22 |
use Data::Dump qw/dump/; |
use Data::Dump qw/dump/; |
23 |
use Storable qw/dclone/; |
use Storable qw/dclone/; |
24 |
|
|
25 |
|
use Proc::Queue size => 1; |
26 |
|
use POSIX ":sys_wait_h"; # imports WNOHANG |
27 |
|
|
28 |
=head1 NAME |
=head1 NAME |
29 |
|
|
30 |
run.pl - start WebPAC indexing |
run.pl - start WebPAC indexing |
84 |
|
|
85 |
Force dump or input and marc record for debugging. |
Force dump or input and marc record for debugging. |
86 |
|
|
87 |
|
=item --parallel 4 |
88 |
|
|
89 |
|
Run databases in parallel (aproximatly same as number of processors in |
90 |
|
machine if you want to use full load) |
91 |
|
|
92 |
|
=item --only-links |
93 |
|
|
94 |
|
Create just links |
95 |
|
|
96 |
|
=item --merge |
97 |
|
|
98 |
|
Create merged index of databases which have links |
99 |
|
|
100 |
=back |
=back |
101 |
|
|
102 |
=cut |
=cut |
113 |
my ($marc_normalize, $marc_output); |
my ($marc_normalize, $marc_output); |
114 |
my $marc_lint = 1; |
my $marc_lint = 1; |
115 |
my $marc_dump = 0; |
my $marc_dump = 0; |
116 |
|
my $parallel = 0; |
117 |
|
my $only_links = 0; |
118 |
|
my $merge = 0; |
119 |
|
|
120 |
GetOptions( |
GetOptions( |
121 |
"limit=i" => \$limit, |
"limit=i" => \$limit, |
131 |
"marc-output=s" => \$marc_output, |
"marc-output=s" => \$marc_output, |
132 |
"marc-lint!" => \$marc_lint, |
"marc-lint!" => \$marc_lint, |
133 |
"marc-dump!" => \$marc_dump, |
"marc-dump!" => \$marc_dump, |
134 |
|
"parallel=i" => \$parallel, |
135 |
|
"only-links!" => \$only_links, |
136 |
|
"merge" => \$merge, |
137 |
); |
); |
138 |
|
|
139 |
$config = LoadFile($config); |
$config = LoadFile($config); |
145 |
my $log = _new WebPAC::Common()->_get_logger(); |
my $log = _new WebPAC::Common()->_get_logger(); |
146 |
$log->info( "-" x 79 ); |
$log->info( "-" x 79 ); |
147 |
|
|
148 |
|
|
149 |
|
my $estcmd_fh; |
150 |
|
my $estcmd_path = './estcmd-merge.sh'; |
151 |
|
if ($merge) { |
152 |
|
open($estcmd_fh, '>', $estcmd_path) || $log->logdie("can't open $estcmd_path: $!"); |
153 |
|
print $estcmd_fh 'cd /data/estraier/_node/ || exit 1',$/; |
154 |
|
print $estcmd_fh 'sudo /etc/init.d/hyperestraier stop',$/; |
155 |
|
$log->info("created merge batch file $estcmd_path"); |
156 |
|
} |
157 |
|
|
158 |
|
|
159 |
my $validate; |
my $validate; |
160 |
$validate = new WebPAC::Validate( |
$validate = new WebPAC::Validate( |
161 |
path => $validate_path, |
path => $validate_path, |
162 |
) if ($validate_path); |
) if ($validate_path); |
163 |
|
|
164 |
|
|
165 |
my $use_indexer = $config->{use_indexer} || 'hyperestraier'; |
my $use_indexer = $config->{use_indexer} || 'hyperestraier'; |
166 |
if ($stats) { |
if ($stats) { |
167 |
$log->debug("option --stats disables update of indexing engine..."); |
$log->debug("option --stats disables update of indexing engine..."); |
177 |
my $start_t = time(); |
my $start_t = time(); |
178 |
|
|
179 |
my @links; |
my @links; |
180 |
my $indexer; |
|
181 |
|
if ($parallel) { |
182 |
|
$log->info("Using $parallel processes for speedup"); |
183 |
|
Proc::Queue::size($parallel); |
184 |
|
} |
185 |
|
|
186 |
while (my ($database, $db_config) = each %{ $config->{databases} }) { |
while (my ($database, $db_config) = each %{ $config->{databases} }) { |
187 |
|
|
188 |
my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter); |
my ($only_database,$only_input) = split(m#/#, $only_filter) if ($only_filter); |
189 |
next if ($only_database && $database !~ m/$only_database/i); |
next if ($only_database && $database !~ m/$only_database/i); |
190 |
|
|
191 |
|
if ($parallel) { |
192 |
|
my $f=fork; |
193 |
|
if(defined ($f) and $f==0) { |
194 |
|
$log->info("Created processes $$ for speedup"); |
195 |
|
} else { |
196 |
|
next; |
197 |
|
} |
198 |
|
} |
199 |
|
|
200 |
|
my $indexer; |
201 |
if ($use_indexer) { |
if ($use_indexer) { |
202 |
my $indexer_config = $config->{$use_indexer} || $log->logdie("can't find '$use_indexer' part in confguration"); |
my $indexer_config = $config->{$use_indexer} || $log->logdie("can't find '$use_indexer' part in confguration"); |
203 |
$indexer_config->{database} = $database; |
$indexer_config->{database} = $database; |
204 |
$indexer_config->{clean} = $clean; |
$indexer_config->{clean} = $clean; |
205 |
$indexer_config->{label} = $db_config->{name}; |
$indexer_config->{label} = $db_config->{name}; |
206 |
|
|
207 |
|
# force clean if database has links |
208 |
|
$indexer_config->{clean} = 1 if ($db_config->{links}); |
209 |
|
|
210 |
if ($use_indexer eq 'hyperestraier') { |
if ($use_indexer eq 'hyperestraier') { |
211 |
|
|
212 |
# open Hyper Estraier database |
# open Hyper Estraier database |
229 |
|
|
230 |
|
|
231 |
# |
# |
232 |
|
# store Hyper Estraier links to other databases |
233 |
|
# |
234 |
|
if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) { |
235 |
|
foreach my $link (@{ $db_config->{links} }) { |
236 |
|
if ($use_indexer eq 'hyperestraier') { |
237 |
|
if ($merge) { |
238 |
|
print $estcmd_fh 'sudo -u www-data estcmd merge ' . $database . ' ' . $link->{to},$/; |
239 |
|
} else { |
240 |
|
$log->info("saving link $database -> $link->{to} [$link->{credit}]"); |
241 |
|
push @links, sub { |
242 |
|
$log->info("adding link $database -> $link->{to} [$link->{credit}]"); |
243 |
|
$indexer->add_link( |
244 |
|
from => $database, |
245 |
|
to => $link->{to}, |
246 |
|
credit => $link->{credit}, |
247 |
|
); |
248 |
|
}; |
249 |
|
} |
250 |
|
} else { |
251 |
|
$log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]"); |
252 |
|
} |
253 |
|
} |
254 |
|
} |
255 |
|
next if ($only_links); |
256 |
|
|
257 |
|
|
258 |
|
# |
259 |
# now WebPAC::Store |
# now WebPAC::Store |
260 |
# |
# |
261 |
my $abs_path = abs_path($0); |
my $abs_path = abs_path($0); |
450 |
) |
) |
451 |
); |
); |
452 |
|
|
453 |
# |
|
454 |
# add Hyper Estraier links to other databases |
# end forked process |
455 |
# |
if ($parallel) { |
456 |
if (ref($db_config->{links}) eq 'ARRAY' && $use_indexer) { |
$log->info("parallel process $$ finished"); |
457 |
foreach my $link (@{ $db_config->{links} }) { |
exit(0); |
|
if ($use_indexer eq 'hyperestraier') { |
|
|
$log->info("saving link $database -> $link->{to} [$link->{credit}]"); |
|
|
push @links, { |
|
|
from => $database, |
|
|
to => $link->{to}, |
|
|
credit => $link->{credit}, |
|
|
}; |
|
|
} else { |
|
|
$log->warn("NOT IMPLEMENTED WITH $use_indexer: adding link $database -> $link->{to} [$link->{credit}]"); |
|
|
} |
|
|
} |
|
458 |
} |
} |
459 |
|
|
460 |
} |
} |
461 |
|
|
462 |
foreach my $link (@links) { |
if ($parallel) { |
463 |
$log->info("adding link $link->{from} -> $link->{to} [$link->{credit}]"); |
# wait all children to finish |
464 |
$indexer->add_link( %{ $link } ); |
sleep(1) while wait != -1; |
465 |
|
$log->info("all parallel processes finished"); |
466 |
|
} |
467 |
|
|
468 |
|
# |
469 |
|
# handle links or merge after indexing |
470 |
|
# |
471 |
|
|
472 |
|
if ($merge) { |
473 |
|
print $estcmd_fh 'sudo /etc/init.d/hyperestraier start',$/; |
474 |
|
close($estcmd_fh); |
475 |
|
chmod 0700, $estcmd_path || $log->warn("can't chmod 0700 $estcmd_path: $!"); |
476 |
|
system $estcmd_path; |
477 |
|
} else { |
478 |
|
foreach my $link (@links) { |
479 |
|
$log->logdie("coderef in link ", Dumper($link), " is ", ref($link), " and not CODE") unless (ref($link) eq 'CODE'); |
480 |
|
$link->(); |
481 |
|
} |
482 |
} |
} |