/[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 765 by dpavlin, Wed Oct 25 21:01:51 2006 UTC revision 856 by dpavlin, Sun May 27 16:00:26 2007 UTC
# Line 8  use lib './lib'; Line 8  use lib './lib';
8    
9  use WebPAC::Common 0.02;  use WebPAC::Common 0.02;
10  use WebPAC::Parser 0.08;  use WebPAC::Parser 0.08;
11  use WebPAC::Input 0.14;  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 21  use Time::HiRes qw/time/; Line 21  use Time::HiRes qw/time/;
21  use File::Slurp;  use File::Slurp;
22  use Data::Dump qw/dump/;  use Data::Dump qw/dump/;
23  use Storable qw/dclone/;  use Storable qw/dclone/;
24    use Pod::Usage qw/pod2usage/;
25    
26  use Proc::Queue size => 1;  use Proc::Queue size => 1;
27  use POSIX ":sys_wait_h"; # imports WNOHANG  use POSIX ":sys_wait_h"; # imports WNOHANG
# Line 31  run.pl - start WebPAC indexing Line 32  run.pl - start WebPAC indexing
32    
33  B<this command will probably go away. Don't get used to it!>  B<this command will probably go away. Don't get used to it!>
34    
35  Options:  =head1 OPTIONS
36    
37  =over 4  =over 4
38    
# Line 67  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
80    
81    Generate MARC file. This will automatically be on if file contains C<marc*> directives.
82    You can use this option as C<--no-marc-generate> to disable MARC generation.
83    
84  =item --marc-lint  =item --marc-lint
85    
86  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 102  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;
121  my $marc_lint = 1;  my $marc_lint = 1;
122  my $marc_dump = 0;  my $marc_dump = 0;
123  my $parallel = 0;  my $parallel = 0;
124  my $only_links = 0;  my $only_links = 0;
125  my $merge = 0;  my $merge = 0;
126    my $help;
127    
128  my $log = _new WebPAC::Common()->_get_logger();  my $log = _new WebPAC::Common()->_get_logger();
129    
# Line 120  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,
142          "marc-lint!" => \$marc_lint,          "marc-lint!" => \$marc_lint,
143          "marc-dump!" => \$marc_dump,          "marc-dump!" => \$marc_dump,
144          "parallel=i" => \$parallel,          "parallel=i" => \$parallel,
145          "only-links!" => \$only_links,          "only-links!" => \$only_links,
146          "merge" => \$merge,          "merge" => \$merge,
147            "help" => \$help,
148  );  );
149    
150    $marc_generate = 0 if ( $validate_delimiters_path );
151    
152    pod2usage(-verbose => 2) if ($help);
153    
154  my $config = new WebPAC::Config( path => $config_path );  my $config = new WebPAC::Config( path => $config_path );
155    
156  #print "config = ",dump($config) if ($debug);  #print "config = ",dump($config) if ($debug);
# Line 135  die "no databases in config file!\n" unl Line 159  die "no databases in config file!\n" unl
159    
160  $log->info( "-" x 79 );  $log->info( "-" x 79 );
161    
162    my $log_file = 'log';
163    
164    if (-e $log_file ) {    # && -s $log_file > 5 * 1024 * 1024) {
165            $log->info("moved old log with ", -s $log_file, " bytes to '${log_file}.old'");
166            rename $log_file, "${log_file}.old" || $log->logwarn("can't rename $log_file to ${log_file}.old: $!");
167    }
168    
169  my $estcmd_fh;  my $estcmd_fh;
170  my $estcmd_path = './estcmd-merge.sh';  my $estcmd_path = './estcmd-merge.sh';
# Line 145  if ($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(
180          path => $validate_path,          path => $validate_path,
181  ) if ($validate_path);          delimiters => $config->webpac('delimiters'),
182            delimiters_path => $validate_delimiters_path,
183    ) if ($validate_path || $validate_delimiters_path);
184    
185  my $use_indexer = $config->use_indexer;  my $use_indexer = $config->use_indexer;
186    $stats ||= $validate;
187  if ($stats) {  if ($stats) {
188          $log->debug("option --stats disables update of indexing engine...");          $log->debug("disabled indexing for stats collection");
189          $use_indexer = undef;          $use_indexer = undef;
190  } else {  } else {
191          $log->info("using $use_indexer indexing engine...");          $log->info("using $use_indexer indexing engine...");
# Line 419  while (my ($database, $db_config) = each Line 450  while (my ($database, $db_config) = each
450                  }                  }
451    
452                  my $marc;                  my $marc;
453                  if ($parser->have_rules( 'marc', $database, $input_name )) {                  if ($marc_generate && $parser->have_rules( 'marc', $database, $input_name )) {
454                          $marc = new WebPAC::Output::MARC(                          $marc = new WebPAC::Output::MARC(
455                                  path => "out/marc/${database}-${input_name}.marc",                                  path => "out/marc/${database}-${input_name}.marc",
456                                  lint => $marc_lint,                                  lint => $marc_lint,
# Line 467  while (my ($database, $db_config) = each Line 498  while (my ($database, $db_config) = each
498    
