| 4 |
|
|
| 5 |
use Cwd qw/abs_path/; |
use Cwd qw/abs_path/; |
| 6 |
use File::Temp qw/tempdir/; |
use File::Temp qw/tempdir/; |
|
use Data::Dumper; |
|
| 7 |
use lib './lib'; |
use lib './lib'; |
| 8 |
|
|
| 9 |
use WebPAC::Common 0.02; |
use WebPAC::Common 0.02; |
| 10 |
use WebPAC::Lookup; |
use WebPAC::Parser 0.08; |
| 11 |
use WebPAC::Input 0.03; |
use WebPAC::Input 0.16; |
| 12 |
use WebPAC::Store 0.03; |
use WebPAC::Store 0.15; |
| 13 |
use WebPAC::Normalize::XML; |
use WebPAC::Normalize 0.22; |
| 14 |
use WebPAC::Output::TT; |
#use WebPAC::Output::TT; |
| 15 |
use WebPAC::Output::Estraier 0.05; |
use WebPAC::Validate 0.11; |
| 16 |
use YAML qw/LoadFile/; |
use WebPAC::Output::MARC; |
| 17 |
use LWP::Simple; |
use WebPAC::Config; |
| 18 |
|
use Getopt::Long; |
| 19 |
|
use File::Path; |
| 20 |
|
use Time::HiRes qw/time/; |
| 21 |
|
use File::Slurp; |
| 22 |
|
use Data::Dump qw/dump/; |
| 23 |
|
use Storable qw/dclone/; |
| 24 |
|
use Pod::Usage qw/pod2usage/; |
| 25 |
|
use LWP::Simple qw//; |
| 26 |
|
|
| 27 |
my $limit = shift @ARGV; |
use POSIX ":sys_wait_h"; # imports WNOHANG |
| 28 |
|
|
| 29 |
my $config = LoadFile('conf/config.yml'); |
=head1 NAME |
| 30 |
|
|
| 31 |
print "config = ",Dumper($config); |
run.pl - start WebPAC indexing |
| 32 |
|
|
| 33 |
die "no databases in config file!\n" unless ($config->{databases}); |
B<this command will probably go away. Don't get used to it!> |
| 34 |
|
|
| 35 |
|
=head1 OPTIONS |
| 36 |
|
|
| 37 |
|
=over 4 |
| 38 |
|
|
| 39 |
|
=item --offset 42 |
| 40 |
|
|
| 41 |
|
start loading (all) databases at offset 42 |
| 42 |
|
|
| 43 |
|
=item --limit 100 |
| 44 |
|
|
| 45 |
|
limit loading to 100 records |
| 46 |
|
|
| 47 |
|
=item --clean |
| 48 |
|
|
| 49 |
|
remove database and Hyper Estraier index before indexing |
| 50 |
|
|
| 51 |
|
=item --only=database_name/input_filter |
| 52 |
|
|
| 53 |
|
reindex just single database (legacy name is --one) |
| 54 |
|
|
| 55 |
|
C</input_filter> is optional part which can be C<name> |
| 56 |
|
or C<type> from input |
| 57 |
|
|
| 58 |
|
=item --config conf/config.yml |
| 59 |
|
|
| 60 |
|
path to YAML configuration file |
| 61 |
|
|
| 62 |
|
=item --stats |
| 63 |
|
|
| 64 |
|
disable indexing, modify_* in configuration and dump statistics about field |
| 65 |
|
and subfield usage for each input |
| 66 |
|
|
| 67 |
|
=item --validate path/to/validation_file |
| 68 |
|
|
| 69 |
|
turn on extra validation of imput records, see L<WebPAC::Validation> |
| 70 |
|
|
| 71 |
|
You can use special variables C<$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 |
| 85 |
|
|
| 86 |
|
By default turned on if normalisation file has C<marc*> directives. You can disable lint |
| 87 |
|
messages with C<--no-marc-lint>. |
| 88 |
|
|
| 89 |
|
=item --marc-dump |
| 90 |
|
|
| 91 |
|
Force dump or input and marc record for debugging. |
| 92 |
|
|
| 93 |
|
=item --parallel 4 |
| 94 |
|
|
| 95 |
|
Run databases in parallel (aproximatly same as number of processors in |
| 96 |
|
machine if you want to use full load) |
| 97 |
|
|
| 98 |
|
=item --only-links |
| 99 |
|
|
| 100 |
|
Create just links |
| 101 |
|
|
| 102 |
|
=item --merge |
| 103 |
|
|
| 104 |
|
Create merged index of databases which have links |
| 105 |
|
|
| 106 |
|
=item --mirror http://www.example.com |
| 107 |
|
|
| 108 |
|
Tries to download input path files from mirror URI |
| 109 |
|
|
| 110 |
|
=back |
| 111 |
|
|
| 112 |
|
=cut |
| 113 |
|
|
| 114 |
|
my $offset; |
| 115 |
|
my $limit; |
| 116 |
|
|
| 117 |
|
my $clean = 0; |
| 118 |
|
my $config_path; |
| 119 |
|
my $debug = 0; |
| 120 |
|
my $only_filter; |
| 121 |
|
my $stats = 0; |
| 122 |
|
my $validate_path; |
| 123 |
|
my $validate_delimiters_path; |
| 124 |
|
my $marc_generate = 1; |
| 125 |
|
my $marc_lint = 1; |
| 126 |
|
my $marc_dump = 0; |
| 127 |
|
my $parallel = 0; |
| 128 |
|
my $only_links = 0; |
| 129 |
|
my $merge = 0; |
| 130 |
|
my $mirror; |
| 131 |
|
my $help; |
| 132 |
|
|
| 133 |
|
my $log = _new WebPAC::Common()->_get_logger(); |
| 134 |
|
|
| 135 |
|
GetOptions( |
| 136 |
|
"limit=i" => \$limit, |
| 137 |
|
"offset=i" => \$offset, |
| 138 |
|
"clean" => \$clean, |
| 139 |
|
"one=s" => \$only_filter, |
| 140 |
|
"only=s" => \$only_filter, |
| 141 |
|
"config=s" => \$config_path, |
| 142 |
|
"debug+" => \$debug, |
| 143 |
|
"stats" => \$stats, |
| 144 |
|
"validate=s" => \$validate_path, |
| 145 |
|
"validate-delimiters=s" => \$validate_delimiters_path, |
| 146 |
|
"marc-generate!" => \$marc_generate, |
| 147 |
|
"marc-lint!" => \$marc_lint, |
| 148 |
|
"marc-dump!" => \$marc_dump, |
| 149 |
|
"parallel=i" => \$parallel, |
| 150 |
|
"only-links!" => \$only_links, |
| 151 |
|
"merge" => \$merge, |
| 152 |
|
"mirror=s" => \$mirror, |
| 153 |
|
"help" => \$help, |
| 154 |
|
); |
| 155 |
|
|
| 156 |
|
$marc_generate = 0 if ( $validate_delimiters_path ); |
| 157 |
|
|
| 158 |
|
pod2usage(-verbose => 2) if ($help); |
| 159 |
|
|
| 160 |
|
my $config = new WebPAC::Config( path => $config_path ); |
| 161 |
|
|
| 162 |
|
WebPAC::Normalize::_debug( $debug - 1 ) if $debug > 1; |
| 163 |
|
|
| 164 |
|
#print "config = ",dump($config) if ($debug); |
| 165 |
|
|
| 166 |
|
die "no databases in config file!\n" unless ($config->databases); |
| 167 |
|
|
| 168 |
|
$log->info( "-" x 79 ); |
| 169 |
|
|
| 170 |
|
my $log_file = 'log'; |
| 171 |
|
|
| 172 |
|
if (-e $log_file ) { # && -s $log_file > 5 * 1024 * 1024) { |
| 173 |
|
$log->info("moved old log with ", -s $log_file, " bytes to '${log_file}.old'"); |
| 174 |
|
rename $log_file, "${log_file}.old" || $log->logwarn("can't rename $log_file to ${log_file}.old: $!"); |
| 175 |
|
} |
| 176 |
|
|
| 177 |
|
my $estcmd_fh; |
| 178 |
|
my $estcmd_path = './estcmd-merge.sh'; |
| 179 |
|
if ($merge) { |
| 180 |
|
open($estcmd_fh, '>', $estcmd_path) || $log->logdie("can't open $estcmd_path: $!"); |
| 181 |
|
print $estcmd_fh 'cd /data/estraier/_node/ || exit 1',$/; |
| 182 |
|
print $estcmd_fh 'sudo /etc/init.d/hyperestraier stop',$/; |
| 183 |
|
$log->info("created merge batch file $estcmd_path"); |
| 184 |
|
} |
| 185 |
|
|
| 186 |
|
my $validate; |
| 187 |
|
$validate = new WebPAC::Validate( |
| 188 |
|
delimiters => $config->webpac('delimiters'), |
| 189 |
|
) if ($validate_path || $validate_delimiters_path); |
| 190 |
|
|
| 191 |
|
my $use_indexer = $config->use_indexer; |
| 192 |
|
$stats ||= $validate; |
| 193 |
|
if ($stats) { |
| 194 |
|
$log->debug("disabled indexing for stats collection"); |
| 195 |
|
$use_indexer = undef; |
| 196 |
|
} elsif ( $use_indexer ) { |
| 197 |
|
$log->info("using $use_indexer indexing engine..."); |
| 198 |
|
} |
| 199 |
|
|
| 200 |
|
# parse normalize files and create source files for lookup and normalization |
| 201 |
|
|
| 202 |
|
my ($only_database,$only_input) = split(m#/#, $only_filter) if $only_filter; |
| 203 |
|
|
| 204 |
|
my $parser = new WebPAC::Parser( |
| 205 |
|
config => $config, |
| 206 |
|
only_database => $only_database, |
| 207 |
|
only_input => $only_input, |
| 208 |
|
); |
| 209 |
|
|
| 210 |
my $total_rows = 0; |
my $total_rows = 0; |
| 211 |
|
my $start_t = time(); |
| 212 |
|
|
| 213 |
while (my ($database, $db_config) = each %{ $config->{databases} }) { |
my @links; |
| 214 |
|
|
| 215 |
my $log = _new WebPAC::Common()->_get_logger(); |
if ($parallel) { |
| 216 |
|
eval 'use Proc::Queue size => 1;'; |
| 217 |
|
die $@ if $@; |
| 218 |
|
$log->info("Using $parallel processes for speedup"); |
| 219 |
|
Proc::Queue::size($parallel); |
