source: main/trunk/greenstone2/perllib/inexport.pm@ 22986

Last change on this file since 22986 was 22986, checked in by kjdon, 14 years ago

need to do a file_block_read before a read when adding new docs through a manifest, in case we have been given a directory. Otherwise, eg for html file, get warnings about all the mages, or else they will be incorporated into the collection as individual docs.

  • Property svn:executable set to *
File size: 35.0 KB
Line 
1###########################################################################
2#
3# inexport.pm -- useful class to support import.pl and export.pl
4# A component of the Greenstone digital library software
5# from the New Zealand Digital Library Project at the
6# University of Waikato, New Zealand.
7#
8# Copyright (C) 1999 New Zealand Digital Library Project
9#
10# This program is free software; you can redistribute it and/or modify
11# it under the terms of the GNU General Public License as published by
12# the Free Software Foundation; either version 2 of the License, or
13# (at your option) any later version.
14#
15# This program is distributed in the hope that it will be useful,
16# but WITHOUT ANY WARRANTY; without even the implied warranty of
17# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18# GNU General Public License for more details.
19#
20# You should have received a copy of the GNU General Public License
21# along with this program; if not, write to the Free Software
22# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
23#
24###########################################################################
25
26package inexport;
27
28use strict;
29
30no strict 'refs'; # allow filehandles to be variables and vice versa
31no strict 'subs'; # allow barewords (eg STDERR) as function arguments
32
33use arcinfo;
34use colcfg;
35use dbutil;
36use doc;
37use plugin;
38use plugout;
39use manifest;
40use inexport;
41use util;
42use scriptutil;
43use FileHandle;
44use gsprintf 'gsprintf';
45use printusage;
46use parse2;
47
48use File::Basename;
49
50sub new
51{
52 my $class = shift (@_);
53 my ($mode,$argv,$options,$opt_listall_options) = @_;
54
55 my $self = { 'xml' => 0, 'mode' => $mode };
56
57 # general options available to all plugins
58 my $arguments = $options->{'args'};
59 my $intArgLeftinAfterParsing = parse2::parse($argv,$arguments,$self,"allow_extra_options");
60 # Parse returns -1 if something has gone wrong
61 if ($intArgLeftinAfterParsing == -1)
62 {
63 &PrintUsage::print_txt_usage($options, "{import.params}");
64 die "\n";
65 }
66
67 my $language = $self->{'language'};
68 # If $language has been specified, load the appropriate resource bundle
69 # (Otherwise, the default resource bundle will be loaded automatically)
70 if ($language && $language =~ /\S/) {
71 &gsprintf::load_language_specific_resource_bundle($language);
72 }
73
74 if ($self->{'listall'}) {
75 if ($self->{'xml'}) {
76 &PrintUsage::print_xml_usage($opt_listall_options);
77 }
78 else
79 {
80 &PrintUsage::print_txt_usage($opt_listall_options,"{export.params}");
81 }
82 die "\n";
83 }
84
85
86 if ($self->{'xml'}) {
87 &PrintUsage::print_xml_usage($options);
88 print "\n";
89 return bless $self, $class;
90 }
91
92 if ($self->{'gli'}) { # the gli wants strings to be in UTF-8
93 &gsprintf::output_strings_in_UTF8;
94 }
95
96 # now check that we had exactly one leftover arg, which should be
97 # the collection name. We don't want to do this earlier, cos
98 # -xml arg doesn't need a collection name
99 # Or if the user specified -h, then we output the usage also
100
101 if ($intArgLeftinAfterParsing != 1 || (@$argv && $argv->[0] =~ /^\-+h/))
102 {
103 &PrintUsage::print_txt_usage($options, "{import.params}");
104 die "\n";
105 }
106
107 $self->{'close_out'} = 0;
108 my $out = $self->{'out'};
109 if ($out !~ /^(STDERR|STDOUT)$/i) {
110 open (OUT, ">$out") ||
111 (&gsprintf(STDERR, "{common.cannot_open_output_file}: $!\n", $out) && die);
112 $out = 'import::OUT';
113 $self->{'close_out'} = 1;
114 }
115 $out->autoflush(1);
116 $self->{'out'} = $out;
117
118 # @ARGV should be only one item, the name of the collection
119 $self->{'collection'} = shift @$argv;
120
121 if ((defined $self->{'jobs'}) && ($self->{'jobs'}>1)) {
122 require ParallelInexport;
123 }
124
125 return bless $self, $class;
126}
127
128sub get_collection
129{
130 my $self = shift @_;
131
132 return $self->{'collection'};
133}
134
135
136sub read_collection_cfg
137{
138 my $self = shift @_;
139 my ($collection,$options) = @_;
140
141 my $collectdir = $self->{'collectdir'};
142 my $site = $self->{'site'};
143 my $out = $self->{'out'};
144
145 if (($collection = &colcfg::use_collection($site, $collection, $collectdir)) eq "") {
146 &PrintUsage::print_txt_usage($options, "{import.params}");
147 die "\n";
148 }
149
150 # add collection's perllib dir into include path in
151 # case we have collection specific modules
152 unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
153
154 # check that we can open the faillog
155 my $faillog = $self->{'faillog'};
156 if ($faillog eq "") {
157 $faillog = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
158 }
159 open (FAILLOG, ">$faillog") ||
160 (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
161
162
163 my $faillogname = $faillog;
164 $faillog = 'inexport::FAILLOG';
165 $faillog->autoflush(1);
166 $self->{'faillog'} = $faillog;
167 $self->{'faillogname'} = $faillogname;
168
169 # Read in the collection configuration file.
170 my ($config_filename, $gs_mode) = &colcfg::get_collect_cfg_name($out);
171 my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
172
173 return ($config_filename,$collectcfg);
174}
175
176sub set_collection_options
177{
178 my $self = shift @_;
179 my ($collectcfg) = @_;
180
181 my $inexport_mode = $self->{'mode'};
182
183 my $verbosity = $self->{'verbosity'};
184 my $debug = $self->{'debug'};
185 my $importdir = $self->{'importdir'};
186 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'} || "";
187 my $out = $self->{'out'};
188
189 # If the infodbtype value wasn't defined in the collect.cfg file, use the default
190 if (!defined($collectcfg->{'infodbtype'}))
191 {
192 $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
193 }
194
195 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
196 $importdir = $collectcfg->{'importdir'};
197 }
198 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
199 $archivedir = $collectcfg->{'archivedir'};
200 }
201 # fill in the default import and archives directories if none
202 # were supplied, turn all \ into / and remove trailing /
203 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
204 $importdir =~ s/[\\\/]+/\//g;
205 $importdir =~ s/\/$//;
206 if (!-e $importdir) {
207 &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
208 die "\n";
209 }
210 $self->{'importdir'} = $importdir;
211
212 if ($archivedir eq "") {
213 if ($inexport_mode eq "import") {
214 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
215 }
216 elsif ($inexport_mode eq "export") {
217 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "export");
218 }
219 else {
220 print STDERR "Warning: Unrecognized import/export mode '$inexport_mode'\n";
221 print STDERR " Defaulting to 'archives' for file output\n";
222 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
223 }
224 }
225
226 $archivedir =~ s/[\\\/]+/\//g;
227 $archivedir =~ s/\/$//;
228 $self->{'archivedir'} = $archivedir;
229
230 if ($verbosity !~ /\d+/) {
231 if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
232 $verbosity = $collectcfg->{'verbosity'};
233 } else {
234 $verbosity = 2; # the default
235 }
236 }
237 $self->{'verbosity'} = $verbosity;
238
239 if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
240 $self->{'manifest'} = $collectcfg->{'manifest'};
241 }
242
243 if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
244 if ($collectcfg->{'gzip'} =~ /^true$/i) {
245 $self->{'gzip'} = 1;
246 }
247 }
248
249 if ($self->{'maxdocs'} !~ /\-?\d+/) {
250 if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
251 $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
252 } else {
253 $self->{'maxdocs'} = -1; # the default
254 }
255 }
256
257 if ((defined $self->{'groupsize'}) && ($self->{'groupsize'} == 1)) {
258 if (defined $collectcfg->{'groupsize'} && $collectcfg->{'groupsize'} =~ /\d+/) {
259 $self->{'groupsize'} = $collectcfg->{'groupsize'};
260 }
261 }
262
263 if (!defined $self->{'OIDtype'}
264 || ($self->{'OIDtype'} !~ /^(hash|incremental|assigned|dirname)$/ )) {
265 if (defined $collectcfg->{'OIDtype'}
266 && $collectcfg->{'OIDtype'} =~ /^(hash|incremental|assigned|dirname)$/) {
267 $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
268 } else {
269 $self->{'OIDtype'} = "hash"; # the default
270 }
271 }
272
273 if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
274 if (defined $collectcfg->{'OIDmetadata'}) {
275 $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
276 } else {
277 $self->{'OIDmetadata'} = "dc.Identifier"; # the default
278 }
279 }
280
281 my $sortmeta = $self->{'sortmeta'};
282 if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
283 $sortmeta = $collectcfg->{'sortmeta'};
284 }
285 # sortmeta cannot be used with group size
286 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
287 if (defined $sortmeta && $self->{'groupsize'} > 1) {
288 &gsprintf($out, "{import.cannot_sort}\n\n");
289 $sortmeta = undef;
290 }
291 $self->{'sortmeta'} = $sortmeta;
292
293 if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
294 $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
295 }
296
297 if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
298 $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
299 }
300 if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
301 $self->{'debug'} = 1;
302 }
303 if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
304 $self->{'gli'} = 1;
305 }
306 $self->{'gli'} = 0 unless defined $self->{'gli'};
307
308 # check keepold and removeold
309 my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
310
311 my ($removeold, $keepold, $incremental, $incremental_mode)
312 = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
313 $self->{'incremental'}, $checkdir,
314 $collectcfg);
315
316 $self->{'removeold'} = $removeold;
317 $self->{'keepold'} = $keepold;
318 $self->{'incremental'} = $incremental;
319 $self->{'incremental_mode'} = $incremental_mode;
320}
321
322sub process_files
323{
324 my $self = shift @_;
325 my ($config_filename,$collectcfg) = @_;
326
327 my $inexport_mode = $self->{'mode'};
328
329 my $verbosity = $self->{'verbosity'};
330 my $debug = $self->{'debug'};
331
332 my $importdir = $self->{'importdir'};
333 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
334
335 my $incremental = $self->{'incremental'};
336 my $incremental_mode = $self->{'incremental_mode'};
337
338 my $removeold = $self->{'removeold'};
339 my $keepold = $self->{'keepold'};
340
341 my $saveas = $self->{'saveas'};
342 my $OIDtype = $self->{'OIDtype'};
343 my $OIDmetadata = $self->{'OIDmetadata'};
344
345 my $out = $self->{'out'};
346 my $faillog = $self->{'faillog'};
347
348 my $maxdocs = $self->{'maxdocs'};
349 my $gzip = $self->{'gzip'};
350 my $groupsize = $self->{'groupsize'};
351 my $sortmeta = $self->{'sortmeta'};
352
353 my $removeprefix = $self->{'removeprefix'};
354 my $removesuffix = $self->{'removesuffix'};
355
356 my $gli = $self->{'gli'};
357
358 my $jobs = $self->{'jobs'};
359 my $epoch = $self->{'epoch'};
360
361 # related to export
362 my $xsltfile = $self->{'xsltfile'};
363 my $group_marc = $self->{'group_marc'};
364 my $mapping_file = $self->{'mapping_file'};
365 my $xslt_mets = $self->{'xslt_mets'};
366 my $xslt_txt = $self->{'xslt_txt'};
367 my $fedora_namespace = $self->{'fedora_namespace'};
368
369 if ($inexport_mode eq "import") {
370 print STDERR "<Import>\n" if $gli;
371 }
372 else {
373 print STDERR "<export>\n" if $gli;
374 }
375
376 my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
377 if ($self->{'manifest'} ne "") {
378 my $manifest_filename = $self->{'manifest'};
379
380 if (!&util::filename_is_absolute($manifest_filename)) {
381 $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
382 }
383
384 $self->{'manifest'} =~ s/[\\\/]+/\//g;
385 $self->{'manifest'} =~ s/\/$//;
386
387 $manifest_lookup->parse($manifest_filename);
388 }
389
390 my $manifest = $self->{'manifest'};
391
392 # load all the plugins
393 my $plugins = [];
394 if (defined $collectcfg->{'plugin'}) {
395 $plugins = $collectcfg->{'plugin'};
396 }
397
398 #some global options for the plugins
399 my @global_opts = ();
400
401 my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $incremental_mode);
402 if (scalar(@$pluginfo) == 0) {
403 &gsprintf($out, "{import.no_plugins_loaded}\n");
404 die "\n";
405 }
406
407 # remove the old contents of the archives directory (and tmp directory) if needed
408 if ($removeold) {
409 if (-e $archivedir) {
410 &gsprintf($out, "{import.removing_archives}\n");
411 &util::rm_r ($archivedir);
412 }
413 my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
414 $tmpdir =~ s/[\\\/]+/\//g;
415 $tmpdir =~ s/\/$//;
416 if (-e $tmpdir) {
417 &gsprintf($out, "{import.removing_tmpdir}\n");
418 &util::rm_r ($tmpdir);
419 }
420 }
421
422 # create the archives dir if needed
423 &util::mk_all_dir($archivedir);
424
425 # read the archive information file
426## my $arcinfo_doc_filename = &util::filename_cat ($archivedir, "archives.inf");
427
428 # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
429 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
430 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
431
432 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir);
433 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
434
435 my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
436 $archive_info->load_info ($arcinfo_doc_filename);
437
438 if ($manifest eq "") {
439 # Load in list of files in import folder from last import (if present)
440 $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
441 }
442
443 ####Use Plugout####
444 my $plugout;
445
446 if ($inexport_mode eq "import") {
447 if (defined $collectcfg->{'plugout'}) {
448 # If a plugout was specified in the collect.cfg file, assume it is sensible
449 # We can't check the name because it could be anything, if it is a custom plugout
450 $plugout = $collectcfg->{'plugout'};
451 }
452 else{
453 if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
454 push @$plugout,"GreenstoneXMLPlugout";
455 }
456 else{
457 push @$plugout,$saveas."Plugout";
458 }
459 }
460 }
461 else {
462 if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
463 $plugout = $collectcfg->{'plugout'};
464 }
465 else{
466 if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
467 push @$plugout,"GreenstoneMETSPlugout";
468 }
469 else{
470 push @$plugout,$saveas."Plugout";
471 }
472 }
473 }
474
475 my $plugout_name = $plugout->[0];
476
477 push @$plugout,("-output_info",$archive_info) if (defined $archive_info);
478 push @$plugout,("-verbosity",$verbosity) if (defined $verbosity);
479 push @$plugout,("-debug") if ($debug);
480 push @$plugout,("-group_size",$groupsize) if (defined $groupsize);
481 push @$plugout,("-gzip_output") if ($gzip);
482 push @$plugout,("-output_handle",$out) if (defined $out);
483
484 push @$plugout,("-xslt_file",$xsltfile) if (defined $xsltfile && $xsltfile ne "");
485
486 if ($plugout_name =~ m/^MARCXMLPlugout$/) {
487 push @$plugout,("-group") if ($group_marc);
488 push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
489 }
490 if ($plugout_name =~ m/^.*METSPlugout$/) {
491 push @$plugout,("-xslt_mets",$xslt_mets) if (defined $xslt_mets && $xslt_mets ne "");
492 push @$plugout,("-xslt_txt",$xslt_txt) if (defined $xslt_txt && $xslt_txt ne "");
493 }
494
495 if ($plugout_name eq "FedoraMETSPlugout") {
496 push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
497 }
498
499
500 my $processor = &plugout::load_plugout($plugout);
501 $processor->setoutputdir ($archivedir);
502 $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
503 $processor->set_OIDtype ($OIDtype, $OIDmetadata);
504
505 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
506
507 if ($removeold) {
508 # occasionally, plugins may want to do something on remove old, eg pharos image indexing
509 &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
510 }
511 if ($manifest eq "") {
512 # process the import directory
513 my $block_hash = {};
514 my $metadata = {};
515 # gobal blocking pass may set up some metadata
516 &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
517
518 if ($incremental || $incremental_mode eq "onlyadd") {
519
520 prime_doc_oid_count($archivedir);
521
522 # Can now work out which files were new, already existed, and have
523 # been deleted
524
525 new_vs_old_import_diff($archive_info,$block_hash,$importdir,
526 $archivedir,$verbosity,$incremental_mode);
527
528 my @new_files = sort keys %{$block_hash->{'new_files'}};
529 if (scalar(@new_files>0)) {
530 print STDERR "New files and modified metadata files since last import:\n ";
531 print STDERR join("\n ",@new_files), "\n";
