|
Line 0
Link Here
|
| 0 |
- |
1 |
#!/usr/bin/perl |
|
|
2 |
|
| 3 |
# Copyright 2013 C & P Bibliography Services |
| 4 |
# |
| 5 |
# This file is part of Koha. |
| 6 |
# |
| 7 |
# Koha is free software; you can redistribute it and/or modify it under the |
| 8 |
# terms of the GNU General Public License as published by the Free Software |
| 9 |
# Foundation; either version 3 of the License, or (at your option) any later |
| 10 |
# version. |
| 11 |
# |
| 12 |
# Koha is distributed in the hope that it will be useful, but WITHOUT ANY |
| 13 |
# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR |
| 14 |
# A PARTICULAR PURPOSE. See the GNU General Public License for more details. |
| 15 |
# |
| 16 |
# You should have received a copy of the GNU General Public License along |
| 17 |
# with Koha; if not, write to the Free Software Foundation, Inc., |
| 18 |
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. |
| 19 |
|
| 20 |
=head1 NAME |
| 21 |
|
| 22 |
dedup_records.pl - clean up duplicate records |
| 23 |
|
| 24 |
=head1 SYNOPSIS |
| 25 |
|
| 26 |
dedup_records.pl --match=1 -a |
| 27 |
|
| 28 |
dedup_records.pl --match="LC-card-number/010a" --select="date" \ |
| 29 |
--limit="authid > 367123592" -a |
| 30 |
|
| 31 |
dedup_records.pl --match="Match/100abcdefghijklmnopqrstuvwxyz" \ |
| 32 |
--select="source=DLC" --select="date" --limit="authtypecode='PERSO_NAME'" -a |
| 33 |
|
| 34 |
=head1 DESCRIPTION |
| 35 |
|
| 36 |
This script will identify duplicate records, and either suggest that you |
| 37 |
merge them (in the case of bibliographic records) or automatically merge |
| 38 |
them for you (in the case of authority records). |
| 39 |
|
| 40 |
=cut |
| 41 |
|
| 42 |
use Modern::Perl; |
| 43 |
use Getopt::Long; |
| 44 |
use Pod::Usage; |
| 45 |
use Time::HiRes qw/time/; |
| 46 |
use POSIX qw/strftime ceil/; |
| 47 |
|
| 48 |
use C4::Context; |
| 49 |
use C4::Matcher; |
| 50 |
use C4::Biblio; |
| 51 |
use C4::AuthoritiesMarc; |
| 52 |
|
| 53 |
sub usage { |
| 54 |
pod2usage( -verbose => 2 ); |
| 55 |
exit; |
| 56 |
} |
| 57 |
|
| 58 |
=head1 OPTIONS |
| 59 |
|
| 60 |
=over 8 |
| 61 |
|
| 62 |
=item B<--help> |
| 63 |
|
| 64 |
Prints this help |
| 65 |
|
| 66 |
=item B<-v|--verbose> |
| 67 |
|
| 68 |
Print verbose log information (warning: very verbose!). |
| 69 |
|
| 70 |
=item B<-t|--test> |
| 71 |
|
| 72 |
Do not actually make any changes to the database, just report what changes |
| 73 |
would be made. |
| 74 |
|
| 75 |
=item B<-r|--report> |
| 76 |
|
| 77 |
Print a report of what happened during the run. |
| 78 |
|
| 79 |
=item B<-l|--limit=S> |
| 80 |
|
| 81 |
Only process those records that match the user-specified WHERE clause |
| 82 |
(the WHERE is implied and should not be included on the command line). |
| 83 |
|
| 84 |
=item B<-a|--authorities> |
| 85 |
|
| 86 |
Check for duplicate authorities rather that duplicate bibliographic records. |
| 87 |
|
| 88 |
=item B<-s|--select=s> |
| 89 |
|
| 90 |
Repeatable. Specify how to identify which record to prefer. See the section |
| 91 |
on SELECTORS below. |
| 92 |
|
| 93 |
=item B<-m|--match=s> |
| 94 |
|
