Line 0
Link Here
|
0 |
- |
1 |
package C4::Catalog::Zebra; |
|
|
2 |
# |
3 |
# This file is part of Koha. |
4 |
# |
5 |
# Koha is free software; you can redistribute it and/or modify it under the |
6 |
# terms of the GNU General Public License as published by the Free Software |
7 |
# Foundation; either version 2 of the License, or (at your option) any later |
8 |
# version. |
9 |
# |
10 |
# Koha is distributed in the hope that it will be useful, but WITHOUT ANY |
11 |
# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR |
12 |
# A PARTICULAR PURPOSE. See the GNU General Public License for more details. |
13 |
# |
14 |
# You should have received a copy of the GNU General Public License along with |
15 |
# Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place, |
16 |
# Suite 330, Boston, MA 02111-1307 USA |
17 |
|
18 |
# Derived from rebuild_zebra.pl (2005-08-11) Paul Poulain and others |
19 |
# Rewriten 02/03/2011 by Tomas Cohen Arazi (tomascohen@gmail.com) |
20 |
# Universidad Nacional de Cordoba / Argentina |
21 |
|
22 |
# Library for managing updates in zebra, usually from zebraqueue |
23 |
|
24 |
use strict; |
25 |
use warnings; |
26 |
use C4::Context; |
27 |
use Getopt::Long; |
28 |
use File::Temp qw/ tempdir /; |
29 |
use File::Path; |
30 |
use Time::HiRes qw(time); |
31 |
use C4::Biblio; |
32 |
use C4::AuthoritiesMarc; |
33 |
|
34 |
use vars qw($VERSION @ISA @EXPORT); |
35 |
|
36 |
BEGIN { |
37 |
# set the version for version checking |
38 |
$VERSION = 0.01; |
39 |
|
40 |
require Exporter; |
41 |
@ISA = qw(Exporter); |
42 |
@EXPORT = qw( |
43 |
&UpdateAuths |
44 |
&UpdateBiblios |
45 |
&UpdateAuthsAndBiblios |
46 |
&IndexZebraqueueRecords |
47 |
); |
48 |
} |
49 |
|
50 |
|
51 |
=head1 NAME |
52 |
|
53 |
C4::Catalog::Zebra |
54 |
|
55 |
Comment: |
56 |
This should be used when merging the rest of the rebuild_zebra.pl indexing logic |
57 |
my $nosanitize = (C4::Context->preference('ZebraNoSanitize')) ? 1 : 0; |
58 |
|
59 |
|
60 |
=head2 UpdateAuths |
61 |
|
62 |
( $num_records_updated ) = &UpdateAuths (); |
63 |
|
64 |
returns the number of updated+deleted authority records |
65 |
|
66 |
=cut |
67 |
|
68 |
sub UpdateAuths |
69 |
{ |
70 |
# Update authorities |
71 |
return IndexZebraqueueRecords('authority'); |
72 |
} |
73 |
|
74 |
=head2 UpdateBiblios |
75 |
|
76 |
( $num_records_updated ) = &UpdateBiblios (); |
77 |
|
78 |
returns the number of updated+deleted biblio records |
79 |
|
80 |
=cut |
81 |
|
82 |
sub UpdateBiblios |
83 |
{ |
84 |
# Update authorities |
85 |
return IndexZebraqueueRecords('biblio'); |
86 |
} |
87 |
|
88 |
=head2 UpdateAuthsAndBiblios |
89 |
|
90 |
( $num_records_updated ) = &UpdateAuthsAndBiblios (); |
91 |
|
92 |
returns the number of updated+deleted authority and biblio records |
93 |
|
94 |
=cut |
95 |
|
96 |
sub UpdateAuthsAndBiblios |
97 |
{ |
98 |
my $ret; |
99 |
# Update authorities |
100 |
$ret = UpdateAuths(); |
101 |
|
102 |
# Update biblios |
103 |
$ret += UpdateBiblios(); |
104 |
|
105 |
return $ret; |
106 |
} |
107 |
|
108 |
=head2 IndexZebraqueueRecords |
109 |
|
110 |
( $num_records_updated ) = &IndexZebraqueueRecords ($record_type); |
111 |
|
112 |
returns the number of updated+deleted $record_type records |
113 |
|
114 |
Comment : |
115 |
$record_type can be either 'biblio' or 'authority' |
116 |
|
117 |
=cut |
118 |
|
119 |
sub IndexZebraqueueRecords |
120 |
{ |
121 |
my ($record_type) = @_; |
122 |
my $as_xml = (C4::Context->preference('ZebraUseXml')) ? 1 : 0; |
123 |
my $noxml = ($as_xml) ? 0 : 1; |
124 |
my $record_format = ($as_xml) ? 'marcxml' : 'iso2709' ; |
125 |
|
126 |
my ($num_records_updated,$num_records_deleted); |
127 |
|
128 |
$num_records_deleted = (IndexZebraqueueByAction('deleted',$record_type,$record_format,$as_xml,$noxml)||0); |
129 |
$num_records_updated = (IndexZebraqueueByAction('updated',$record_type,$record_format,$as_xml,$noxml)||0); |
130 |
|
131 |
return $num_records_deleted + $num_records_updated; |
132 |
} |
133 |
|
134 |
=head2 IndexZebraqueueByAction |
135 |
|
136 |
( $num_records_updated ) = &IndexZebraqueueByAction ($action,$record_type, |
137 |
$record_format,$as_xml,$noxml); |
138 |
|
139 |
returns the number of updated+deleted $record_type records |
140 |
|
141 |
Comment : |
142 |
$record_type can be 'biblio' or 'authority' |
143 |
$record_format can be 'marcxml' or 'iso2709' |
144 |
$action can be 'updated' or 'deleted' |
145 |
$as_xml and $noxml are maintained for legacy reasons, one is enough. They |
146 |
indicate whether to use marcxml for indexing in zebra or iso2709. They should |
147 |
all be deduced from C4::Context->preference('ZebraUseXml'). |
148 |
|
149 |
=cut |
150 |
|
151 |
sub IndexZebraqueueByAction |
152 |
{ |
153 |
my ($action,$record_type,$record_format,$as_xml,$noxml) = @_; |
154 |
my ($num_records_exported,$ret,$zaction); |
155 |
|
156 |
if ($action eq 'updated' or $action eq 'deleted') { |
157 |
# get records by action |
158 |
my $entries = select_zebraqueue_records($record_type, $action); |
159 |
# Create tmp dir |
160 |
my $directory = File::Temp->newdir(); |
161 |
|
162 |
# get records from zebraqueue, export to file for zebraidx |
163 |
if ($action eq 'updated') { |
164 |
$zaction = 'update'; |
165 |
$num_records_exported = export_marc_records_from_list($record_type, |
166 |
$entries, "$directory", $as_xml, $noxml); |
167 |
} else { |
168 |
# $action eq 'deleted' |
169 |
$zaction = 'delete'; |
170 |
$num_records_exported = generate_deleted_marc_records($record_type, |
171 |
$entries, "$directory", $as_xml); |
172 |
} |
173 |
|
174 |
if ($num_records_exported) { |
175 |
# log export |
176 |
my $time = localtime(time); |
177 |
print "$time $num_records_exported $record_type record(s) exported for $zaction\n"; |
178 |
# TODO error handling / and better logging |
179 |
$ret = DoIndexing($record_type,$zaction,"$directory",$record_format); |
180 |
if ($ret) { |
181 |
print "$time $num_records_exported $record_type record(s) $action\n"; |
182 |
mark_zebraqueue_batch_done($entries); |
183 |
print "$time $num_records_exported $record_type record(s) marked done in zebraqueue\n"; |
184 |
} |
185 |
# /TODO |
186 |
} |
187 |
} else { |
188 |
# Wrong action |
189 |
$ret = -1; |
190 |
} |
191 |
|
192 |
return $ret; |
193 |
} |
194 |
|
195 |
|
196 |
sub select_zebraqueue_records { |
197 |
my ($record_type, $update_type) = @_; |
198 |
|
199 |
my $dbh = C4::Context->dbh; |
200 |
my $server = ($record_type eq 'biblio') ? 'biblioserver' : 'authorityserver'; |
201 |
my $op = ($update_type eq 'deleted') ? 'recordDelete' : 'specialUpdate'; |
202 |
|
203 |
my $sth = $dbh->prepare(<<'SQL'); |
204 |
SELECT id, biblio_auth_number |
205 |
FROM zebraqueue |
206 |
WHERE server = ? |
207 |
AND operation = ? |
208 |
AND done = 0 |
209 |
ORDER BY id DESC; |
210 |
SQL |
211 |
|
212 |
$sth->execute($server, $op); |
213 |
my $entries = $sth->fetchall_arrayref({}); |
214 |
} |
215 |
|
216 |
sub mark_zebraqueue_batch_done { |
217 |
my ($entries) = @_; |
218 |
|
219 |
my $dbh = C4::Context->dbh; |
220 |
|
221 |
$dbh->{AutoCommit} = 0; |
222 |
my $sth = $dbh->prepare("UPDATE zebraqueue SET done = 1 WHERE id = ?"); |
223 |
$dbh->commit(); |
224 |
foreach my $id (map { $_->{id} } @$entries) { |
225 |
$sth->execute($id); |
226 |
} |
227 |
$dbh->{AutoCommit} = 1; |
228 |
} |
229 |
|
230 |
sub export_marc_records_from_list { |
231 |
my ($record_type, $entries, $directory, $as_xml, $noxml) = @_; |
232 |
my $verbose_logging = (C4::Context->preference('ZebraqueueVerboseLogging')) ? 1 : 0; |
233 |
|
234 |
my $num_exported = 0; |
235 |
open (OUT, ">:utf8 ", "$directory/exported_records") or die $!; |
236 |
my $i = 0; |
237 |
my %found = (); |
238 |
foreach my $record_number ( map { $_->{biblio_auth_number} } |
239 |
grep { !$found{ $_->{biblio_auth_number} }++ } |
240 |
@$entries ) { |
241 |
print "." if ( $verbose_logging ); |
242 |
print "\r$i" unless ($i++ %100 or !$verbose_logging); |
243 |
my ($marc) = get_corrected_marc_record($record_type, $record_number, $noxml); |
244 |
if (defined $marc) { |
245 |
# FIXME - when more than one record is exported and $as_xml is true, |
246 |
# the output file is not valid XML - it's just multiple <record> elements |
247 |
# strung together with no single root element. zebraidx doesn't seem |
248 |
# to care, though, at least if you're using the GRS-1 filter. It does |
249 |
# care if you're using the DOM filter, which requires valid XML file(s). |
250 |
print OUT ($as_xml) ? $marc->as_xml_record() : $marc->as_usmarc(); |
251 |
$num_exported++; |
252 |
} |
253 |
} |
254 |
print "\nRecords exported: $num_exported\n" if ( $verbose_logging ); |
255 |
close OUT; |
256 |
return $num_exported; |
257 |
} |
258 |
|
259 |
sub generate_deleted_marc_records { |
260 |
my ($record_type, $entries, $directory, $as_xml) = @_; |
261 |
my $verbose_logging = (C4::Context->preference('ZebraqueueVerboseLogging')) ? 1 : 0; |
262 |
|
263 |
my $num_exported = 0; |
264 |
open (OUT, ">:utf8 ", "$directory/exported_records") or die $!; |
265 |
my $i = 0; |
266 |
foreach my $record_number (map { $_->{biblio_auth_number} } @$entries ) { |
267 |
print "\r$i" unless ($i++ %100 or !$verbose_logging); |
268 |
print "." if ( $verbose_logging ); |
269 |
|
270 |
my $marc = MARC::Record->new(); |
271 |
if ($record_type eq 'biblio') { |
272 |
fix_biblio_ids($marc, $record_number, $record_number); |
273 |
} else { |
274 |
fix_authority_id($marc, $record_number); |
275 |
} |
276 |
if (C4::Context->preference("marcflavour") eq "UNIMARC") { |
277 |
fix_unimarc_100($marc); |
278 |
} |
279 |
|
280 |
print OUT ($as_xml) ? $marc->as_xml_record() : $marc->as_usmarc(); |
281 |
$num_exported++; |
282 |
} |
283 |
print "\nRecords exported: $num_exported\n" if ( $verbose_logging ); |
284 |
close OUT; |
285 |
return $num_exported; |
286 |
} |
287 |
|
288 |
sub get_corrected_marc_record { |
289 |
my ($record_type, $record_number, $noxml) = @_; |
290 |
|
291 |