532 }
533
534 if ($incremental) {
535 # only look for deletions if we are truely incremental
536 my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
537 # Filter out any in gsdl/tmp area
538 my @filtered_deleted_files = ();
539 my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
540 my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
541 $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
542 $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
543
544 foreach my $df (@deleted_files) {
545 next if ($df =~ m/^$gsdl_tmp_area/);
546 next if ($df =~ m/^$collect_tmp_area/);
547
548 push(@filtered_deleted_files,$df);
549 }
550
551
552 @deleted_files = @filtered_deleted_files;
553
554 if (scalar(@deleted_files)>0) {
555 print STDERR "Files deleted since last import:\n ";
556 print STDERR join("\n ",@deleted_files), "\n";
557
558
559 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
560
561 mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
562 }
563
564 my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
565
566 if (scalar(@reindex_files)>0) {
567 print STDERR "Files to reindex since last import:\n ";
568 print STDERR join("\n ",@reindex_files), "\n";
569 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
570 mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
571 }
572
573 }
574
575 # Play it safe, and run through the entire folder, only processing new or edited files
576
577 if ((defined $jobs) && ($jobs > 1))
578 {
579 # if jobs are set to >1, run in parallel using MPI helper
580 # [hs, 1 july 2010]
581 &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
582 $self->{'collection'}, $self->{'site'});
583 }
584 else
585 {
586 &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
587 }
588 }
589 else {
590 if ((defined $jobs) && ($jobs > 1))
591 {
592 # if jobs are set to >1, run in parallel using MPI helper
593 # [hs, 1 july 2010]
594 &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
595 $self->{'collection'}, $self->{'site'});
596 }
597 else
598 {
599 &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
600 }
601 }
602
603 }
604 else
605 {
606 #
607 # 1. Process delete files first
608 #
609
610 my @deleted_files = keys %{$manifest_lookup->{'delete'}};
611 my @full_deleted_files = ();
612
613 # ensure all filenames are absolute
614 foreach my $df (@deleted_files) {
615 my $full_df =
616 (&util::filename_is_absolute($df))
617 ? $df
618 : &util::filename_cat($importdir,$df);
619
620 push(@full_deleted_files,$full_df);
621 }
622
623 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
624 mark_docs_for_deletion($archive_info,{},
625 \@full_deleted_files,
626 $archivedir, $verbosity, "delete");
627
628
629 #
630 # 2. Now files for reindexing
631 #
632
633 my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
634 my @full_reindex_files = ();
635
636 # ensure all filenames are absolute
637 foreach my $rf (@reindex_files) {
638 my $full_rf =
639 (&util::filename_is_absolute($rf))
640 ? $rf
641 : &util::filename_cat($importdir,$rf);
642
643 push(@full_reindex_files,$full_rf);
644 }
645
646 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
647 mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
648
649 # And now ensure the new version of the file processed by appropriate
650 # plugin
651 foreach my $full_rf (@full_reindex_files) {
652 &plugin::read ($pluginfo, "", $full_rf, {}, {}, $processor, $maxdocs, 0, $gli);
653 }
654
655
656 #
657 # 3. Now finally any new files
658 #
659
660 foreach my $file (keys %{$manifest_lookup->{'index'}}) {
661 my $block_hash = {};
662 my $metadata = {};
663 &plugin::file_block_read($pluginfo, $importdir, $file, $block_hash, $metadata, $gli);
664 &plugin::read ($pluginfo, $importdir, $file, $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
665 }
666
667
668 }
669
670 if ($saveas eq "FedoraMETS") {
671 # create collection "doc obj" for Fedora that contains
672 # collection-level metadata
673
674 my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
675 $doc_obj->set_OID("collection");
676
677 my $col_name = undef;
678 my $col_meta = $collectcfg->{'collectionmeta'};
679
680 if (defined $col_meta) {
681 store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
682 store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description