| 220 |
|
} |
| 221 |
|
|
| 222 |
# |
sub create_ds_config { |
| 223 |
# open Hyper Estraier database |
my ($db_config, $database, $input, $mfn) = @_; |
| 224 |
# |
my $c = dclone( $db_config ); |
| 225 |
|
$c->{_} = $database || $log->logconfess("need database"); |
| 226 |
|
$c->{_mfn} = $mfn || $log->logconfess("need mfn"); |
| 227 |
|
$c->{input} = $input || $log->logconfess("need input"); |
| 228 |
|
return $c; |
| 229 |
|
} |
| 230 |
|
|
| 231 |
my $est_config = $config->{hyperestraier} || $log->logdie("can't find 'hyperestraier' part in confguration"); |
foreach my $database ( sort keys %{ $config->databases } ) { |
| 232 |
$est_config->{database} = $database; |
my $db_config = $config->databases->{$database}; |
| 233 |
|
|
| 234 |
$log->info("using HyperEstraier URL $est_config->{masterurl}"); |
next if ($only_database && $database !~ m/$only_database/i); |
| 235 |
|
|
| 236 |
|
if ($parallel) { |
| 237 |
|
my $f=fork; |
| 238 |
|
if(defined ($f) and $f==0) { |
| 239 |
|
$log->info("Created processes $$ for speedup"); |
| 240 |
|
} else { |
| 241 |
|
next; |
| 242 |
|
} |
| 243 |
|
} |
| 244 |
|
|
| 245 |
|
my $indexer; |
| 246 |
|
if ($use_indexer && $parser->have_rules( 'search', $database )) { |
| 247 |
|
|
| 248 |
|
my $cfg_name = $use_indexer; |
| 249 |
|
$cfg_name =~ s/\-.*$//; |
| 250 |
|
|
| 251 |
|
my $indexer_config = $config->get( $cfg_name ) || $log->logdie("can't find '$cfg_name' part in confguration"); |
| 252 |
|
$indexer_config->{database} = $database; |
| 253 |
|
$indexer_config->{clean} = $clean; |
| 254 |
|
$indexer_config->{label} = $db_config->{name}; |
| 255 |
|
|
| 256 |
|
# force clean if database has links |
| 257 |
|
$indexer_config->{clean} = 1 if ($db_config->{links}); |
| 258 |
|
|
| 259 |
|
if ($use_indexer eq 'hyperestraier') { |
| 260 |
|
|
| 261 |
|
# open Hyper Estraier database |
| 262 |
|
require WebPAC::Output::Estraier; |
| 263 |
|
$indexer = new WebPAC::Output::Estraier( %{ $indexer_config } ); |
| 264 |
|
|
| 265 |
|
} elsif ($use_indexer eq 'hyperestraier-native') { |
| 266 |
|
|
| 267 |
|
# open Hyper Estraier database |
| 268 |
|
require WebPAC::Output::EstraierNative; |
| 269 |
|
$indexer = new WebPAC::Output::EstraierNative( %{ $indexer_config } ); |
| 270 |
|
|
| 271 |
|
} elsif ($use_indexer eq 'kinosearch') { |
| 272 |
|
|
| 273 |
|
die "no longer supported"; |
| 274 |
|
|
| 275 |
|
} else { |
| 276 |
|
$log->logdie("unknown use_indexer: $use_indexer"); |
| 277 |
|
} |
| 278 |
|
|
| 279 |
|
$log->logdie("can't continue without valid indexer") unless ($indexer); |
| 280 |
|
} |
| 281 |
|
|
|
my $est = new WebPAC::Output::Estraier( |
|
|
%{ $est_config }, |
|
|
); |
|
| 282 |
|
|
| 283 |
# |
# |
| 284 |
# now WebPAC::Store |
# store Hyper Estraier links to other databases |
| 285 |
# |
# |
| 286 |
my $abs_path = abs_path($0); |
if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) { |
| 287 |
$abs_path =~ s#/[^/]*$#/#; |
foreach my $link (@{ $db_config->{links} }) { |
| 288 |
|
if ($use_indexer eq 'hyperestraier') { |
| 289 |
|
if ($merge) { |
| 290 |
|
print $estcmd_fh 'sudo -u www-data estcmd merge ' . $database . ' ' . $link->{to},$/; |
| 291 |
|
} else { |
| 292 |
|
$log->info("saving link $database -> $link->{to} [$link->{credit}]"); |
| 293 |
|
push @links, sub { |
| 294 |
|
$log->info("adding link $database -> $link->{to} [$link->{credit}]"); |
| 295 |
|
$indexer->add_link( |
| 296 |
|
from => $database, |
| 297 |
|
to => $link->{to}, |
| 298 |
|
credit => $link->{credit}, |
| 299 |
|
); |
| 300 |
|
}; |
| 301 |
|
} |
| 302 |
|
} else { |
| 303 |
|
$log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]"); |
| 304 |
|
} |
| 305 |
|
} |
| 306 |
|
} |
| 307 |
|
next if ($only_links); |
| 308 |
|
|
|