my $marc = get_raw_marc_record($record_type, $record_number, $noxml); |
292 |
|
293 |
if (defined $marc) { |
294 |
fix_leader($marc); |
295 |
if ($record_type eq 'biblio') { |
296 |
my $succeeded = fix_biblio_ids($marc, $record_number); |
297 |
return unless $succeeded; |
298 |
} else { |
299 |
fix_authority_id($marc, $record_number); |
300 |
} |
301 |
if (C4::Context->preference("marcflavour") eq "UNIMARC") { |
302 |
fix_unimarc_100($marc); |
303 |
} |
304 |
} |
305 |
|
306 |
return $marc; |
307 |
} |
308 |
|
309 |
sub get_raw_marc_record { |
310 |
my ($record_type, $record_number, $noxml) = @_; |
311 |
my $dbh = C4::Context->dbh; |
312 |
|
313 |
my $marc; |
314 |
if ($record_type eq 'biblio') { |
315 |
if ($noxml) { |
316 |
my $fetch_sth = $dbh->prepare_cached("SELECT marc FROM biblioitems WHERE biblionumber = ?"); |
317 |
$fetch_sth->execute($record_number); |
318 |
if (my ($blob) = $fetch_sth->fetchrow_array) { |
319 |
$marc = MARC::Record->new_from_usmarc($blob); |
320 |
$fetch_sth->finish(); |
321 |
} else { |
322 |
return; # failure to find a bib is not a problem - |
323 |
# a delete could have been done before |
324 |
# trying to process a record update |
325 |
} |
326 |
} else { |
327 |
eval { $marc = GetMarcBiblio($record_number); }; |
328 |
if ($@) { |
329 |
# here we do warn since catching an exception |
330 |
# means that the bib was found but failed |
331 |
# to be parsed |
332 |
warn "error retrieving biblio $record_number"; |
333 |
return; |
334 |
} |
335 |
} |
336 |
} else { |
337 |
eval { $marc = GetAuthority($record_number); }; |
338 |
if ($@) { |
339 |
warn "error retrieving authority $record_number"; |
340 |
return; |
341 |
} |
342 |
} |
343 |
return $marc; |
344 |
} |
345 |
|
346 |
sub fix_leader { |
347 |
# FIXME - this routine is suspect |
348 |
# It blanks the Leader/00-05 and Leader/12-16 to |
349 |
# force them to be recalculated correct when |
350 |
# the $marc->as_usmarc() or $marc->as_xml() is called. |
351 |
# But why is this necessary? It would be a serious bug |
352 |
# in MARC::Record (definitely) and MARC::File::XML (arguably) |
353 |
# if they are emitting incorrect leader values. |
354 |
my $marc = shift; |
355 |
|
356 |
my $leader = $marc->leader; |
357 |
substr($leader, 0, 5) = ' '; |
358 |
substr($leader, 10, 7) = '22 '; |
359 |
$marc->leader(substr($leader, 0, 24)); |
360 |
} |
361 |
|
362 |
sub fix_biblio_ids { |
363 |
# FIXME - it is essential to ensure that the biblionumber is present, |
364 |
# otherwise, Zebra will choke on the record. However, this |
365 |
# logic belongs in the relevant C4::Biblio APIs. |
366 |
my $marc = shift; |
367 |
my $biblionumber = shift; |
368 |
my $biblioitemnumber; |
369 |
my $dbh = C4::Context->dbh; |
370 |
|
371 |
if (@_) { |
372 |
$biblioitemnumber = shift; |
373 |
} else { |
374 |
my $sth = $dbh->prepare( |
375 |
"SELECT biblioitemnumber FROM biblioitems WHERE biblionumber=?"); |
376 |
$sth->execute($biblionumber); |
377 |
($biblioitemnumber) = $sth->fetchrow_array; |
378 |
$sth->finish; |
379 |
unless ($biblioitemnumber) { |
380 |
warn "failed to get biblioitemnumber for biblio $biblionumber"; |
381 |
return 0; |
382 |
} |
383 |
} |
384 |
|
385 |
# FIXME - this is cheating on two levels |
386 |
# 1. C4::Biblio::_koha_marc_update_bib_ids is meant to be an internal function |
387 |