683 }
684 $processor->process($doc_obj);
685 }
686
687 &plugin::end($pluginfo, $processor);
688
689 &plugin::deinit($pluginfo, $processor);
690
691 # Store the value of OIDCount (used in doc.pm) so it can be
692 # restored correctly to this value on an incremental build
693 store_doc_oid_count($archivedir);
694
695 # write out the archive information file
696 $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
697 $processor->close_group_output() if $processor->is_group();
698
699 # for backwards compatability with archvies.inf file
700 if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
701 $archive_info->save_info($arcinfo_doc_filename);
702 }
703 else {
704 $archive_info->save_revinfo_db($arcinfo_src_filename);
705 }
706
707 return $pluginfo;
708}
709
710
711sub generate_statistics
712{
713 my $self = shift @_;
714 my ($pluginfo) = @_;
715
716 my $inexport_mode = $self->{'mode'};
717
718 my $statsfile = $self->{'statsfile'};
719 my $out = $self->{'out'};
720 my $faillogname = $self->{'faillogname'};
721 my $gli = $self->{'gli'};
722 my $jobs = $self->{'jobs'};
723
724 # write out import stats
725
726 if ((!defined $jobs) || ($jobs == 1))
727 {
728 # only output statistics if there are multiple jobs
729 # [hs, 1 july 2010]
730
731 my $close_stats = 0;
732 if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
733 if (open (STATS, ">$statsfile")) {
734 $statsfile = 'import::STATS';
735 $close_stats = 1;
736 } else {
737 &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
738 &gsprintf($out, "{import.stats_backup}\n");
739 $statsfile = 'STDERR';
740 }
741 }
742
743 &gsprintf($out, "\n");
744 &gsprintf($out, "*********************************************\n");
745 &gsprintf($out, "{$inexport_mode.complete}\n");
746 &gsprintf($out, "*********************************************\n");
747
748 &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
749 if ($close_stats) {
750 close STATS;
751 }
752 }
753
754 close OUT if $self->{'close_out'};
755 close FAILLOG;
756}
757
758
759sub store_collectionmeta
760{
761 my ($collectionmeta,$field,$doc_obj) = @_;
762
763 my $section = $doc_obj->get_top_section();
764
765 my $field_hash = $collectionmeta->{$field};
766
767 foreach my $k (keys %$field_hash)
768 {
769 my $val = $field_hash->{$k};
770
771 ### print STDERR "*** $k = $field_hash->{$k}\n";
772
773 my $md_label = "ex.$field";
774
775
776 if ($k =~ m/^\[l=(.*?)\]$/)
777 {
778
779 my $md_suffix = $1;
780 $md_label .= "^$md_suffix";
781 }
782
783
784 $doc_obj->add_utf8_metadata($section,$md_label, $val);
785
786 # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
787 # while "collectionname" in GS2 is called "name" in GS3.
788 # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
789 if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
790 {
791 $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
792 }
793
794 }
795}
796
797
798sub oid_count_file {
799 my ($archivedir) = @_;
800 return &util::filename_cat ($archivedir, "OIDcount");
801}
802
803
804sub prime_doc_oid_count
805{
806 my ($archivedir) = @_;
807 my $oid_count_filename = &oid_count_file($archivedir);
808
809 if (-e $oid_count_filename) {
810 if (open(OIDIN,"<$oid_count_filename")) {
811 my $OIDcount = <OIDIN>;
812 chomp $OIDcount;
813 close(OIDIN);
814
815 $doc::OIDcount = $OIDcount;
816 }
817 else {
818
819 print STDERR "Warning: unable to read document OID count from $oid_count_filename\n";
820 print STDERR "Setting value to 0\n";
821 }
822 }
823
824}
825
826sub store_doc_oid_count
827{
828 # Use the file "OIDcount" in the archives directory to record
829 # what value doc.pm got up to
830
831 my ($archivedir) = @_;
832 my $oid_count_filename = &oid_count_file($archivedir);
833
834
835 if (open(OIDOUT,">$oid_count_filename")) {
836 print OIDOUT $doc::OIDcount, "\n";
837
838 close(OIDOUT);
839 }
840 else {
841 print STDERR "Warning: unable to store document OID count\n";
842 }
843}
844
845
846
847sub new_vs_old_import_diff
848{
849 my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
850
851 # Get the infodbtype value for this collection from the arcinfo object
852 my $infodbtype = $archive_info->{'infodbtype'};
853
854 # in this method, we want to know if metadata files are modified or not.
