/[webpac2]/trunk/run.pl
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Diff of /trunk/run.pl

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 833 by dpavlin, Wed May 23 20:04:13 2007 UTC revision 866 by dpavlin, Thu Jun 21 12:53:41 2007 UTC
# Line 68  and subfield usage for each input Line 68  and subfield usage for each input
68    
69  turn on extra validation of imput records, see L<WebPAC::Validation>  turn on extra validation of imput records, see L<WebPAC::Validation>
70    
71    You can use special variables C<$database> and $C<$input> in this parametar
72    like C<--validate 'conf/validate/$database-$input'> to construct filename
73    
74    =item --validate-delimiters path/to/validate_delimiters_file
75    
76    this option is used with C<--validate> to turn on extra validation of
77    delimiters. If file is non existant, it will be created on first run.
78    
79  =item --marc-generate  =item --marc-generate
80    
81  Generate MARC file. This will automatically be on if file contains C<marc*> directives.  Generate MARC file. This will automatically be on if file contains C<marc*> directives.
# Line 108  my $debug = 0; Line 116  my $debug = 0;
116  my $only_filter;  my $only_filter;
117  my $stats = 0;  my $stats = 0;
118  my $validate_path;  my $validate_path;
119    my $validate_delimiters_path;
120  my $marc_generate = 1;  my $marc_generate = 1;
121  my $marc_lint = 1;  my $marc_lint = 1;
122  my $marc_dump = 0;  my $marc_dump = 0;
# Line 128  GetOptions( Line 137  GetOptions(
137          "debug+" => \$debug,          "debug+" => \$debug,
138          "stats" => \$stats,          "stats" => \$stats,
139          "validate=s" => \$validate_path,          "validate=s" => \$validate_path,
140            "validate-delimiters=s" => \$validate_delimiters_path,
141          "marc-generate!" => \$marc_generate,          "marc-generate!" => \$marc_generate,
142          "marc-lint!" => \$marc_lint,          "marc-lint!" => \$marc_lint,
143          "marc-dump!" => \$marc_dump,          "marc-dump!" => \$marc_dump,
# Line 137  GetOptions( Line 147  GetOptions(
147          "help" => \$help,          "help" => \$help,
148  );  );
149    
150    $marc_generate = 0 if ( $validate_delimiters_path );
151    
152  pod2usage(-verbose => 2) if ($help);  pod2usage(-verbose => 2) if ($help);
153    
154  my $config = new WebPAC::Config( path => $config_path );  my $config = new WebPAC::Config( path => $config_path );
# Line 163  if ($merge) { Line 175  if ($merge) {
175          $log->info("created merge batch file $estcmd_path");          $log->info("created merge batch file $estcmd_path");
176  }  }
177    
   
178  my $validate;  my $validate;
179  $validate = new WebPAC::Validate(  $validate = new WebPAC::Validate(
         path => $validate_path,  
180          delimiters => $config->webpac('delimiters'),          delimiters => $config->webpac('delimiters'),
181  ) if ($validate_path);  ) if ($validate_path || $validate_delimiters_path);
   