# 2. Making sure that the biblionumber and biblioitemnumber are correct and |
388 |
# present in the MARC::Record object ought to be part of GetMarcBiblio. |
389 |
# |
390 |
# On the other hand, this better for now than what rebuild_zebra.pl used to |
391 |
# do, which was duplicate the code for inserting the biblionumber |
392 |
# and biblioitemnumber |
393 |
C4::Biblio::_koha_marc_update_bib_ids($marc, '', $biblionumber, $biblioitemnumber); |
394 |
|
395 |
return 1; |
396 |
} |
397 |
|
398 |
sub fix_authority_id { |
399 |
# FIXME - as with fix_biblio_ids, the authid must be present |
400 |
# for Zebra's sake. However, this really belongs |
401 |
# in C4::AuthoritiesMarc. |
402 |
my ($marc, $authid) = @_; |
403 |
unless ($marc->field('001') and $marc->field('001')->data() eq $authid){ |
404 |
$marc->delete_field($marc->field('001')); |
405 |
$marc->insert_fields_ordered(MARC::Field->new('001',$authid)); |
406 |
} |
407 |
} |
408 |
|
409 |
sub fix_unimarc_100 { |
410 |
# FIXME - again, if this is necessary, it belongs in C4::AuthoritiesMarc. |
411 |
my $marc = shift; |
412 |
|
413 |
my $string; |
414 |
if ( length($marc->subfield( 100, "a" )) == 35 ) { |
415 |
$string = $marc->subfield( 100, "a" ); |
416 |
my $f100 = $marc->field(100); |
417 |
$marc->delete_field($f100); |
418 |
} |
419 |
else { |
420 |
$string = POSIX::strftime( "%Y%m%d", localtime ); |
421 |
$string =~ s/\-//g; |
422 |
$string = sprintf( "%-*s", 35, $string ); |
423 |
} |
424 |
substr( $string, 22, 6, "frey50" ); |
425 |
unless ( length($marc->subfield( 100, "a" )) == 35 ) { |
426 |
$marc->delete_field($marc->field(100)); |
427 |
$marc->insert_grouped_field(MARC::Field->new( 100, "", "", "a" => $string )); |
428 |
} |
429 |
} |
430 |
|
431 |
=head2 DoIndexing |
432 |
|
433 |
( $error_code ) = &DoIndexing($record_type,$op,$record_dir,$record_format); |
434 |
|
435 |
returns the corresponding zebraidx error code |
436 |
|
437 |
Comment : |
438 |
$record_type can be 'biblio' or 'authority' |
439 |
$zaction can be 'delete' or 'update' |
440 |
$record_dir is the directory where the exported records are |
441 |
$record_format can be 'marcxml' or 'iso2709' |
442 |
|
443 |
=cut |
444 |
|
445 |
sub DoIndexing { |
446 |
my ($record_type, $zaction, $record_dir, $record_format) = @_; |
447 |
my $zebra_server = ($record_type eq 'biblio') ? 'biblioserver' : 'authorityserver'; |
448 |
my $zebra_db_name = ($record_type eq 'biblio') ? 'biblios' : 'authorities'; |
449 |
my $zebra_config = C4::Context->zebraconfig($zebra_server)->{'config'}; |
450 |
my $zebra_db_dir = C4::Context->zebraconfig($zebra_server)->{'directory'}; |
451 |
my $noshadow = (C4::Context->preference('ZebraNoshadow')) ? '-n' : ''; |
452 |
my $zebraidx_log_opt = " -v none,fatal "; |
453 |
|
454 |
# TODO better error handling!! |
455 |
system("zebraidx -c $zebra_config $zebraidx_log_opt $noshadow -g $record_format -d $zebra_db_name $zaction $record_dir"); |
456 |
system("zebraidx -c $zebra_config $zebraidx_log_opt -g $record_format -d $zebra_db_name commit") unless $noshadow; |
457 |
# /TODO |
458 |
|
459 |
return 1; |
460 |
} |
461 |
|
462 |
|
463 |
END { } |
464 |
|
465 |
1; |
466 |
__END__ |
467 |
|
468 |
=head1 AUTHOR |
469 |
|
470 |
Koha Development Team <http://koha-community.org/> |
471 |
|
472 |
Tomas Cohen Arazi tomascohen@gmail.com |
473 |
|
474 |
=cut |