/[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 803 by dpavlin, Fri Mar 2 00:32:21 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;
# 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    =item --marc-generate
72    
73    Generate MARC file. This will automatically be on if file contains C<marc*> directives.
74    You can use this option as C<--no-marc-generate> to disable MARC generation.
75    
76  =item --marc-lint  =item --marc-lint
77    
78  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 108  my $debug = 0;
108  my $only_filter;  my $only_filter;
109  my $stats = 0;  my $stats = 0;
110  my $validate_path;  my $validate_path;
111    my $marc_generate = 1;
112  my $marc_lint = 1;  my $marc_lint = 1;
113  my $marc_dump = 0;  my $marc_dump = 0;
114  my $parallel = 0;  my $parallel = 0;
115  my $only_links = 0;  my $only_links = 0;
116  my $merge = 0;  my $merge = 0;
117    my $help;
118    
119  my $log = _new WebPAC::Common()->_get_logger();  my $log = _new WebPAC::Common()->_get_logger();
120    
# Line 120  GetOptions( Line 128  GetOptions(
128          "debug+" => \$debug,          "debug+" => \$debug,
129          "stats" => \$stats,          "stats" => \$stats,
130          "validate=s" => \$validate_path,          "validate=s" => \$validate_path,
131            "marc-generate!" => \$marc_generate,
132          "marc-lint!" => \$marc_lint,          "marc-lint!" => \$marc_lint,
133          "marc-dump!" => \$marc_dump,          "marc-dump!" => \$marc_dump,
134          "parallel=i" => \$parallel,          "parallel=i" => \$parallel,
135          "only-links!" => \$only_links,          "only-links!" => \$only_links,
136          "merge" => \$merge,          "merge" => \$merge,
137            "help" => \$help,
138  );  );
139    
140    pod2usage(-verbose => 2) if ($help);
141    
142  my $config = new WebPAC::Config( path => $config_path );  my $config = new WebPAC::Config( path => $config_path );
143    
144  #print "config = ",dump($config) if ($debug);  #print "config = ",dump($config) if ($debug);
# Line 135  die "no databases in config file!\n" unl Line 147  die "no databases in config file!\n" unl
147    
148  $log->info( "-" x 79 );  $log->info( "-" x 79 );
149    
150    my $log_file = 'log';
151    
152    if (-e $log_file ) {    # && -s $log_file > 5 * 1024 * 1024) {
153            $log->info("moved old log with ", -s $log_file, " bytes to '${log_file}.old'");
154            rename $log_file, "${log_file}.old" || $log->logwarn("can't rename $log_file to ${log_file}.old: $!");
155    }
156    
157  my $estcmd_fh;  my $estcmd_fh;
158  my $estcmd_path = './estcmd-merge.sh';  my $estcmd_path = './estcmd-merge.sh';
# Line 153  $validate = new WebPAC::Validate( Line 171  $validate = new WebPAC::Validate(
171    
172    
173  my $use_indexer = $config->use_indexer;  my $use_indexer = $config->use_indexer;
174    $stats ||= $validate;
175  if ($stats) {  if ($stats) {
176          $log->debug("option --stats disables update of indexing engine...");          $log->debug("disabled indexing for stats collection");
177          $use_indexer = undef;          $use_indexer = undef;
178  } else {  } else {
179          $log->info("using $use_indexer indexing engine...");          $log->info("using $use_indexer indexing engine...");
# Line 419  while (my ($database, $db_config) = each Line 438  while (my ($database, $db_config) = each
438                  }                  }
439    
440                  my $marc;                  my $marc;
441                  if ($parser->have_rules( 'marc', $database, $input_name )) {                  if ($marc_generate && $parser->have_rules( 'marc', $database, $input_name )) {
442                          $marc = new WebPAC::Output::MARC(                          $marc = new WebPAC::Output::MARC(
443                                  path => "out/marc/${database}-${input_name}.marc",                                  path => "out/marc/${database}-${input_name}.marc",
444                                  lint => $marc_lint,                                  lint => $marc_lint,
# Line 467  while (my ($database, $db_config) = each Line 486  while (my ($database, $db_config) = each
486    
487                          my $row = $input_db->fetch || next;                          my $row = $input_db->fetch || next;
488    
489                            $total_rows++;
490    
491                          my $mfn = $row->{'000'}->[0];                          my $mfn = $row->{'000'}->[0];
492    
493                          if (! $mfn || $mfn !~ m#^\d+$#) {                          if (! $mfn || $mfn !~ m#^\d+$#) {
# Line 477  while (my ($database, $db_config) = each Line 498  while (my ($database, $db_config) = each
498    
499    
500                          if ($validate) {                          if ($validate) {
501                                  if ( my $errors = $validate->validate_errors( $row, $input_db->dump ) ) {                                  if ( my $errors = $validate->validate_rec( $row, $input_db->dump_ascii ) ) {
502                                          $log->error( "MFN $mfn validation error:\n",                                          $log->error( "MFN $mfn validation error:\n",
503                                                  $validate->report_error( $errors )                                                  $validate->report_error( $errors )
504                                          );                                          );
505                                  }                                  }
506                                    next;   # validation doesn't create any output
507                          }                          }
508    
509                          my $ds = WebPAC::Normalize::data_structure(                          my $ds = WebPAC::Normalize::data_structure(
# Line 530  while (my ($database, $db_config) = each Line 552  while (my ($database, $db_config) = each
552    
553                                  $log->info("Created $i instances of MFN $mfn\n") if ($i > 1);                                  $log->info("Created $i instances of MFN $mfn\n") if ($i > 1);
554                          }                          }
   
                         $total_rows++;  
555                  }                  }
556    
557                  if ($validate) {                  if ($validate) {
Legend:
Removed from v.765  
changed lines
  Added in v.803
  ViewVC Help
Powered by ViewVC 1.1.26