my $db_path = $config->{webpac}->{db_path} . '/' . $database; |
|
| 309 |
|
|
| 310 |
$log->info("working on $database in $db_path"); |
# |
| 311 |
|
# now WebPAC::Store |
| 312 |
|
# |
| 313 |
|
my $store = new WebPAC::Store({ |
| 314 |
|
debug => $debug, |
| 315 |
|
}); |
| 316 |
|
|
| 317 |
my $db = new WebPAC::Store( |
|
| 318 |
path => $db_path, |
# |
| 319 |
database => $database, |
# prepare output |
| 320 |
debug => 1, |
# |
| 321 |
); |
my @outputs = force_array( $db_config->{output}, sub { |
| 322 |
|
$log->error("Database $database doesn't have any outputs defined. Do you want to remove it from configuration?" ); |
| 323 |
|
} ); |
| 324 |
|
|
| 325 |
|
my @output_modules; |
| 326 |
|
|
| 327 |
|
foreach my $output ( @outputs ) { |
| 328 |
|
|
| 329 |
|
#warn '## output = ',dump( $output ); |
| 330 |
|
|
| 331 |
|
my $module = $output->{module} || $log->logdie("need module in output section of $database"); |
| 332 |
|
$module = 'WebPAC::Output::' . $module unless $module =~ m/::/; |
| 333 |
|
|
| 334 |
|
$log->debug("loading output module $module"); |
| 335 |
|
eval "require $module"; |
| 336 |
|
|
| 337 |
|
# add database to arugemnts for output filter |
| 338 |
|
$output->{database} = $database; |
| 339 |
|
$output->{clean} = $clean; |
| 340 |
|
|
| 341 |
|
$log->debug("calling $module->new(",dump( $output ),")"); |
| 342 |
|
my $out = new $module->new( $output ); |
| 343 |
|
if ( $out->init ) { |
| 344 |
|
push @output_modules, $out; |
| 345 |
|
} else { |
| 346 |
|
$log->warn("SKIPPED $module"); |
| 347 |
|
} |
| 348 |
|
} |
| 349 |
|
|
| 350 |
|
|
| 351 |
# |
# |
| 352 |
# now, iterate through input formats |
# now, iterate through input formats |
| 353 |
# |
# |
| 354 |
|
|
| 355 |
my @inputs; |
|
| 356 |
if (ref($db_config->{input}) eq 'ARRAY') { |
my @inputs = force_array( $db_config->{input}, sub { |
|
@inputs = @{ $db_config->{input} }; |
|
|
} elsif ($db_config->{input}) { |
|
|
push @inputs, $db_config->{input}; |
|
|
} else { |
|
| 357 |
$log->info("database $database doesn't have inputs defined"); |
$log->info("database $database doesn't have inputs defined"); |
| 358 |
} |
} ); |
| 359 |
|
|
| 360 |
my @supported_inputs = keys %{ $config->{webpac}->{inputs} }; |
if ( -e 'out/debug' ) { # FIXME flag? |
| 361 |
|
my $out; |
| 362 |
|
foreach my $i ( @inputs ) { |
| 363 |
|
warn dump( $i ); |
| 364 |
|
next unless defined $i->{normalize}; |
| 365 |
|
warn dump( $i->{normalize} ); |
| 366 |
|
foreach my $normalize ( @{ $i->{normalize} } ) { |
| 367 |
|
my $path = $normalize->{path}; |
| 368 |
|
$out .= qq/\n##\n## $path\n##\n\n/; |
| 369 |
|
$out .= read_file( $path ); |
| 370 |
|
} |
| 371 |
|
} |
| 372 |
|
my $all = "out/debug/all-normalize.pl"; |
| 373 |
|
write_file( $all, $out ); |
| 374 |
|
warn "### all normalize for this input saved to: $all"; |
| 375 |
|
}; |
| 376 |
|
|
| 377 |
foreach my $input (@inputs) { |
foreach my $input (@inputs) { |
| 378 |
|
|
| 379 |
|
my $input_name = $input->{name} || $log->logdie("input without a name isn't valid: ",dump($input)); |
| 380 |
|
|
| 381 |
|
if ( $input->{skip} ) { |
| 382 |
|
$log->info("skip $input_name"); |
| 383 |
|
next; |
| 384 |
|
} |
| 385 |
|
|
| 386 |
|
next if ($only_input && ($input_name !~ m#$only_input#i && $input->{type} !~ m#$only_input#i)); |
| 387 |
|
|
| 388 |
my $type = lc($input->{type}); |
my $type = lc($input->{type}); |
| 389 |
|
|
| 390 |
die "I know only how to handle input types ", join(",", @supported_inputs), " not '$type'!\n" unless (grep(/$type/, @supported_inputs)); |
# FIXME check if input module exists |
| 391 |
|
my $input_module = $input->{module}; |
| 392 |
|
|
| 393 |
|
if ( ! $input_module ) { |
| 394 |
|
if ( grep(/$type/, $config->webpac('inputs')) ) { |
| 395 |
|
