source: main/trunk/greenstone2/perllib/inexport.pm@ 23402

Last change on this file since 23402 was 23402, checked in by max, 13 years ago

New "psuedo" constructor added that is useful to call from CGI Perl scripts
  • Property svn:executable set to *
File size: 36.9 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 = 'inexport::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
128# Simplified version of the contstructor for use with CGI scripts
129sub newCGI
130{
131 my $class = shift (@_);
132 my ($mode,$collect) = @_;
133
134 my $self = { 'xml' => 0, 'mode' => $mode };
135
136 $self->{'out'} = STDERR;
137 $self->{'site'} = "";
138 $self->{'collectdir'} = &util::filename_cat($ENV{'GSDLHOME'},"collect");
139 $self->{'faillog'} = "";
140
141 $self->{'collection'} = $collect;
142
143 return bless $self, $class;
144}
145sub get_collection
146{
147 my $self = shift @_;
148
149 return $self->{'collection'};
150}
151
152
153sub read_collection_cfg
154{
155 my $self = shift @_;
156 my ($collection,$options) = @_;
157
158 my $collectdir = $self->{'collectdir'};
159 my $site = $self->{'site'};
160 my $out = $self->{'out'};
161
162 if (($collection = &colcfg::use_collection($site, $collection, $collectdir)) eq "") {
163 &PrintUsage::print_txt_usage($options, "{import.params}");
164 die "\n";
165 }
166
167 # add collection's perllib dir into include path in
168 # case we have collection specific modules
169 unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
170
171 # check that we can open the faillog
172 my $faillog = $self->{'faillog'};
173 if ($faillog eq "") {
174 $faillog = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
175 }
176 open (FAILLOG, ">$faillog") ||
177 (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
178
179
180 my $faillogname = $faillog;
181 $faillog = 'inexport::FAILLOG';
182 $faillog->autoflush(1);
183 $self->{'faillog'} = $faillog;
184 $self->{'faillogname'} = $faillogname;
185
186 # Read in the collection configuration file.
187 my ($config_filename, $gs_mode) = &colcfg::get_collect_cfg_name($out);
188 my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
189
190 return ($config_filename,$collectcfg);
191}
192
193sub set_collection_options
194{
195 my $self = shift @_;
196 my ($collectcfg) = @_;
197
198 my $inexport_mode = $self->{'mode'};
199
200 my $verbosity = $self->{'verbosity'};
201 my $debug = $self->{'debug'};
202 my $importdir = $self->{'importdir'};
203 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'} || "";
204 my $out = $self->{'out'};
205
206 # If the infodbtype value wasn't defined in the collect.cfg file, use the default
207 if (!defined($collectcfg->{'infodbtype'}))
208 {
209 $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
210 }
211 if ($collectcfg->{'infodbtype'} eq "gdbm-txtgz") {
212 # we can't use the text version for archives dbs.
213 $collectcfg->{'infodbtype'} = "gdbm";
214 }
215 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
216 $importdir = $collectcfg->{'importdir'};
217 }
218 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
219 $archivedir = $collectcfg->{'archivedir'};
220 }
221 # fill in the default import and archives directories if none
222 # were supplied, turn all \ into / and remove trailing /
223 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
224 $importdir =~ s/[\\\/]+/\//g;
225 $importdir =~ s/\/$//;
226 if (!-e $importdir) {
227 &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
228 die "\n";
229 }
230 $self->{'importdir'} = $importdir;
231
232 if ($archivedir eq "") {
233 if ($inexport_mode eq "import") {
234 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
235 }
236 elsif ($inexport_mode eq "export") {
237 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "export");
238 }
239 else {
240 print STDERR "Warning: Unrecognized import/export mode '$inexport_mode'\n";
241 print STDERR " Defaulting to 'archives' for file output\n";
242 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
243 }
244 }
245
246 $archivedir =~ s/[\\\/]+/\//g;