| 95 |
Specifies the matching rule to use. This can be the numeric ID of a matching |
| 96 |
rule that you have already configured (preferred), or you can specify a matching |
| 97 |
rule on the command-line in the following format: |
| 98 |
|
| 99 |
<index1>/<tag1><subfield1>[##<index2>/<tag2><subfield2>[##...]] |
| 100 |
|
| 101 |
Examples: |
| 102 |
|
| 103 |
at/152b##he-main/2..a##he/2..bxyzt##ident/009@ |
| 104 |
authtype/152b##he-main,ext/2..a##he,ext/2..bxyz |
| 105 |
sn,ne,st-numeric/001##authtype/152b##he-main,ext/2..a##he,ext/2..bxyz |
| 106 |
|
| 107 |
=item B<-c|--check=s> |
| 108 |
|
| 109 |
Only relevant when you are using a matching rule specified on the command |
| 110 |
line. Specifies sanity checks to use to ensure that the records are really |
| 111 |
duplicate. The format is <tag1><subfields1>[,<tag2><subfields2>[,...]] |
| 112 |
|
| 113 |
Examples: |
| 114 |
|
| 115 |
200abxyz will check subfields a,b,x,y,z of 200 fields |
| 116 |
009@,152b will check 009 data and 152$b subfields |
| 117 |
|
| 118 |
=back |
| 119 |
|
| 120 |
=cut |
| 121 |
|
| 122 |
my ( |
| 123 |
$match, $check, $verbose, $test_only, $help, |
| 124 |
$limit, $report, $authorities, @select |
| 125 |
); |
| 126 |
my $max_matches = 50; |
| 127 |
|
| 128 |
my $result = GetOptions( |
| 129 |
'h|help' => \$help, |
| 130 |
'v|verbose' => \$verbose, |
| 131 |
't|test' => \$test_only, |
| 132 |
'r|report' => \$report, |
| 133 |
'l|limit=s' => \$limit, |
| 134 |
'a|authorities' => \$authorities, |
| 135 |
's|select=s' => \@select, |
| 136 |
'm|match=s' => \$match, |
| 137 |
'c|check=s' => \$check, |
| 138 |
); |
| 139 |
|
| 140 |
if ( $help || !$match ) { |
| 141 |
usage(); |
| 142 |
} |
| 143 |
|
| 144 |
my $starttime = time(); |
| 145 |
|
| 146 |
my $sql = |
| 147 |
$authorities |
| 148 |
? 'SELECT authid FROM auth_header' |
| 149 |
: 'SELECT biblionumber FROM biblio'; |
| 150 |
|
| 151 |
if ($limit) { |
| 152 |
$sql .= ' WHERE ' . $limit; |
| 153 |
} |
| 154 |
|
| 155 |
@select = ('date') unless @select; |
| 156 |
|
| 157 |
my @selectors = (); |
| 158 |
|
| 159 |
=head1 SELECTORS |
| 160 |
|
| 161 |
This script supports a number of selectors for choosing which record |
| 162 |
is "better." |
| 163 |
|
| 164 |
=over 8 |
| 165 |
|
| 166 |
=cut |
| 167 |
|
| 168 |
foreach my $sel (@select) { |
| 169 |
for ($sel) { |
| 170 |
|
| 171 |
=item B<score> |
| 172 |
|
| 173 |
Prefer the record which is the best match based on the specified matching |
| 174 |
rule. This will probably only be useful in cases where the matching rule |
| 175 |
will not match the source record, since the source record will automatically |
| 176 |
be given a score of 2 * the matching rule threshold if it wasn't picked up |
| 177 |
by the matcher. |
| 178 |
|
| 179 |
=cut |
| 180 |
|
| 181 |
when (/^score$/) { |
| 182 |
push @selectors, sub { return $_[0]->{'score'} || 0 } |
| 183 |
} |
| 184 |
|
| 185 |
=item B<date> |
| 186 |
|
| 187 |
Prefer the record which is newer based on the 005 field. |
| 188 |
|
| 189 |
=cut |
| 190 |
|
| 191 |
when (/^date$/) { |
| 192 |
push @selectors, sub { |
| 193 |
return ( |
| 194 |