855 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
856
857 my $archiveinf_timestamp = -M $arcinfo_doc_filename;
858
859 # First convert all files to absolute form
860 # This is to support the situation where the import folder is not
861 # the default
862
863 my $prev_all_files = $archive_info->{'prev_import_filelist'};
864 my $full_prev_all_files = {};
865
866 foreach my $prev_file (keys %$prev_all_files) {
867
868 if (!&util::filename_is_absolute($prev_file)) {
869 my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
870 $full_prev_all_files->{$full_prev_file} = $prev_file;
871 }
872 else {
873 $full_prev_all_files->{$prev_file} = $prev_file;
874 }
875 }
876
877
878 # Figure out which are the new files, existing files and so
879 # by implication the files from the previous import that are not
880 # there any more => mark them for deletion
881 foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
882
883 my $full_curr_file = $curr_file;
884
885 # entry in 'all_files' is moved to either 'existing_files',
886 # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
887
888 if (!&util::filename_is_absolute($curr_file)) {
889 # add in import dir to make absolute
890 $full_curr_file = &util::filename_cat($importdir,$curr_file);
891 }
892
893 # figure out if new file or not
894 if (defined $full_prev_all_files->{$full_curr_file}) {
895 # delete it so that only files that need deleting are left
896 delete $full_prev_all_files->{$full_curr_file};
897
898 # had it before. is it a metadata file?
899 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
900
901 # is it modified??
902 if (-M $full_curr_file < $archiveinf_timestamp) {
903 print STDERR "*** Detected a modified metadata file: $full_curr_file\n" if $verbosity > 2;
904 # its newer than last build
905 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
906 }
907 }
908 else {
909 if ($incremental_mode eq "all") {
910
911 # had it before
912 $block_hash->{'existing_files'}->{$full_curr_file} = 1;
913
914 }
915 else {
916 # Warning in "onlyadd" mode, but had it before!
917 print STDERR "Warning: File $full_curr_file previously imported.\n";
918 print STDERR " Treating as new file\n";
919
920 $block_hash->{'new_files'}->{$full_curr_file} = 1;
921
922 }
923 }
924 }
925 else {
926 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
927 # the new file is the special sort of file greenstone uses
928 # to attach metadata to src documents
929 # i.e metadata.xml
930 # (but note, the filename used is not constrained in
931 # Greenstone to always be this)
932
933 print STDERR "***** Detected new metadata file: $full_curr_file\n" if $verbosity > 2;
934 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
935 }
936 else {
937 $block_hash->{'new_files'}->{$full_curr_file} = 1;
938 }
939 }
940
941
942 delete $block_hash->{'all_files'}->{$curr_file};
943 }
944
945
946
947
948 # Deal with complication of new or modified metadata files by forcing
949 # everything from this point down in the file hierarchy to
950 # be freshly imported.
951 #
952 # This may mean files that have not changed are reindexed, but does
953 # guarantee by the end of processing all new metadata is correctly
954 # associated with the relevant document(s).
955
956 foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
957 my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
958
959 $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
960 $situated_dir =~ s/\\/\\\\/g; # need to protect windows slash \ in regular expression
961
962 # Go through existing_files, and mark anything that is contained
963 # within 'situated_dir' to be reindexed (in case some of the metadata
964 # attaches to one of these files)
965
966 my $reindex_files = [];
967
968 foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
969
970 if ($existing_f =~ m/^$situated_dir/) {
971 push(@$reindex_files,$existing_f);
972 $block_hash->{'reindex_files'}->{$existing_f} = 1;
973 delete $block_hash->{'existing_files'}->{$existing_f};
974
975 }
976 }
977
978 # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
979 # (or equivalent)
980 $block_hash->{'new_files'}->{$new_mdf} = 1;
981
982 }
983
984 # go through remaining existing files and work out what has changed and needs to be reindexed.