247 $archivedir =~ s/\/$//;
248 $self->{'archivedir'} = $archivedir;
249
250 if ($verbosity !~ /\d+/) {
251 if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
252 $verbosity = $collectcfg->{'verbosity'};
253 } else {
254 $verbosity = 2; # the default
255 }
256 }
257 $self->{'verbosity'} = $verbosity;
258
259 if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
260 $self->{'manifest'} = $collectcfg->{'manifest'};
261 }
262
263 if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
264 if ($collectcfg->{'gzip'} =~ /^true$/i) {
265 $self->{'gzip'} = 1;
266 }
267 }
268
269 if ($self->{'maxdocs'} !~ /\-?\d+/) {
270 if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
271 $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
272 } else {
273 $self->{'maxdocs'} = -1; # the default
274 }
275 }
276
277 if ((defined $self->{'groupsize'}) && ($self->{'groupsize'} == 1)) {
278 if (defined $collectcfg->{'groupsize'} && $collectcfg->{'groupsize'} =~ /\d+/) {
279 $self->{'groupsize'} = $collectcfg->{'groupsize'};
280 }
281 }
282
283 if (!defined $self->{'OIDtype'}
284 || ($self->{'OIDtype'} !~ /^(hash|incremental|assigned|dirname)$/ )) {
285 if (defined $collectcfg->{'OIDtype'}
286 && $collectcfg->{'OIDtype'} =~ /^(hash|incremental|assigned|dirname)$/) {
287 $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
288 } else {
289 $self->{'OIDtype'} = "hash"; # the default
290 }
291 }
292
293 if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
294 if (defined $collectcfg->{'OIDmetadata'}) {
295 $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
296 } else {
297 $self->{'OIDmetadata'} = "dc.Identifier"; # the default
298 }
299 }
300
301 my $sortmeta = $self->{'sortmeta'};
302 if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
303 $sortmeta = $collectcfg->{'sortmeta'};
304 }
305 # sortmeta cannot be used with group size
306 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
307 if (defined $sortmeta && $self->{'groupsize'} > 1) {
308 &gsprintf($out, "{import.cannot_sort}\n\n");
309 $sortmeta = undef;
310 }
311 $self->{'sortmeta'} = $sortmeta;
312
313 if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
314 $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
315 }
316
317 if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
318 $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
319 }
320 if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
321 $self->{'debug'} = 1;
322 }
323 if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
324 $self->{'gli'} = 1;
325 }
326 $self->{'gli'} = 0 unless defined $self->{'gli'};
327
328 # check keepold and removeold
329 my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
330
331 my ($removeold, $keepold, $incremental, $incremental_mode)
332 = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
333 $self->{'incremental'}, $checkdir,
334 $collectcfg);
335
336 $self->{'removeold'} = $removeold;
337 $self->{'keepold'} = $keepold;
338 $self->{'incremental'} = $incremental;
339 $self->{'incremental_mode'} = $incremental_mode;
340}
341
342sub process_files
343{
344 my $self = shift @_;
345 my ($config_filename,$collectcfg) = @_;
346
347 my $inexport_mode = $self->{'mode'};
348
349 my $verbosity = $self->{'verbosity'};
350 my $debug = $self->{'debug'};
351
352 my $importdir = $self->{'importdir'};
353 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
354
355 my $incremental = $self->{'incremental'};
356 my $incremental_mode = $self->{'incremental_mode'};
357
358 my $removeold = $self->{'removeold'};
359 my $keepold = $self->{'keepold'};
360
361 my $saveas = $self->{'saveas'};
362 my $OIDtype = $self->{'OIDtype'};
363 my $OIDmetadata = $self->{'OIDmetadata'};
364
365 my $out = $self->{'out'};
366 my $faillog = $self->{'faillog'};
367
368 my $maxdocs = $self->{'maxdocs'};
369 my $gzip = $self->{'gzip'};
370 my $groupsize = $self->{'groupsize'};
371 my $sortmeta = $self->{'sortmeta'};
372
373 my $removeprefix = $self->{'removeprefix'};
374 my $removesuffix = $self->{'removesuffix'};