$input_module = $config->webpac('inputs')->{$type}; |
| 396 |
|
} else { |
| 397 |
|
$log->logdie("I know only how to handle input types ", join(",", $config->webpac('inputs') ), " not '$type'!" ); |
| 398 |
|
} |
| 399 |
|
} |
| 400 |
|
|
| 401 |
|
my @lookups = $parser->have_lookup_create($database, $input); |
| 402 |
|
|
| 403 |
my $lookup = new WebPAC::Lookup( |
$log->info("working on $database/$input_name with $input_module on $input->{path}", |
| 404 |
lookup_file => $input->{lookup}, |
@lookups ? " creating lookups: ".join(", ", @lookups) : "" |
| 405 |
); |
); |
| 406 |
|
|
| 407 |
my $input_module = $config->{webpac}->{inputs}->{$type}; |
if ($stats) { |
| 408 |
|
# disable modification of records if --stats is in use |
| 409 |
|
delete($input->{modify_records}); |
| 410 |
|
delete($input->{modify_file}); |
| 411 |
|
} |
| 412 |
|
|
| 413 |
$log->info("working on input $input->{path} [$input->{type}] using $input_module"); |
if ( $mirror ) { |
| 414 |
|
my $path = $input->{path} || die "no input path in ",dump( $input ); |
| 415 |
|
|
| 416 |
|
my $base = $path; |
| 417 |
|
$base =~ s{/[^/]+$}{}; |
| 418 |
|
mkpath $base unless -e $base; |
| 419 |
|
|
| 420 |
|
my $rc = LWP::Simple::mirror( "$mirror/$path", $path ); |
| 421 |
|
if (LWP::Simple::is_error( $rc )) { |
| 422 |
|
die "can't mirror $mirror/$path -> $path [$rc]"; |
| 423 |
|
} else { |
| 424 |
|
$log->info( "mirror ", $path, " [$rc] ", -s $path, " bytes" ); |
| 425 |
|
} |
| 426 |
|
|
| 427 |
|
} |
| 428 |
|
|
| 429 |
my $input_db = new WebPAC::Input( |
my $input_db = new WebPAC::Input( |
| 430 |
module => $input_module, |
module => $input_module, |
| 431 |
code_page => $config->{webpac}->{webpac_encoding}, |
limit => $limit || $input->{limit}, |
| 432 |
limit_mfn => $input->{limit}, |
offset => $offset, |
| 433 |
lookup => $lookup, |
recode => $input->{recode}, |
| 434 |
|
stats => $stats, |
| 435 |
|
modify_records => $input->{modify_records}, |
| 436 |
|
modify_file => $input->{modify_file}, |
| 437 |
|
input_config => $input, |
| 438 |
); |
); |
| 439 |
$log->logdie("can't create input using $input_module") unless ($input); |
$log->logdie("can't create input using $input_module") unless ($input); |
| 440 |
|
|
| 441 |
|
if (defined( $input->{lookup} )) { |
| 442 |
|
$log->warn("$database/$input_name has depriciated lookup definition, removing it..."); |
| 443 |
|
delete( $input->{lookup} ); |
| 444 |
|
} |
| 445 |
|
|
| 446 |
|
my $lookup_coderef; |
| 447 |
|
|
| 448 |
|
if (@lookups) { |
| 449 |
|
|
| 450 |
|
my $rules = $parser->lookup_create_rules($database, $input) || $log->logdie("no rules found for $database/$input"); |
| 451 |
|
|
| 452 |
|
$lookup_coderef = sub { |
| 453 |
|
my $rec = shift || die "need rec!"; |
| 454 |
|
my $mfn = $rec->{'000'}->[0] || die "need mfn in 000"; |
| 455 |
|
|
| 456 |
|
WebPAC::Normalize::data_structure( |
| 457 |
|
row => $rec, |
| 458 |
|
rules => $rules, |
| 459 |
|
config => create_ds_config( $db_config, $database, $input, $mfn ), |
| 460 |
|
); |
| 461 |
|
|
| 462 |
|
#warn "current lookup: ", dump(WebPAC::Normalize::_get_lookup()); |
| 463 |
|
}; |
| 464 |
|
|
| 465 |
|
WebPAC::Normalize::_set_lookup( undef ); |
| 466 |
|
|
| 467 |
|
$log->debug("created lookup_coderef using:\n$rules"); |
| 468 |
|
|
| 469 |
|
}; |
| 470 |
|
|
| 471 |
|
my $lookup_jar; |
| 472 |
|
|
| 473 |
my $maxmfn = $input_db->open( |
my $maxmfn = $input_db->open( |
| 474 |
path => $input->{path}, |
path => $input->{path}, |
| 475 |
code_page => $input->{encoding}, # database encoding |
input_encoding => $input->{encoding}, # database encoding |
| 476 |
); |
lookup_coderef => $lookup_coderef, |
| 477 |
|
lookup => $lookup_jar, |
| 478 |
|
%{ $input }, |
| 479 |
|
load_row => sub { |
| 480 |
|