defined $_[0]->{'record'}->field('005') |
| 195 |
? $_[0]->{'record'}->field('005')->data() |
| 196 |
: 0 |
| 197 |
); |
| 198 |
} |
| 199 |
} |
| 200 |
|
| 201 |
=item B<source=ABC> |
| 202 |
|
| 203 |
MARC21 only. Prefer records which come from ABC based on the 003 field. |
| 204 |
|
| 205 |
=cut |
| 206 |
|
| 207 |
when (/^source=(.*)$/) { |
| 208 |
push @selectors, sub { |
| 209 |
return ( |
| 210 |
defined $_[0]->{'record'}->field('003') |
| 211 |
&& $_[0]->{'record'}->field('003')->data() eq $1 |
| 212 |
? 1 |
| 213 |
: 0 |
| 214 |
); |
| 215 |
} |
| 216 |
} |
| 217 |
|
| 218 |
=item B<usage> |
| 219 |
|
| 220 |
Authorities only. Prefer the record used in the most bibliographic records. |
| 221 |
|
| 222 |
=cut |
| 223 |
|
| 224 |
when (/^usage$/) { |
| 225 |
if ($authorities) { |
| 226 |
push @selectors, sub { |
| 227 |
return CountUsage( $_[0]->{'record_id'} ); |
| 228 |
} |
| 229 |
} |
| 230 |
} |
| 231 |
|
| 232 |
=item B<ppn> |
| 233 |
|
| 234 |
UNIMARC only. Prefer records which have a PPN in the 009 field. |
| 235 |
|
| 236 |
=cut |
| 237 |
|
| 238 |
when (/^ppn$/) { |
| 239 |
push @selectors, sub { |
| 240 |
return defined $_[0]->{'record'}->field('009') |
| 241 |
&& $_[0]->{'record'}->field('009')->data() ? 1 : 0; |
| 242 |
} |
| 243 |
} |
| 244 |
} |
| 245 |
} |
| 246 |
|
| 247 |
=back |
| 248 |
|
| 249 |
=cut |
| 250 |
|
| 251 |
my $GetRecord; |
| 252 |
if ($authorities) { |
| 253 |
$GetRecord = \&C4::AuthoritiesMarc::GetAuthority; |
| 254 |
} |
| 255 |
else { |
| 256 |
$GetRecord = \&C4::Biblio::GetBiblio; |
| 257 |
} |
| 258 |
|
| 259 |
my $matcher; |
| 260 |
|
| 261 |
if ( $match =~ m#/# ) { |
| 262 |
my @matchers = split( '##', $match ); |
| 263 |
my $cnt = 0; |
| 264 |
$matcher = C4::Matcher->new( $authorities ? 'authority' : 'biblio' ); |
| 265 |
$matcher->threshold( 1000 * scalar(@matchers) - 1 ); |
| 266 |
$matcher->code('TEMP'); |
| 267 |
$matcher->description('Temporary matcher for deduplication run'); |
| 268 |
while ( $matchers[ $cnt++ ] =~ m#^([^/]+)/([0-9]{3})(.*)$# ) { |
| 269 |
$matcher->add_simple_matchpoint( $1, 1000, $2, $3, 0, 0, '' ); |
| 270 |
} |
| 271 |
my @checks = split( ',', $check ); |
| 272 |
$cnt = 0; |
| 273 |
while ( $checks[ $cnt++ ] =~ m#^([0-9]{3})(.*)$# ) { |
| 274 |
$matcher->add_simple_required_check( $1, $2, 0, 0, $1, $2, 0, 0, '' ); |
| 275 |
} |
| 276 |
} |
| 277 |
else { |
| 278 |
$matcher = C4::Matcher->fetch($match); |
| 279 |
} |
| 280 |
|
| 281 |
unless ($matcher) { |
| 282 |
die "Unrecognized matcher. Bailing out.\n"; |
| 283 |
} |
| 284 |
|
| 285 |
print "Retrieving records for checking using query: $sql\n" if $verbose; |
| 286 |
my $dbh = C4::Context->dbh; |
| 287 |
my $sth = $dbh->prepare($sql); |
| 288 |
$sth->execute(); |
| 289 |
|
| 290 |
my %merges = (); |
| 291 |
my %rematches = (); |
| 292 |
my $counter = 0; |
| 293 |
my $checked = 0; |
| 294 |
my $bibsmerged = 0; |
| 295 |
while ( my ($recordid) = $sth->fetchrow_array() ) { |
| 296 |
my $record; |
| 297 |
$counter++; |
| 298 |
print "Checking for matches for record #$counter (id $recordid).\n" |
| 299 |
if $verbose; |
| 300 |