375
376 my $gli = $self->{'gli'};
377
378 my $jobs = $self->{'jobs'};
379 my $epoch = $self->{'epoch'};
380
381 # related to export
382 my $xsltfile = $self->{'xsltfile'};
383 my $group_marc = $self->{'group_marc'};
384 my $mapping_file = $self->{'mapping_file'};
385 my $xslt_mets = $self->{'xslt_mets'};
386 my $xslt_txt = $self->{'xslt_txt'};
387 my $fedora_namespace = $self->{'fedora_namespace'};
388
389 if ($inexport_mode eq "import") {
390 print STDERR "<Import>\n" if $gli;
391 }
392 else {
393 print STDERR "<export>\n" if $gli;
394 }
395
396 my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
397 if ($self->{'manifest'} ne "") {
398 my $manifest_filename = $self->{'manifest'};
399
400 if (!&util::filename_is_absolute($manifest_filename)) {
401 $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
402 }
403
404 $self->{'manifest'} =~ s/[\\\/]+/\//g;
405 $self->{'manifest'} =~ s/\/$//;
406
407 $manifest_lookup->parse($manifest_filename);
408 }
409
410 my $manifest = $self->{'manifest'};
411
412 # load all the plugins
413 my $plugins = [];
414 if (defined $collectcfg->{'plugin'}) {
415 $plugins = $collectcfg->{'plugin'};
416 }
417
418 my $plugin_incr_mode = $incremental_mode;
419 if ($manifest ne "") {
420 # if we have a manifest file, then we pretend we are fully incremental for plugins
421 $plugin_incr_mode = "all";
422 }
423 #some global options for the plugins
424 my @global_opts = ();
425
426 my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode);
427 if (scalar(@$pluginfo) == 0) {
428 &gsprintf($out, "{import.no_plugins_loaded}\n");
429 die "\n";
430 }
431
432 # remove the old contents of the archives directory (and tmp directory) if needed
433 if ($removeold) {
434 if (-e $archivedir) {
435 &gsprintf($out, "{import.removing_archives}\n");
436 &util::rm_r ($archivedir);
437 }
438 my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
439 $tmpdir =~ s/[\\\/]+/\//g;
440 $tmpdir =~ s/\/$//;
441 if (-e $tmpdir) {
442 &gsprintf($out, "{import.removing_tmpdir}\n");
443 &util::rm_r ($tmpdir);
444 }
445 }
446
447 # create the archives dir if needed
448 &util::mk_all_dir($archivedir);
449
450 # read the archive information file
451
452 # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
453 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
454 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
455
456 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir);
457 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
458
459 my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
460 $archive_info->load_info ($arcinfo_doc_filename);
461
462 if ($manifest eq "") {
463 # Load in list of files in import folder from last import (if present)
464 $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
465 }
466
467 ####Use Plugout####
468 my $plugout;
469
470 if ($inexport_mode eq "import") {
471 if (defined $collectcfg->{'plugout'}) {
472 # If a plugout was specified in the collect.cfg file, assume it is sensible
473 # We can't check the name because it could be anything, if it is a custom plugout
474 $plugout = $collectcfg->{'plugout'};
475 }
476 else{
477 if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
478 push @$plugout,"GreenstoneXMLPlugout";
479 }
480 else{
481 push @$plugout,$saveas."Plugout";
482 }
483 }
484 }
485 else {
486 if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
487 $plugout = $collectcfg->{'plugout'};
488 }
489 else{
490 if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
491 push @$plugout,"GreenstoneMETSPlugout";
492 }
493 else{
494 push @$plugout,$saveas."Plugout";
495 }
496 }
497 }
498
499 my $plugout_name = $plugout->[0];
500
501 push @$plugout,("-output_info",$archive_info) if (defined $archive_info);
502 push @$plugout,("-verbosity",$verbosity) if (defined $verbosity);
503 push @$plugout,("-debug") if ($debug);
504 push @$plugout,("-group_size",$groupsize) if (defined $groupsize);
505 push @$plugout,("-gzip_output") if ($gzip);
506 push @$plugout,("-output_handle",$out) if (defined $out);