my $a = shift; |
| 481 |
|
return $store->load_row( |
| 482 |
|
database => $database, |
| 483 |
|
input => $input_name, |
| 484 |
|
id => $a->{id}, |
| 485 |
|
); |
| 486 |
|
}, |
| 487 |
|
save_row => sub { |
| 488 |
|
my $a = shift; |
| 489 |
|
return $store->save_row( |
| 490 |
|
database => $database, |
| 491 |
|
input => $input_name, |
| 492 |
|
id => $a->{id}, |
| 493 |
|
row => $a->{row}, |
| 494 |
|
); |
| 495 |
|
}, |
| 496 |
|
|
|
my $n = new WebPAC::Normalize::XML( |
|
|
# filter => { 'foo' => sub { shift } }, |
|
|
db => $db, |
|
|
lookup_regex => $lookup->regex, |
|
|
lookup => $lookup, |
|
|
prefix => $input->{name}, |
|
| 497 |
); |
); |
| 498 |
|
|
| 499 |
my $normalize_path = $input->{normalize}->{path}; |
my $lookup_data = WebPAC::Normalize::_get_lookup(); |
| 500 |
|
|
| 501 |
if ($normalize_path =~ m/\.xml$/i) { |
if (defined( $lookup_data->{$database}->{$input_name} )) { |
| 502 |
$n->open( |
$log->debug("created following lookups: ", sub { dump( $lookup_data ) } ); |
| 503 |
tag => $input->{normalize}->{tag}, |
|
| 504 |
xml_file => $input->{normalize}->{path}, |
foreach my $key (keys %{ $lookup_data->{$database}->{$input_name} }) { |
| 505 |
); |
$store->save_lookup( |
| 506 |
} elsif ($normalize_path =~ m/\.(?:yml|yaml)$/i) { |
database => $database, |
| 507 |
$n->open_yaml( |
input => $input_name, |
| 508 |
path => $normalize_path, |
key => $key, |
| 509 |
tag => $input->{normalize}->{tag}, |
data => $lookup_data->{$database}->{$input_name}->{$key}, |
| 510 |
|
); |
| 511 |
|
} |
| 512 |
|
} |
| 513 |
|
|
| 514 |
|
my $report_fh; |
| 515 |
|
if ($stats || $validate) { |
| 516 |
|
my $out_report = 'out/report'; # FIXME move to config |
| 517 |
|
mkpath $out_report unless -e $out_report; |
| 518 |
|
my $path = "$out_report/${database}-${input_name}.txt"; |
| 519 |
|
open($report_fh, '>', $path) || $log->logdie("can't open $path: $!"); |
| 520 |
|
|
| 521 |
|
print $report_fh "Report for database '$database' input '$input_name' records ", |
| 522 |
|
$offset || 1, "-", $limit || $input->{limit} || $maxmfn, "\n\n"; |
| 523 |
|
$log->info("Generating report file $path"); |
| 524 |
|
|
| 525 |
|
if ( $validate ) { |
| 526 |
|
$validate->read_validate_file( $validate->fill_in( $validate_path, database => $database, input => $input_name ) ) if ( $validate_path ); |
| 527 |
|
$validate->read_validate_delimiters_file( $validate->fill_in( $validate_delimiters_path, database => $database, input => $input_name ) ) if ( $validate_delimiters_path ); |
| 528 |
|
} |
| 529 |
|
} |
| 530 |
|
|
| 531 |
|
my $marc; |
| 532 |
|
if ($marc_generate && $parser->have_rules( 'marc', $database, $input_name )) { |
| 533 |
|
|
| 534 |
|
my $out_marc = 'out/marc'; # FIXME move to config |
| 535 |
|
mkpath $out_marc unless -e $out_marc; |
| 536 |
|
|
| 537 |
|
$marc = new WebPAC::Output::MARC( |
| 538 |
|
path => "$out_marc/${database}-${input_name}.marc", |
| 539 |
|
lint => $marc_lint, |
| 540 |
|
dump => $marc_dump, |
| 541 |
); |
); |
| 542 |
} |
} |
| 543 |
|
|
| 544 |
for ( 0 ... $input_db->size ) { |
my $rules = $parser->normalize_rules($database,$input_name); |
| 545 |
|
if ( ! $rules ) { |
| 546 |
|
$log->logwarn("no normalize rules for $database/$input_name", $input_db->input_module->can('normalize') ? " using normalize from input module" : ''); |
| 547 |
|
next; |
| 548 |
|
} |
| 549 |
|
|
| 550 |
my $row = $input_db->fetch || next; |
$log->debug("parsed normalize rules:\n$rules"); |
| 551 |
|
|
| 552 |
my $mfn = $row->{'000'}->[0] || die "can't find MFN"; |
# reset position in database |
| 553 |
|
$input_db->seek(1); |
| 554 |
|
|
| 555 |
my $ds = $n->data_structure($row); |
# generate name of config key for indexer (strip everything after -) |
| 556 |
|