499                          my $row = $input_db->fetch || next;                          my $row = $input_db->fetch || next;
500    
501                            $total_rows++;
502    
503                          my $mfn = $row->{'000'}->[0];                          my $mfn = $row->{'000'}->[0];
504    
505                          if (! $mfn || $mfn !~ m#^\d+$#) {                          if (! $mfn || $mfn !~ m#^\d+$#) {
# Line 477  while (my ($database, $db_config) = each Line 510  while (my ($database, $db_config) = each
510    
511    
512                          if ($validate) {                          if ($validate) {
513                                  if ( my $errors = $validate->validate_errors( $row, $input_db->dump ) ) {                                  if ( my $errors = $validate->validate_rec( $row, $input_db->dump_ascii ) ) {
514                                          $log->error( "MFN $mfn validation error:\n",                                          $log->error( "MFN $mfn validation error:\n",
515                                                  $validate->report_error( $errors )                                                  $validate->report_error( $errors )
516                                          );                                          );
517                                  }                                  }
518                                    next;   # validation doesn't create any output
519                          }                          }
520    
521                          my $ds = WebPAC::Normalize::data_structure(                          my $ds = WebPAC::Normalize::data_structure(
# Line 522  while (my ($database, $db_config) = each Line 556  while (my ($database, $db_config) = each
556                                          $marc->add(                                          $marc->add(
557                                                  id => $mfn . ( $i ? "/$i" : '' ),                                                  id => $mfn . ( $i ? "/$i" : '' ),
558                                                  fields => $fields,                                                  fields => $fields,
559                                                  leader => WebPAC::Normalize::marc_leader(),                                                  leader => WebPAC::Normalize::_get_marc_leader(),
560                                                  row => $row,                                                  row => $row,
561                                          );                                          );
562                                          $i++;                                          $i++;
# Line 530  while (my ($database, $db_config) = each Line 564  while (my ($database, $db_config) = each
564    
565                                  $log->info("Created $i instances of MFN $mfn\n") if ($i > 1);                                  $log->info("Created $i instances of MFN $mfn\n") if ($i > 1);
566                          }                          }
   
                         $total_rows++;  
567                  }                  }
568    
569                  if ($validate) {                  if ($validate) {
# Line 540  while (my ($database, $db_config) = each Line 572  while (my ($database, $db_config) = each
572                                  $log->info("validation errors:\n$errors\n" );                                  $log->info("validation errors:\n$errors\n" );
573                                  print $report_fh "$errors\n" if ($report_fh);                                  print $report_fh "$errors\n" if ($report_fh);
574                          }                          }
575    
576                            print $report_fh "\nAll possible subfields/delimiter templates:\n", $validate->delimiters_templates( report => 1, current_input => 1 ), "\n\n";
577    
578                            # must be last thing that touches $validate for this input
579                            $validate->reset;
580                  }                  }
581    
582                  if ($stats) {                  if ($stats) {
# Line 552  while (my ($database, $db_config) = each Line 589  while (my ($database, $db_config) = each
589                  $marc->finish if ($marc);                  $marc->finish if ($marc);
590    
591                  # close report                  # close report
592                  close($report_fh) if ($report_fh)                  close($report_fh) if ($report_fh);
   
593          }          }
594    
595          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));          eval { $indexer->finish } if ($indexer && $indexer->can('finish'));
# Line 580  if ($parallel) { Line 616  if ($parallel) {
616          $log->info("all parallel processes finished");          $log->info("all parallel processes finished");
617  }  }
618    
619    # save new delimiters if needed
620    $validate->save_delimiters_templates if ( $validate );
621    
622  #  #
623  # handle links or merge after indexing  # handle links or merge after indexing
624  #  #
Legend:
Removed from v.765  
changed lines
  Added in v.856
  ViewVC Help
Powered by ViewVC 1.1.26