507
508 push @$plugout,("-xslt_file",$xsltfile) if (defined $xsltfile && $xsltfile ne "");
509
510 if ($plugout_name =~ m/^MARCXMLPlugout$/) {
511 push @$plugout,("-group") if ($group_marc);
512 push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
513 }
514 if ($plugout_name =~ m/^.*METSPlugout$/) {
515 push @$plugout,("-xslt_mets",$xslt_mets) if (defined $xslt_mets && $xslt_mets ne "");
516 push @$plugout,("-xslt_txt",$xslt_txt) if (defined $xslt_txt && $xslt_txt ne "");
517 }
518
519 if ($plugout_name eq "FedoraMETSPlugout") {
520 push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
521 }
522
523
524 my $processor = &plugout::load_plugout($plugout);
525 $processor->setoutputdir ($archivedir);
526 $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
527 $processor->set_OIDtype ($OIDtype, $OIDmetadata);
528
529 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
530
531 if ($removeold) {
532 # occasionally, plugins may want to do something on remove old, eg pharos image indexing
533 &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
534 }
535
536 # process the import directory
537 my $block_hash = {};
538 $block_hash->{'new_files'} = {};
539 $block_hash->{'reindex_files'} = {};
540 my $metadata = {};
541
542 # gobal blocking pass may set up some metadata
543 &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
544
545 if ($manifest ne "") {
546 #
547 # 1. Process delete files first
548 #
549 my @deleted_files = keys %{$manifest_lookup->{'delete'}};
550 my @full_deleted_files = ();
551
552 # ensure all filenames are absolute
553 foreach my $df (@deleted_files) {
554 my $full_df =
555 (&util::filename_is_absolute($df))
556 ? $df
557 : &util::filename_cat($importdir,$df);
558
559 if (-d $full_df) {
560 &add_dir_contents_to_list($full_df, \@full_deleted_files);
561 } else {
562 push(@full_deleted_files,$full_df);
563 }
564 }
565
566 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
567 mark_docs_for_deletion($archive_info,{},
568 \@full_deleted_files,
569 $archivedir, $verbosity, "delete");
570
571
572 #
573 # 2. Now files for reindexing
574 #
575
576 my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
577 my @full_reindex_files = ();
578 # ensure all filenames are absolute
579 foreach my $rf (@reindex_files) {
580 my $full_rf =
581 (&util::filename_is_absolute($rf))
582 ? $rf
583 : &util::filename_cat($importdir,$rf);
584
585 if (-d $full_rf) {
586 &add_dir_contents_to_list($full_rf, \@full_reindex_files);
587 } else {
588 push(@full_reindex_files,$full_rf);
589 }
590 }
591
592 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
593 mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
594
595 # And now to ensure the new version of the file processed by
596 # appropriate plugin, we need to add it to block_hash reindex list
597 foreach my $full_rf (@full_reindex_files) {
598 $block_hash->{'reindex_files'}->{$full_rf} = 1;
599 }
600
601
602 #
603 # 3. Now finally any new files - add to block_hash new_files list
604 #
605
606 my @new_files = keys %{$manifest_lookup->{'index'}};
607 my @full_new_files = ();
608
609 foreach my $nf (@new_files) {
610 # ensure filename is absolute
611 my $full_nf =
612 (&util::filename_is_absolute($nf))
613 ? $nf
614 : &util::filename_cat($importdir,$nf);
615
616 if (-d $full_nf) {
617 &add_dir_contents_to_list($full_nf, \@full_new_files);
618 } else {
619 push(@full_new_files,$full_nf);
620 }
621 }
622
623 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
624 my $arcinfodb_map = {};
625 &dbutil::read_infodb_file($collectcfg->{'infodbtype'}, $arcinfo_src_filename, $arcinfodb_map);
626 foreach my $f (@full_new_files) {
627 # check that we haven't seen it already
628 if (defined $arcinfodb_map->{$f}) {
629 # TODO make better warning
630 print STDERR "Warning: $f already in src archive, \n";
631 } else {
632 $block_hash->{'new_files'}->{$f} = 1;
633 }
634 }
635
636 undef $arcinfodb_map;
637 }
638 else {
639 # if incremental, we read through the import folder to see whats changed.