my $indexer_config = $use_indexer; |
| 557 |
|
$indexer_config =~ s/^(\w+)-?.*$/$1/g if ($indexer_config); |
| 558 |
|
|
| 559 |
|
my $lookup_hash; |
| 560 |
|
my $depends = $parser->depends($database,$input_name); |
| 561 |
|
|
| 562 |
|
if ($depends) { |
| 563 |
|
$log->debug("$database/$input_name depends on: ", dump($depends)) if ($depends); |
| 564 |
|
$log->logdie("parser->depends didn't return HASH") unless (ref($depends) eq 'HASH'); |
| 565 |
|
|
| 566 |
|
foreach my $db (keys %$depends) { |
| 567 |
|
foreach my $i (keys %{$depends->{$db}}) { |
| 568 |
|
foreach my $k (keys %{$depends->{$db}->{$i}}) { |
| 569 |
|
my $t = time(); |
| 570 |
|
$log->debug("loading lookup $db/$i"); |
| 571 |
|
$lookup_hash->{$db}->{$i}->{$k} = $store->load_lookup( |
| 572 |
|
database => $db, |
| 573 |
|
input => $i, |
| 574 |
|
key => $k, |
| 575 |
|
); |
| 576 |
|
$log->debug(sprintf("lookup $db/$i took %.2fs", time() - $t)); |
| 577 |
|
} |
| 578 |
|
} |
| 579 |
|
} |
| 580 |
|
|
| 581 |
$est->add( |
$log->debug("lookup_hash = ", sub { dump( $lookup_hash ) }); |
| 582 |
id => $input->{name} . "#" . $mfn, |
} |
| 583 |
ds => $ds, |
|
| 584 |
type => $config->{hyperestraier}->{type}, |
|
| 585 |
); |
# setup input name for all output filters |
| 586 |
|
foreach my $out ( @output_modules ) { |
| 587 |
|
if ( $out->can('input') ) { |
| 588 |
|
$out->input( $input_name ); |
| 589 |
|
} else { |
| 590 |
|
$log->warn("output filter ",ref($out)," doesn't support input name"); |
| 591 |
|
} |
| 592 |
|
} |
| 593 |
|
|
| 594 |
|
|
| 595 |
|
foreach my $pos ( 0 ... $input_db->size ) { |
| 596 |
|
|
| 597 |
|
my $row = $input_db->fetch || next; |
| 598 |
|
|
| 599 |
$total_rows++; |
$total_rows++; |
| 600 |
|
|
| 601 |
|
my $mfn = $row->{'000'}->[0]; |
| 602 |
|
|
| 603 |
|
if (! $mfn || $mfn !~ m{^\d+$}) { |
| 604 |
|
$log->warn("record $pos doesn't have valid MFN but '$mfn', using $pos"); |
| 605 |
|
$mfn = $pos; |
| 606 |
|
push @{ $row->{'000'} }, $pos; |
| 607 |
|
} |
| 608 |
|
|
| 609 |
|
foreach my $out ( @output_modules ) { |
| 610 |
|
$out->add_row( $mfn, $row ) if $out->can('add_row'); |
| 611 |
|
} |
| 612 |
|
|
| 613 |
|
if ($validate) { |
| 614 |
|
if ( my $errors = $validate->validate_rec( $row, $input_db->dump_ascii ) ) { |
| 615 |
|
$log->error( "MFN $mfn validation error:\n", |
| 616 |
|
$validate->report_error( $errors ) |
| 617 |
|
); |
| 618 |
|
} |
| 619 |
|
next; # validation doesn't create any output |
| 620 |
|
} |
| 621 |
|
|
| 622 |
|
my $ds; |
| 623 |
|
|
| 624 |
|
if ($rules) { |
| 625 |
|
|
| 626 |
|
$ds = WebPAC::Normalize::data_structure( |
| 627 |
|
row => $row, |
| 628 |
|
rules => $rules, |
| 629 |
|
lookup => $lookup_hash, |
| 630 |
|
config => create_ds_config( $db_config, $database, $input, $mfn ), |
| 631 |
|
marc_encoding => 'utf-8', |
| 632 |
|
load_row_coderef => sub { |
| 633 |
|
my ($database,$input,$mfn) = @_; |
| 634 |
|
#warn "### load_row($database,$input,$mfn) from data_structure\n"; |
| 635 |
|
return $store->load_row( |
| 636 |
|
database => $database, |
| 637 |
|
input => $input, |
| 638 |
|
id => $mfn, |
| 639 |
|
); |
| 640 |
|
}, |
| 641 |
|
); |
| 642 |
|
|
| 643 |
|
} elsif ( $input_db->input_module->can('normalize') ) { |
| 644 |
|
$ds = $input_db->input_module->normalize( $mfn ); |
| 645 |
|
} |
| 646 |
|
|
| 647 |
|
if ( $ds ) { |
| 648 |
|
$log->debug("ds = ", sub { dump($ds) }); |
| 649 |
|
|
| 650 |
|
$store->save_ds( |
| 651 |
|
database => $database, |
| 652 |
|
input => $input_name, |
| 653 |
|
id => $mfn, |
| 654 |
|
ds => $ds, |
| 655 |
|
) if !$stats; |
| 656 |
|
|
| 657 |
|
$indexer->add( |
| 658 |
|
id => "${input_name}/${mfn}", |
| 659 |
|
ds => $ds, |
| 660 |
|
type => $config->get($indexer_config)->{type}, |
| 661 |
|
) if $indexer; |