182    
183  my $use_indexer = $config->use_indexer;  my $use_indexer = $config->use_indexer;
184  $stats ||= $validate;  $stats ||= $validate;
185  if ($stats) {  if ($stats) {
186          $log->debug("disabled indexing for stats collection");          $log->debug("disabled indexing for stats collection");
187          $use_indexer = undef;          $use_indexer = undef;
188  } else {  } elsif ( $use_indexer ) {
189          $log->info("using $use_indexer indexing engine...");          $log->info("using $use_indexer indexing engine...");
190  }  }
191    
# Line 203  sub create_ds_config { Line 212  sub create_ds_config {
212          return $c;          return $c;
213  }  }
214    
215  while (my ($database, $db_config) = each %{ $config->databases }) {  foreach my $database ( sort keys %{ $config->databases } ) {
216            my $db_config = $config->databases->{$database};
217    
218          my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter);          my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter);
219          next if ($only_database && $database !~ m/$only_database/i);          next if ($only_database && $database !~ m/$only_database/i);
# Line 234  while (my ($database, $db_config) = each Line 244  while (my ($database, $db_config) = each
244                  if ($use_indexer eq 'hyperestraier') {                  if ($use_indexer eq 'hyperestraier') {
245    
246                          # open Hyper Estraier database                          # open Hyper Estraier database
247                          use WebPAC::Output::Estraier '0.10';                          require WebPAC::Output::Estraier;
248                          $indexer = new WebPAC::Output::Estraier( %{ $indexer_config } );                          $indexer = new WebPAC::Output::Estraier( %{ $indexer_config } );
249                                    
250                  } elsif ($use_indexer eq 'hyperestraier-native') {                  } elsif ($use_indexer eq 'hyperestraier-native') {
251    
252                          # open Hyper Estraier database                          # open Hyper Estraier database
253                          use WebPAC::Output::EstraierNative;                          require WebPAC::Output::EstraierNative;
254                          $indexer = new WebPAC::Output::EstraierNative( %{ $indexer_config } );                          $indexer = new WebPAC::Output::EstraierNative( %{ $indexer_config } );
255    
256                  } elsif ($use_indexer eq 'kinosearch') {                  } elsif ($use_indexer eq 'kinosearch') {
257    
258                          # open KinoSearch                          # open KinoSearch
259                          use WebPAC::Output::KinoSearch;                          require WebPAC::Output::KinoSearch;
260                          $indexer_config->{clean} = 1 unless (-e $indexer_config->{index_path});                          $indexer_config->{clean} = 1 unless (-e $indexer_config->{index_path});
261                          $indexer = new WebPAC::Output::KinoSearch( %{ $indexer_config } );                          $indexer = new WebPAC::Output::KinoSearch( %{ $indexer_config } );
262    
# Line 254  while (my ($database, $db_config) = each Line 264  while (my ($database, $db_config) = each
264                          $log->logdie("unknown use_indexer: $use_indexer");                          $log->logdie("unknown use_indexer: $use_indexer");
265                  }                  }
266    
267                  $log->logide("can't continue without valid indexer") unless ($indexer);                  $log->logdie("can't continue without valid indexer") unless ($indexer);
268          }          }
269    
270    
# Line 436  while (my ($database, $db_config) = each Line 446  while (my ($database, $db_config) = each
446                          print $report_fh "Report for database '$database' input '$input_name' records ",                          print $report_fh "Report for database '$database' input '$input_name' records ",
447                                  $offset || 1, "-", $limit || $input->{limit} || $maxmfn, "\n\n";                                  $offset || 1, "-", $limit || $input->{limit} || $maxmfn, "\n\n";
448                          $log->info("Generating report file $path");                          $log->info("Generating report file $path");
449    
450                            if ( $validate ) {
451                                    $validate->read_validate_file( $validate->fill_in( $validate_path, database => $database, input => $input_name ) ) if ( $validate_path );
452                                    $validate->read_validate_delimiters_file( $validate->fill_in( $validate_delimiters_path, database => $database, input => $input_name ) ) if ( $validate_delimiters_path );
453                            }
454                  }                  }
455    
456                  my $marc;                  my $marc;
# Line 562  while (my ($database, $db_config) = each Line 577  while (my ($database, $db_config) = each
577                                  print $report_fh "$errors\n" if ($report_fh);                                  print $report_fh "$errors\n" if ($report_fh);
578                          }                          }
579    
580                          print $report_fh "\nAll possible subfields/delimiter templates:\n", $validate->delimiters_templates, "\n\n";                          print $report_fh "\nAll possible subfields/delimiter templates:\n", $validate->delimiters_templates( report => 1, current_input => 1 ), "\n\n";
581    
582                            # must be last thing that touches $validate for this input
583                            $validate->reset;
584                  }                  }
585    
586                  if ($stats) {                  if ($stats) {
# Line 575  while (my ($database, $db_config) = each Line 593  while (my ($database, $db_config) = each
593                  $marc->finish if ($marc);                  $marc->finish if ($marc);
594    
595                  # close report                  # close report
596                  close($report_fh) if ($report_fh)                  close($report_fh) if ($report_fh);
   
597          }          }
598    
599          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));
# Line 603  if ($parallel) { Line 620  if ($parallel) {
620          $log->info("all parallel processes finished");          $log->info("all parallel processes finished");
621  }  }
622    
623    # save new delimiters if needed
624    $validate->save_delimiters_templates if ( $validate_delimiters_path );
625    
626  #  #
627  # handle links or merge after indexing  # handle links or merge after indexing
628  #  #
Legend:
Removed from v.833  
changed lines
  Added in v.866
  ViewVC Help
Powered by ViewVC 1.1.26