next if ( $merges{$recordid} ); |
| 301 |
$checked++; |
| 302 |
$record = $GetRecord->($recordid); |
| 303 |
|
| 304 |
my @matches = (); |
| 305 |
if ( defined $matcher ) { |
| 306 |
@matches = $matcher->get_matches( $record, $max_matches ); |
| 307 |
} |
| 308 |
if ( scalar(@matches) > 1 ) { |
| 309 |
print "Found matches for $recordid.\n" if $verbose; |
| 310 |
my $foundself; |
| 311 |
foreach my $rec (@matches) { |
| 312 |
$foundself = 1 if $rec->{'record_id'} == $recordid; |
| 313 |
$rec->{'record'} = $GetRecord->( $rec->{'record_id'} ); |
| 314 |
my @weights = (); |
| 315 |
foreach my $selector (@selectors) { |
| 316 |
push @weights, $selector->($rec); |
| 317 |
} |
| 318 |
while ( scalar @weights < 5 ) { push @weights, 0; } |
| 319 |
$rec->{weights} = \@weights; |
| 320 |
} |
| 321 |
unless ($foundself) { |
| 322 |
my $rec = { |
| 323 |
'record_id' => $recordid, |
| 324 |
'score' => $matcher->threshold() * 2, |
| 325 |
'record' => $record |
| 326 |
}; |
| 327 |
my @weights = (); |
| 328 |
foreach my $selector (@selectors) { |
| 329 |
push @weights, $selector->($rec); |
| 330 |
} |
| 331 |
while ( scalar @weights < 5 ) { push @weights, 0; } |
| 332 |
$rec->{weights} = \@weights; |
| 333 |
} |
| 334 |
@matches = |
| 335 |
sort { |
| 336 |
$b->{'weights'}->[0] <=> $a->{'weights'}->[0] |
| 337 |
|| $b->{'weights'}->[1] <=> $a->{'weights'}->[1] |
| 338 |
|| $b->{'weights'}->[2] <=> $a->{'weights'}->[2] |
| 339 |
|| $b->{'weights'}->[3] <=> $a->{'weights'}->[3] |
| 340 |
|| $b->{'weights'}->[4] <=> $a->{'weights'}->[4] |
| 341 |
} @matches; |
| 342 |
print "Identified the following records for merging: " |
| 343 |
. join( ', ', map { $_->{'record_id'} } @matches ) . ".\n" |
| 344 |
if $verbose; |
| 345 |
if ( $merges{ $matches[0]->{'record_id'} } |
| 346 |
&& $merges{ $matches[0]->{'record_id'} } != |
| 347 |
$matches[0]->{'record_id'} ) |
| 348 |
{ |
| 349 |
print |
| 350 |
"Preferred record $matches[0]->{'record_id'} has already been merged into $merges{$matches[0]->{'record_id'}}.\n" |
| 351 |
if $verbose; |
| 352 |
$rematches{ $matches[0]->{'record_id'} } = |
| 353 |
$merges{ $matches[0]->{'record_id'} }; |
| 354 |
next; |
| 355 |
} |
| 356 |
foreach my $ii ( 0 .. $#matches ) { |
| 357 |
next if $merges{ $matches[$ii]->{'record_id'} }; |
| 358 |
$merges{ $matches[$ii]->{'record_id'} } = |
| 359 |
$matches[0]->{'record_id'}; |
| 360 |
if ( $ii > 0 ) { |
| 361 |
if ($authorities) { |
| 362 |
print "Merging authority " |
| 363 |
. $matches[$ii]->{'record_id'} |
| 364 |
. " into " |
| 365 |
. $matches[0]->{'record_id'} . ".\n" |
| 366 |
if $verbose; |
| 367 |
if ( !$test_only ) { |
| 368 |
my $editedbibs = merge( |
| 369 |
$matches[$ii]->{'record_id'}, |
| 370 |
$matches[$ii]->{'record'}, |
| 371 |
$matches[0]->{'record_id'}, |
| 372 |
$matches[0]->{'record'} |
| 373 |
); |
| 374 |
print |
| 375 |
"Changed $editedbibs bibliographic records in the course of merging " |
| 376 |
. $matches[$ii]->{'record_id'} |
| 377 |
. " into " |
| 378 |
. $matches[0]->{'record_id'} . ".\n" |
| 379 |
if $verbose; |
| 380 |
|
| 381 |
$bibsmerged += $editedbibs; |
| 382 |
} |
| 383 |
} |
| 384 |