640
641 if ($incremental || $incremental_mode eq "onlyadd") {
642 prime_doc_oid_count($archivedir);
643
644 # Can now work out which files were new, already existed, and have
645 # been deleted
646
647 new_vs_old_import_diff($archive_info,$block_hash,$importdir,
648 $archivedir,$verbosity,$incremental_mode);
649
650 my @new_files = sort keys %{$block_hash->{'new_files'}};
651 if (scalar(@new_files>0)) {
652 print STDERR "New files and modified metadata files since last import:\n ";
653 print STDERR join("\n ",@new_files), "\n";
654 }
655
656 if ($incremental) {
657 # only look for deletions if we are truely incremental
658 my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
659 # Filter out any in gsdl/tmp area
660 my @filtered_deleted_files = ();
661 my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
662 my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
663 $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
664 $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
665
666 foreach my $df (@deleted_files) {
667 next if ($df =~ m/^$gsdl_tmp_area/);
668 next if ($df =~ m/^$collect_tmp_area/);
669
670 push(@filtered_deleted_files,$df);
671 }
672
673
674 @deleted_files = @filtered_deleted_files;
675
676 if (scalar(@deleted_files)>0) {
677 print STDERR "Files deleted since last import:\n ";
678 print STDERR join("\n ",@deleted_files), "\n";
679
680
681 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
682
683 mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
684 }
685
686 my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
687
688 if (scalar(@reindex_files)>0) {
689 print STDERR "Files to reindex since last import:\n ";
690 print STDERR join("\n ",@reindex_files), "\n";
691 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
692 mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
693 }
694
695 }
696 }
697 }
698
699 # now, whichever mode we are in, we can process the entire import folder
700 if ((defined $jobs) && ($jobs > 1))
701 {
702 # if jobs are set to >1, run in parallel using MPI helper
703 # [hs, 1 july 2010]
704 &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
705 $self->{'collection'}, $self->{'site'});
706 }
707 else
708 {
709 &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
710 }
711
712
713 if ($saveas eq "FedoraMETS") {
714 # create collection "doc obj" for Fedora that contains
715 # collection-level metadata
716
717 my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
718 $doc_obj->set_OID("collection");
719
720 my $col_name = undef;
721 my $col_meta = $collectcfg->{'collectionmeta'};
722
723 if (defined $col_meta) {
724 store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
725 store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description
726 }
727 $processor->process($doc_obj);
728 }
729
730 &plugin::end($pluginfo, $processor);
731
732 &plugin::deinit($pluginfo, $processor);
733
734 # Store the value of OIDCount (used in doc.pm) so it can be
735 # restored correctly to this value on an incremental build
736 store_doc_oid_count($archivedir);
737
738 # write out the archive information file
739 $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
740 $processor->close_group_output() if $processor->is_group();
741
742 # for backwards compatability with archvies.inf file
743 if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
744 $archive_info->save_info($arcinfo_doc_filename);
745 }
746 else {
747 $archive_info->save_revinfo_db($arcinfo_src_filename);
748 }
749
750 return $pluginfo;
751}
752
753
754sub generate_statistics
755{
756 my $self = shift @_;
757 my ($pluginfo) = @_;
758
759 my $inexport_mode = $self->{'mode'};
760
761 my $statsfile = $self->{'statsfile'};
762 my $out = $self->{'out'};
763 my $faillogname = $self->{'faillogname'};
764 my $gli = $self->{'gli'};
765 my $jobs = $self->{'jobs'};
766
767 # write out import stats
768
769 if ((!defined $jobs) || ($jobs == 1))
770 {
771 # only output statistics if there are multiple jobs
772 # [hs, 1 july 2010]
773
774 my $close_stats = 0;
775 if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
776 if (open (STATS, ">$statsfile")) {
777 $statsfile = 'inexport::STATS';
778 $close_stats = 1;
779 } else {
780 &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
781 &gsprintf($out, "{import.stats_backup}\n");
782 $statsfile = 'STDERR';
783 }
784 }
785
786 &gsprintf($out, "\n");
787 &gsprintf($out, "*********************************************\n");
788 &gsprintf($out, "{$inexport_mode.complete}\n");
789 &gsprintf($out, "*********************************************\n");
790
791 &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
792 if ($close_stats) {
793 close STATS;
794 }
795 }
796
797 close OUT if $self->{'close_out'};
798 close FAILLOG;
799}
800
801
802sub store_collectionmeta
803{
804 my ($collectionmeta,$field,$doc_obj) = @_;
805
806 my $section = $doc_obj->get_top_section();
807
808 my $field_hash = $collectionmeta->{$field};
809
810 foreach my $k (keys %$field_hash)
811 {
812 my $val = $field_hash->{$k};
813
814 ### print STDERR "*** $k = $field_hash->{$k}\n";
815
816 my $md_label = "ex.$field";
817
818
819 if ($k =~ m/^\[l=(.*?)\]$/)
820 {
821
822 my $md_suffix = $1;
823 $md_label .= "^$md_suffix";
824 }
825
826
827 $doc_obj->add_utf8_metadata($section,$md_label, $val);
828
829 # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
830 # while "collectionname" in GS2 is called "name" in GS3.
