/[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 790 by dpavlin, Wed Dec 13 10:08:27 2006 UTC revision 851 by dpavlin, Sun May 27 11:14:40 2007 UTC
# Line 12  use WebPAC::Input 0.16; Line 12  use WebPAC::Input 0.16;
12  use WebPAC::Store 0.14;  use WebPAC::Store 0.14;
13  use WebPAC::Normalize 0.22;  use WebPAC::Normalize 0.22;
14  use WebPAC::Output::TT;  use WebPAC::Output::TT;
15  use WebPAC::Validate 0.06;  use WebPAC::Validate 0.11;
16  use WebPAC::Output::MARC;  use WebPAC::Output::MARC;
17  use WebPAC::Config;  use WebPAC::Config;
18  use Getopt::Long;  use Getopt::Long;
# 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    =item --validate-delimiters path/to/validate_delimiters_file
72    
73    this option is used with C<--validate> to turn on extra validation of
74    delimiters. If file is non existant, it will be created on first run.
75    
76    =item --marc-generate
77    
78    Generate MARC file. This will automatically be on if file contains C<marc*> directives.
79    You can use this option as C<--no-marc-generate> to disable MARC generation.
80    
81  =item --marc-lint  =item --marc-lint
82    
83  By default turned on if normalisation file has C<marc*> directives. You can disable lint  By default turned on if normalisation file has C<marc*> directives. You can disable lint
# Line 103  my $debug = 0; Line 113  my $debug = 0;
113  my $only_filter;  my $only_filter;
114  my $stats = 0;  my $stats = 0;
115  my $validate_path;  my $validate_path;
116    my $validate_delimiters_path;
117    my $marc_generate = 1;
118  my $marc_lint = 1;  my $marc_lint = 1;
119  my $marc_dump = 0;  my $marc_dump = 0;
120  my $parallel = 0;  my $parallel = 0;
# Line 122  GetOptions( Line 134  GetOptions(
134          "debug+" => \$debug,          "debug+" => \$debug,
135          "stats" => \$stats,          "stats" => \$stats,
136          "validate=s" => \$validate_path,          "validate=s" => \$validate_path,
137            "validate-delimiters=s" => \$validate_delimiters_path,
138            "marc-generate!" => \$marc_generate,
139          "marc-lint!" => \$marc_lint,          "marc-lint!" => \$marc_lint,
140          "marc-dump!" => \$marc_dump,          "marc-dump!" => \$marc_dump,
141          "parallel=i" => \$parallel,          "parallel=i" => \$parallel,
# Line 130  GetOptions( Line 144  GetOptions(
144          "help" => \$help,          "help" => \$help,
145  );  );
146    
147    $marc_generate = 0 if ( $validate_delimiters_path );
148    
149  pod2usage(-verbose => 2) if ($help);  pod2usage(-verbose => 2) if ($help);
150    
151  my $config = new WebPAC::Config( path => $config_path );  my $config = new WebPAC::Config( path => $config_path );
# Line 156  if ($merge) { Line 172  if ($merge) {
172          $log->info("created merge batch file $estcmd_path");          $log->info("created merge batch file $estcmd_path");
173  }  }
174    
   
175  my $validate;  my $validate;
176  $validate = new WebPAC::Validate(  $validate = new WebPAC::Validate(
177          path => $validate_path,          path => $validate_path,
178  ) if ($validate_path);          delimiters => $config->webpac('delimiters'),
179            delimiters_path => $validate_delimiters_path,
180    ) if ($validate_path || $validate_delimiters_path);
181    
182  my $use_indexer = $config->use_indexer;  my $use_indexer = $config->use_indexer;
183  $stats ||= $validate;  $stats ||= $validate;
# Line 431  while (my ($database, $db_config) = each Line 447  while (my ($database, $db_config) = each
447                  }                  }
448    
449                  my $marc;                  my $marc;
450                  if ($parser->have_rules( 'marc', $database, $input_name )) {                  if ($marc_generate && $parser->have_rules( 'marc', $database, $input_name )) {
451                          $marc = new WebPAC::Output::MARC(                          $marc = new WebPAC::Output::MARC(
452                                  path => "out/marc/${database}-${input_name}.marc",                                  path => "out/marc/${database}-${input_name}.marc",
453                                  lint => $marc_lint,                                  lint => $marc_lint,
# Line 537  while (my ($database, $db_config) = each Line 553  while (my ($database, $db_config) = each
553                                          $marc->add(                                          $marc->add(
554                                                  id => $mfn . ( $i ? "/$i" : '' ),                                                  id => $mfn . ( $i ? "/$i" : '' ),
555                                                  fields => $fields,                                                  fields => $fields,
556                                                  leader => WebPAC::Normalize::marc_leader(),                                                  leader => WebPAC::Normalize::_get_marc_leader(),
557                                                  row => $row,                                                  row => $row,
558                                          );                                          );
559                                          $i++;                                          $i++;
# Line 553  while (my ($database, $db_config) = each Line 569  while (my ($database, $db_config) = each
569                                  $log->info("validation errors:\n$errors\n" );                                  $log->info("validation errors:\n$errors\n" );
570                                  print $report_fh "$errors\n" if ($report_fh);                                  print $report_fh "$errors\n" if ($report_fh);
571                          }                          }
572    
573                            print $report_fh "\nAll possible subfields/delimiter templates:\n", $validate->delimiters_templates( report => 1, current_input => 1 ), "\n\n";
574    
575                            # must be last thing that touches $validate for this input
576                            $validate->reset;
577                  }                  }
578    
579                  if ($stats) {                  if ($stats) {
# Line 565  while (my ($database, $db_config) = each Line 586  while (my ($database, $db_config) = each
586                  $marc->finish if ($marc);                  $marc->finish if ($marc);
587    
588                  # close report                  # close report
589                  close($report_fh) if ($report_fh)                  close($report_fh) if ($report_fh);
   
590          }          }
591    
592          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));
# Line 593  if ($parallel) { Line 613  if ($parallel) {
613          $log->info("all parallel processes finished");          $log->info("all parallel processes finished");
614  }  }
615    
616    # save new delimiters if needed
617    $validate->save_delimiters_templates if ( $validate );
618    
619  #  #
620  # handle links or merge after indexing  # handle links or merge after indexing
621  #  #
Legend:
Removed from v.790  
changed lines
  Added in v.851
  ViewVC Help
Powered by ViewVC 1.1.26