Return to updateRS.pl CVS log

Up to [Local Repository] / db / prgsrc

Adding Roma 7's files

    1: #!/usr/local/bin/perl -w
    2: 
    3: =head1 NAME
    4: 
    5: updateRS.pl - Скрипт для занесения в таблицы русского поиска базы 
    6: B<$base> информации о вопросах 
    7: 
    8: =head1 SYNOPSIS
    9: 
   10: updateRS.pl QuestionNumber
   11: 
   12: updateRS.pl 
   13: 
   14: 
   15: =head1 DESCRIPTION
   16: 
   17: Скрипт ищет в таблице Questions вопросы с нулевым ProcessedBySearch,
   18: добавляет информацию в таблицы word2question, nests, nf. Поле 
   19: ProcessedBySearch устанавливается в 1. Обрабатывает QuestionNumber
   20: вопросов. Если параметр QuestionNumber не указан, работает пока не 
   21: обработает все вопросы.
   22: 
   23: 
   24: 
   25: =head1 AUTHOR
   26: 
   27: Роман Семизаров
   28: 
   29: 
   30: =cut
   31: 
   32: 
   33: use locale;
   34: use DBI;
   35: use POSIX qw (locale_h);
   36: use chgkfiles;
   37: use dbchgk;
   38: 
   39: 
   40: do "common.pl";
   41: do "chgk.cnf";
   42:                require "check.pl";
   43: 
   44: open (STDERR,">$stderr") if $stderr;
   45: 
   46: 
   47: open (UNKNOWN,">$unknown");
   48: 
   49: $qlimit=shift||500000;
   50: 
   51:           
   52: my $nf;
   53: 
   54: #open WARN, ">$warnings";
   55: 
   56: %forbidden=checktable('equalto')? getequalto : ();
   57: 
   58: 
   59: if ((uc 'а') ne 'А') {die "!Koi8-r locale not installed!\n"};
   60: 
   61: getquestions(QuestionId, Question, Answer, Comments, Authors, Sources,"ProcessedBySearch IS NULL");
   62: 
   63: 
   64: 
   65: print "Loading dictionaries\n";
   66: 
   67: die "No dictionaries! Check your chgk.cnf" unless scalar @dictionaries;
   68: 
   69: foreach $d(@dictionaries)
   70: {
   71:      print "Loading $d\n";
   72:      open (DICT, $d) || print "              Not found\n";
   73:      while ( <DICT> )
   74:      {
   75:      	chomp;
   76:      	s/\s*$//;
   77:        	($aa,$b)=split(/\//);
   78:     	$a= uc $aa;
   79:     	$words{$a}.=$b || "!";
   80:      }
   81:      close(DICT);
   82: }
   83: 
   84: die "No dictionaries found! Check your chgk.cnf" unless scalar keys %words;
   85: 
   86: 
   87: print "Getting words...\n";
   88: 
   89: 
   90: $sch=0;
   91: while ((++$sch<=$qlimit) && (($id, @ss) = getrow, $id)) 
   92: {
   93:    if ($forbidden{$id}) {next};
   94:    print "\n$id ";
   95:    foreach $fieldnumber (0..$#ss) #перебираем поля
   96:    {
   97:       $text=$ss[$fieldnumber];
   98:       next unless $text;
   99:       $text=~tr/ёЁ/еЕ/;
  100:       $text=~s/(${RLrl})p(${RLrl})/$1p$2/gom;
  101:       $text=~s/p(${RLrl})/р$1/gom;
  102:       $text=~s/(${RLrl})p/$1р/gom;
  103:       $text=~s/\s+/ /gmo;
  104:       @list= $text=~m/(?:(?:${RLrl})+)|(?:[A-Za-z0-9]+)/gom;
  105: 
  106:       foreach $wordnumber(0..$#list)
  107:       {
  108:            $word=$list[$wordnumber];
  109: 
  110:            if (@n=knownword(uc $word)) 
  111:            {
  112:                 incnf($_) foreach @n;
  113:                 updateword2question($_,packword($fieldnumber, $id,$wordnumber),1)
  114:                      foreach (@n);
  115: print ".";
  116:            }
  117:            else {
  118:              if ($word=~/^${RLrl}+$/o){ # Русское слово
  119:             # проанализировать по таблице аффиксов, 
  120:             # проверить наличие начальных форм в 
  121:             # nf, а если таких нет, то 
  122:             # и по словарю. 
  123:               
  124:                $nf=&checkit(uc $word,\%words);
  125: print "!";
  126:                if (!$nf) {
  127:                            $nf=(uc $word)."/!";
  128:                            print UNKNOWN "$nf\n" if $unknown;
  129:                          }
  130: 
  131:               }  else {# нерусское слово
  132:                          $nf=(uc $word)."/!";
  133:                          print UNKNOWN "$nf\n" if $unknown;
  134:                       }
  135: 
  136:                foreach $n (split ' ', $nf)
  137:                {
  138:                    ($f,$flag)=split '/', $n;
  139:                    if ($nfnumber=knownnf($f))
  140:                    {
  141:                        addnest(uc $word,$nfnumber);
  142:                        incnf($nfnumber);
  143:                        updateword2question($nfnumber,packword($fieldnumber,
  144:                           $id,$wordnumber),1)           
  145:                    }
  146:                    else 
  147:                    {
  148:                       $nfnumber=addnf(0, $f, $flag,1);
  149:                       addnest(uc $word,$nfnumber);
  150:                       updateword2question($nfnumber,packword($fieldnumber,
  151:                           $id,$wordnumber),0)           
  152:                    }
  153:                }
  154:            }
  155: 
  156:       }
  157:    }
  158:    searchmark($id);
  159: }
  160: 
  161: 
  162: