  82 'an' => '6009,6109,6119',
  83 'series' => 440*,490*,
  84 'host-item' => '952a,952c',
  85 \"WHERE variable='NoZebraIndexes'");
  86
  87         %index = GetNoZebraIndexes();
  88     }
  89 }
  90 $|=1;
  91
  92 $dbh->{AutoCommit} = 0;
  93
  94 print "***********************************\n";
  95 print "***** building BIBLIO indexes *****\n";
  96 print "***********************************\n";
  97
  98 my $sth;
  99 $sth=$dbh->prepare("select biblionumber from biblioitems order by biblionumber $limit");
 100 $sth->execute();
 101 my $i=0;
 102 my %result;
 103 while (my ($biblionumber) = $sth->fetchrow) {
 104         $i++;
 105         print "\r$i";
 106         my  $record;
 107     eval{
 108             $record = GetMarcBiblio($biblionumber);
 109     };
 110     if($@){
 111             print "  There was some pb getting biblionumber : ".$biblionumber."\n";
 112             next;
 113     }
 114     next unless $record;
 115     # get title of the record (to store the 10 first letters with the index)
 116     my ($titletag,$titlesubfield) = GetMarcFromKohaField('biblio.title');
 117     my $title = lc($record->subfield($titletag,$titlesubfield));
 118
 119     # remove blancks comma (that could cause problem when decoding the string for CQL retrieval) and regexp specific values
 120     $title =~ s/ |\.|,|;|\[|\]|\(|\)|\*|-|'|=|://g;
 121     # limit to 10 char, should be enough, and limit the DB size
 122     $title = substr($title,0,10);
 123     #parse each field
 124     foreach my $field ($record->fields()) {
 125         #parse each subfield
 126         next if $field->tag <10;
 127         foreach my $subfield ($field->subfields()) {
 128             my $tag = $field->tag();
 129             my $subfieldcode = $subfield->[0];
 130             my $indexed=0;
 131             # check each index to see if the subfield is stored somewhere
 132             # otherwise, store it in __RAW__ index
 133             foreach my $key (keys %index) {
 134                 if ($index{$key} =~ /$tag\*/ or $index{$key} =~ /$tag$subfieldcode/) {
 135                     $indexed=1;
 136                     my $line= lc $subfield->[1];
 137                     # remove meaningless value in the field...
 138                     $line =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|<|>|&|\+|\*|\/|=|:/ /g;
 139                     # ... and split in words
 140                     foreach (split / /,$line) {
 141                         next unless $_; # skip  empty values (multiple spaces)
 142                         # remove any accented char
 143                         # if the entry is already here, improve weight
 144                         if ($result{$key}->{"$_"} =~ /$biblionumber,\Q$title\E\-(\d);/) {
 145                             my $weight=$1+1;
 146                             $result{$key}->{"$_"} =~ s/$biblionumber,\Q$title\E\-(\d);//;
 147                             $result{$key}->{"$_"} .= "$biblionumber,$title-$weight;";
 148                         # otherwise, create it, with weight=1
 149                         } else {
 150                             $result{$key}->{"$_"}.="$biblionumber,$title-1;";
 151                         }
 152                     }
 153                 }
 154             }
 155             # the subfield is not indexed, store it in __RAW__ index anyway
 156             unless ($indexed) {
 157                 my $line= lc $subfield->[1];
 158                 $line =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|<|>|&|\+|\*|\/|=/ /g;
 159                 foreach (split / /,$line) {
 160                         next unless $_;
 161 #                     warn $record->as_formatted."$_ =>".$title;
 162                         if ($result{__RAW__}->{"$_"} =~ /$biblionumber,\Q$title\E\-(\d);/) {
 163                             my $weight=$1+1;
 164 #                             $weight++;
 165                             $result{__RAW__}->{"$_"} =~ s/$biblionumber,\Q$title\E\-(\d);//;
 166                             $result{__RAW__}->{"$_"} .= "$biblionumber,$title-$weight;";
 167                         } else {
 168                             $result{__RAW__}->{"$_"}.="$biblionumber,$title-1;";
 169                         }
 170                 }
 171             }
 172         }
 173     }
 174    $dbh->commit() if (0 == $i % $commitnum);
 175 }
 176 $dbh->commit;
 177
 178
 179 print "\nInserting records...\n";
 180 $i=0;
 181
 182 my $commitnum = 100;
 183 $dbh->{AutoCommit} = 0;
 184
 185 $sth = $dbh->prepare("INSERT INTO nozebra (server,indexname,value,biblionumbers) VALUES ('biblioserver',?,?,?)");
 186 foreach my $key (keys %result) {
 187     foreach my $index (keys %{$result{$key}}) {
 188         if (length($result{$key}->{$index}) > 1000000) {
 189             print "very long index (".length($result{$key}->{$index}).")for $key / $index. update mySQL config file if you have an error just after this warning (max_paquet_size parameter)\n";
 190         }
 191         print "\r$i";
 192         $i++;
 193         $sth->execute($key,$index,$result{$key}->{$index});
 194         $dbh->commit() if (0 == $i % $commitnum);
 195     }
 196    $dbh->commit() if (0 == $i % $commitnum);
 197 }
 198 $dbh->commit;
 199
 200
 201 print "\nbiblios done\n";
 202
 203 print "\n***********************************\n";
 204 print "***** building AUTHORITIES indexes *****\n";
 205 print "***********************************\n";
 206
 207 $sth=$dbh->prepare("select authid from auth_header order by authid $limit");
 208 $sth->execute();
 209 $i=0;
 210 %result = ();
 211 while (my ($authid) = $sth->fetchrow) {
 212     $i++;
 213     print "\r$i";
 214     my $record;
 215     eval{
 216         $record = GetAuthority($authid);
 217     };
 218     if($@){
 219         print "  There was some pb getting authnumber : ".$authid."\n";
 220         next;
 221     }
 222
 223     my %index;
 224     # for authorities, the "title" is the $a mainentry
 225     my $authref = C4::AuthoritiesMarc::GetAuthType(C4::AuthoritiesMarc::GetAuthTypeCode($authid));
 226
 227     warn "ERROR : authtype undefined for ".$record->as_formatted unless $authref;
 228     my $title = $record->subfield($authref->{auth_tag_to_report},'a');
 229     $index{'mainmainentry'}= $authref->{'auth_tag_to_report'}.'a';
 230     $index{'mainentry'}    = $authref->{'auth_tag_to_report'}.'*';
 231     $index{'auth_type'}    = '152b';
 232
 233     # remove blancks comma (that could cause problem when decoding the string for CQL retrieval) and regexp specific values
 234     $title =~ s/ |\.|,|;|\[|\]|\(|\)|\*|-|'|:|=//g;
 235     $title = quotemeta $title;
 236     # limit to 10 char, should be enough, and limit the DB size
 237     $title = substr($title,0,10);
 238     #parse each field
 239     foreach my $field ($record->fields()) {
 240         #parse each subfield
 241         next if $field->tag <10;
 242         foreach my $subfield ($field->subfields()) {
 243             my $tag = $field->tag();
 244             my $subfieldcode = $subfield->[0];
 245             my $indexed=0;
 246             # check each index to see if the subfield is stored somewhere
 247             # otherwise, store it in __RAW__ index
 248             foreach my $key (keys %index) {
 249                 if ($index{$key} =~ /$tag\*/ or $index{$key} =~ /$tag$subfieldcode/) {
 250                     $indexed=1;
 251                     my $line= lc $subfield->[1];
 252                     # remove meaningless value in the field...
 253                     $line =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|<|>|&|\+|\*|\/|=|:/ /g;
 254                     # ... and split in words
 255                     foreach (split / /,$line) {
 256                         next unless $_; # skip  empty values (multiple spaces)
 257                         # if the entry is already here, improve weight
 258                         if ($result{$key}->{"$_"} =~ /$authid,$title\-(\d);/) {
 259                             my $weight=$1+1;
 260                             $result{$key}->{"$_"} =~ s/$authid,$title\-(\d);//;
 261                             $result{$key}->{"$_"} .= "$authid,$title-$weight;";
 262                         # otherwise, create it, with weight=1
 263                         } else {
 264                             $result{$key}->{"$_"}.="$authid,$title-1;";
 265                         }
 266                     }
 267                 }
 268             }
 269             # the subfield is not indexed, store it in __RAW__ index anyway
 270             unless ($indexed) {
 271                 my $line= lc $subfield->[1];
 272                 $line =~ s/-|\.|\?|,|;|!|'|\(|\)|\[|\]|{|}|"|<|>|&|\+|\*|\/|=/ /g;
 273                 foreach (split / /,$line) {
 274                         next unless $_;
 275 #                     warn $record->as_formatted."$_ =>".$title;
 276                         if ($result{__RAW__}->{"$_"} =~ /$authid,$title\-(\d);/) {
 277                             my $weight=$1+1;
 278 #                             $weight++;
 279                             $result{__RAW__}->{"$_"} =~ s/$authid,$title\-(\d);//;
 280                             $result{__RAW__}->{"$_"} .= "$authid,$title-$weight;";
 281                         } else {
 282                             $result{__RAW__}->{"$_"}.="$authid,$title-1;";
 283                         }
 284                 }
 285             }
 286         }
 287     }
 288    $dbh->commit() if (0 == $i % $commitnum);
 289 }
 290 $dbh->commit;
 291
 292
 293
 294 print "\nInserting...\n";
 295 $i=0;
 296
 297 my $commitnum = 100;
 298 $dbh->{AutoCommit} = 0;
 299 $sth = $dbh->prepare("INSERT INTO nozebra (server,indexname,value,biblionumbers) VALUES ('authorityserver',?,?,?)");
 300 foreach my $key (keys %result) {
 301     foreach my $index (keys %{$result{$key}}) {
 302         if (length($result{$key}->{$index}) > 1000000) {
 303             print "very long index (".length($result{$key}->{$index}).")for $key / $index. update mySQL config file if you have an error just after this warning (max_paquet_size parameter)\n";
 304         }
 305         print "\r$i";
 306         $i++;
 307         $sth->execute($key,$index,$result{$key}->{$index});
 308         $dbh->commit() if (0 == $i % $commitnum);
 309     }
 310    $dbh->commit() if (0 == $i % $commitnum);
 311 }
 312 $dbh->commit;
 313 print "\nauthorities done\n";