else { |
| 385 |
print "Bib " |
| 386 |
. $matches[$ii]->{'record_id'} |
| 387 |
. " should probably be merged into " |
| 388 |
. $matches[0]->{'record_id'} . ".\n"; |
| 389 |
} |
| 390 |
} |
| 391 |
} |
| 392 |
} |
| 393 |
} |
| 394 |
|
| 395 |
$authorities = 'yes' if $authorities; |
| 396 |
$match = $matcher->code() unless $match =~ m#/#; |
| 397 |
$check = '(none)' unless $check; |
| 398 |
my $endtime = time(); |
| 399 |
my $totaltime = ceil( ( $endtime - $starttime ) * 1000 ); |
| 400 |
$starttime = strftime( '%D %T', localtime($starttime) ); |
| 401 |
$endtime = strftime( '%D %T', localtime($endtime) ); |
| 402 |
|
| 403 |
my @preferred = |
| 404 |
map { $_ eq $merges{$_} ? $_ : () } sort { $a <=> $b } keys %merges; |
| 405 |
my $keepers = scalar(@preferred); |
| 406 |
my @replaced = |
| 407 |
map { $_ ne $merges{$_} ? "$_ => $merges{$_}" : () } |
| 408 |
sort { $a <=> $b } keys %merges; |
| 409 |
my $obsoleted = scalar(@replaced); |
| 410 |
my @rematched = |
| 411 |
map { $_ ne $rematches{$_} ? "$_ (already replaced by $rematches{$_})" : () } |
| 412 |
sort { $a <=> $b } keys %rematches; |
| 413 |
my $needrematch = scalar(@rematched); |
| 414 |
my $summary = <<_SUMMARY_; |
| 415 |
|
| 416 |
Deduplication report |
| 417 |
======================================================= |
| 418 |
Match rule: $match |
| 419 |
Check rule: $check |
| 420 |
Selectors: @select |
| 421 |
Run started at: $starttime |
| 422 |
Run ended at: $endtime |
| 423 |
Total run time: $totaltime ms |
| 424 |
Number of records retrieved: $counter |
| 425 |
Number of records checked: $checked |
| 426 |
Number of records chosen: $keepers |
| 427 |
Number of records obsoleted: $obsoleted |
| 428 |
Records requiring relink/rerun: $needrematch |
| 429 |
Bibs merged on authority run: $bibsmerged |
| 430 |
_SUMMARY_ |
| 431 |
$summary .= "\n**** Ran in test mode only ****\n" if $test_only; |
| 432 |
|
| 433 |
print $summary; |
| 434 |
|
| 435 |
if ($report) { |
| 436 |
if (@preferred) { |
| 437 |
print <<_PREFERRED_HEADER_; |
| 438 |
|
| 439 |
Records identified as preferred |
| 440 |
------------------------------------------------------- |
| 441 |
|
| 442 |
_PREFERRED_HEADER_ |
| 443 |
|
| 444 |
print join( "\n", @preferred ) . "\n\n"; |
| 445 |
} |
| 446 |
|
| 447 |
if (@replaced) { |
| 448 |
print <<_REPLACED_HEADER_; |
| 449 |
|
| 450 |
Records replaced |
| 451 |
------------------------------------------------------- |
| 452 |
|
| 453 |
_REPLACED_HEADER_ |
| 454 |
|
| 455 |
print join( "\n", @replaced ) . "\n\n"; |
| 456 |
} |
| 457 |
|
| 458 |
if (@rematched) { |
| 459 |
print <<_REMATCHES_HEADER_; |
| 460 |
|
| 461 |
Records requiring index update to match |
| 462 |
------------------------------------------------------- |
| 463 |
|
| 464 |
_REMATCHES_HEADER_ |
| 465 |
|
| 466 |
print join( "\n", @rematched ) . "\n\n"; |
| 467 |
|
| 468 |
print "Please re-run this script after running the linker.\n"; |
| 469 |
} |
| 470 |
|
| 471 |
print $summary; |
| 472 |
} |
| 473 |
|
| 474 |
=head1 AUTHOR |
| 475 |
|
| 476 |
Jared Camins-Esakov <jcamins@cpbibliography.com> |
| 477 |
|
| 478 |
=cut |