985 my @existing_files = sort keys %{$block_hash->{'existing_files'}};
986
987 my $reindex_files = [];
988
989 foreach my $existing_filename (@existing_files) {
990 if (-M $existing_filename < $archiveinf_timestamp) {
991 # file is newer than last build
992
993 my $existing_file = $existing_filename;
994 #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
995
996 #my $collectdir_resafe = &util::filename_to_regex($collectdir);
997 #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
998
999 print STDERR "**** Reindexing existing file: $existing_file\n";
1000
1001 push(@$reindex_files,$existing_file);
1002 $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1003 }
1004
1005 }
1006
1007
1008 # By this point full_prev_all_files contains the files
1009 # mentioned in archiveinf-src.db but are not in the 'import'
1010 # folder (or whatever was specified through -importdir ...)
1011
1012 # This list can contain files that were created in the 'tmp' or
1013 # 'cache' areas (such as screen-size and thumbnail images).
1014 #
1015 # In building the final list of files to delete, we test to see if
1016 # it exists on the filesystem and if it does (unusual for a "normal"
1017 # file in import, but possible in the case of 'tmp' files),
1018 # supress it from going into the final list
1019
1020 my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1021
1022 my @deleted_files = values %$full_prev_all_files;
1023 map { my $curr_file = $_;
1024 my $full_curr_file = $curr_file;
1025
1026 if (!&util::filename_is_absolute($curr_file)) {
1027 # add in import dir to make absolute
1028
1029 $full_curr_file = &util::filename_cat($collectdir,$curr_file);
1030 }
1031
1032
1033 if (!-e $full_curr_file) {
1034 $block_hash->{'deleted_files'}->{$curr_file} = 1;
1035 }
1036 } @deleted_files;
1037
1038
1039
1040}
1041
1042
1043# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1044# $mode is 'delete' or 'reindex'
1045sub mark_docs_for_deletion
1046{
1047 my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1048
1049 my $mode_text = "deleted from index";
1050 if ($mode eq "reindex") {
1051 $mode_text = "reindexed";
1052 }
1053
1054 # Get the infodbtype value for this collection from the arcinfo object
1055 my $infodbtype = $archive_info->{'infodbtype'};
1056
1057 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1058 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1059
1060
1061 # record files marked for deletion in arcinfo
1062 foreach my $file (@$deleted_files) {
1063 # use 'archiveinf-src' info database file to look up all the OIDs
1064 # that this file is used in (note in most cases, it's just one OID)
1065
1066 my $src_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
1067 my $src_rec = &dbutil::convert_infodb_string_to_hash($src_rec_string);
1068 my $oids = $src_rec->{'oid'};
1069 my $file_record_deleted = 0;
1070
1071 # delete the src record
1072 my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1073 &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
1074 &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1075
1076
1077 foreach my $oid (@$oids) {
1078
1079 # find the source doc (the primary file that becomes this oid)
1080 my $doc_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1081 my $doc_rec = &dbutil::convert_infodb_string_to_hash($doc_rec_string);
1082 my $doc_source_file = $doc_rec->{'src-file'}->[0];
1083 if (!&util::filename_is_absolute($doc_source_file)) {
1084 $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1085 }
1086
1087 if ($doc_source_file ne $file) {
1088 # its an associated or metadata file
1089
1090 # mark source doc for reimport as one of its assoc files has changed or deleted
1091 $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1092
1093 }
1094 my $curr_status = $archive_info->get_status_info($oid);
1095 if (defined($curr_status) && (($curr_status ne "D"))) {
1096 if ($verbosity>1) {
1097 print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1098 }
1099 # mark oid for deletion (it will be deleted or reimported)
1100 $archive_info->set_status_info($oid,"D");
1101 my $val = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1102 $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1103
1104 my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1105 my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1106
1107 &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1108 &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1109 }
1110 }
1111
1112 }
1113 # now go through and check that we haven't marked any primary files for reindex (because their associated files have changed/deleted) when they have been deleted themselves. only in delete mode.
1114 if ($mode eq "delete") {
1115 foreach my $file (@$deleted_files) {
1116 if (defined $block_hash->{'reindex_files'}->{$file}) {
1117 delete $block_hash->{'reindex_files'}->{$file};
1118 }
1119 }
1120 }
1121
1122
1123}
1124
1125
1126
11271;
Note: See TracBrowser for help on using the repository browser.