| 662 |
|
|
| 663 |
|
foreach my $out ( @output_modules ) { |
| 664 |
|
$out->add( $mfn, $ds ) if $out->can('add'); |
| 665 |
|
} |
| 666 |
|
|
| 667 |
|
} else { |
| 668 |
|
$log->warn("record $pos didn't produce any output after normalization rules!") unless $marc; |
| 669 |
|
} |
| 670 |
|
if ($marc) { |
| 671 |
|
my $i = 0; |
| 672 |
|
|
| 673 |
|
while (my $fields = WebPAC::Normalize::MARC::_get_marc_fields( fetch_next => 1 ) ) { |
| 674 |
|
$marc->add( |
| 675 |
|
id => $mfn . ( $i ? "/$i" : '' ), |
| 676 |
|
fields => $fields, |
| 677 |
|
leader => WebPAC::Normalize::MARC::_get_marc_leader(), |
| 678 |
|
row => $row, |
| 679 |
|
); |
| 680 |
|
$i++; |
| 681 |
|
} |
| 682 |
|
|
| 683 |
|
$log->info("Created $i instances of MFN $mfn\n") if ($i > 1); |
| 684 |
|
} |
| 685 |
|
|
| 686 |
} |
} |
| 687 |
|
|
| 688 |
}; |
if ($validate) { |
| 689 |
|
my $errors = $validate->report; |
| 690 |
|
if ($errors) { |
| 691 |
|
$log->info("validation errors:\n$errors\n" ); |
| 692 |
|
print $report_fh "$errors\n" if ($report_fh); |
| 693 |
|
} |
| 694 |
|
|
| 695 |
$log->info("$total_rows records indexed"); |
print $report_fh "\nAll possible subfields/delimiter templates:\n", $validate->delimiters_templates( report => 1, current_input => 1 ), "\n\n"; |
| 696 |
|
|
| 697 |
# |
# must be last thing that touches $validate for this input |
| 698 |
# add Hyper Estraier links to other databases |
$validate->reset; |
| 699 |
# |
} |
| 700 |
if (ref($db_config->{links}) eq 'ARRAY') { |
|
| 701 |
foreach my $link (@{ $db_config->{links} }) { |
if ($stats) { |
| 702 |
$log->info("adding link $database -> $link->{to} [$link->{credit}]"); |
my $s = $input_db->stats; |
| 703 |
$est->add_link( |
$log->info("statistics of fields usage:\n$s"); |
| 704 |
from => $database, |
print $report_fh "Statistics of fields usage:\n$s" if ($report_fh); |
|
to => $link->{to}, |
|
|
credit => $link->{credit}, |
|
|
); |
|
| 705 |
} |
} |
| 706 |
|
|
| 707 |
|
# close MARC file |
| 708 |
|
$marc->finish if ($marc); |
| 709 |
|
|
| 710 |
|
# close report |
| 711 |
|
close($report_fh) if ($report_fh); |
| 712 |
|
} |
| 713 |
|
|
| 714 |
|
$indexer->finish if $indexer && $indexer->can('finish'); |
| 715 |
|
|
| 716 |
|
foreach my $out ( @output_modules ) { |
| 717 |
|
$out->finish if $out->can('finish'); |
| 718 |
|
} |
| 719 |
|
|
| 720 |
|
my $dt = time() - $start_t; |
| 721 |
|
$log->info("$total_rows records ", $indexer ? "indexed " : "", |
| 722 |
|
sprintf("in %.2f sec [%.2f rec/sec]", |
| 723 |
|
$dt, ($total_rows / $dt) |
| 724 |
|
) |
| 725 |
|
); |
| 726 |
|
|
| 727 |
|
|
| 728 |
|
# end forked process |
| 729 |
|
if ($parallel) { |
| 730 |
|
$log->info("parallel process $$ finished"); |
| 731 |
|
exit(0); |
| 732 |
} |
} |
| 733 |
|
|
| 734 |
} |
} |
| 735 |
|
|
| 736 |
|
if ($parallel) { |
| 737 |
|
# wait all children to finish |
| 738 |
|
sleep(1) while wait != -1; |
| 739 |
|
$log->info("all parallel processes finished"); |
| 740 |
|
} |
| 741 |
|
|
| 742 |
|
# save new delimiters if needed |
| 743 |
|
$validate->save_delimiters_templates if ( $validate_delimiters_path ); |
| 744 |
|
|
| 745 |
|
# |
| 746 |
|
# handle links or merge after indexing |
| 747 |
|
# |
| 748 |
|
|
| 749 |
|
if ($merge) { |
| 750 |
|
print $estcmd_fh 'sudo /etc/init.d/hyperestraier start',$/; |
| 751 |
|
close($estcmd_fh); |
| 752 |
|
chmod 0700, $estcmd_path || $log->warn("can't chmod 0700 $estcmd_path: $!"); |
| 753 |
|
system $estcmd_path; |
| 754 |
|
} else { |
| 755 |
|
foreach my $link (@links) { |
| 756 |
|
$log->logdie("coderef in link ", Dumper($link), " is ", ref($link), " and not CODE") unless (ref($link) eq 'CODE'); |
| 757 |
|
$link->(); |
| 758 |
|
} |
| 759 |
|
} |
Parent Directory
| 
Revision Log
| 
Patch