831 # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
832 if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
833 {
834 $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
835 }
836
837 }
838}
839
840
841sub oid_count_file {
842 my ($archivedir) = @_;
843 return &util::filename_cat ($archivedir, "OIDcount");
844}
845
846
847sub prime_doc_oid_count
848{
849 my ($archivedir) = @_;
850 my $oid_count_filename = &oid_count_file($archivedir);
851
852 if (-e $oid_count_filename) {
853 if (open(OIDIN,"<$oid_count_filename")) {
854 my $OIDcount = <OIDIN>;
855 chomp $OIDcount;
856 close(OIDIN);
857
858 $doc::OIDcount = $OIDcount;
859 }
860 else {
861
862 print STDERR "Warning: unable to read document OID count from $oid_count_filename\n";
863 print STDERR "Setting value to 0\n";
864 }
865 }
866
867}
868
869sub store_doc_oid_count
870{
871 # Use the file "OIDcount" in the archives directory to record
872 # what value doc.pm got up to
873
874 my ($archivedir) = @_;
875 my $oid_count_filename = &oid_count_file($archivedir);
876
877
878 if (open(OIDOUT,">$oid_count_filename")) {
879 print OIDOUT $doc::OIDcount, "\n";
880
881 close(OIDOUT);
882 }
883 else {
884 print STDERR "Warning: unable to store document OID count\n";
885 }
886}
887
888
889
890sub new_vs_old_import_diff
891{
892 my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
893
894 # Get the infodbtype value for this collection from the arcinfo object
895 my $infodbtype = $archive_info->{'infodbtype'};
896
897 # in this method, we want to know if metadata files are modified or not.
898 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
899
900 my $archiveinf_timestamp = -M $arcinfo_doc_filename;
901
902 # First convert all files to absolute form
903 # This is to support the situation where the import folder is not
904 # the default
905
906 my $prev_all_files = $archive_info->{'prev_import_filelist'};
907 my $full_prev_all_files = {};
908
909 foreach my $prev_file (keys %$prev_all_files) {
910
911 if (!&util::filename_is_absolute($prev_file)) {
912 my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
913 $full_prev_all_files->{$full_prev_file} = $prev_file;
914 }
915 else {
916 $full_prev_all_files->{$prev_file} = $prev_file;
917 }
918 }
919
920
921 # Figure out which are the new files, existing files and so
922 # by implication the files from the previous import that are not
923 # there any more => mark them for deletion
924 foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
925
926 my $full_curr_file = $curr_file;
927
928 # entry in 'all_files' is moved to either 'existing_files',
929 # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
930
931 if (!&util::filename_is_absolute($curr_file)) {
932 # add in import dir to make absolute
933 $full_curr_file = &util::filename_cat($importdir,$curr_file);
934 }
935
936 # figure out if new file or not
937 if (defined $full_prev_all_files->{$full_curr_file}) {
938 # delete it so that only files that need deleting are left
939 delete $full_prev_all_files->{$full_curr_file};
940
941 # had it before. is it a metadata file?
942 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
943
944 # is it modified??
945 if (-M $full_curr_file < $archiveinf_timestamp) {
946 print STDERR "*** Detected a modified metadata file: $full_curr_file\n" if $verbosity > 2;
947 # its newer than last build
948 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
949 }
950 }
951 else {
952 if ($incremental_mode eq "all") {
953
954 # had it before
955 $block_hash->{'existing_files'}->{$full_curr_file} = 1;
956
957 }
958 else {
959 # Warning in "onlyadd" mode, but had it before!
960 print STDERR "Warning: File $full_curr_file previously imported.\n";
961 print STDERR " Treating as new file\n";
962
963 $block_hash->{'new_files'}->{$full_curr_file} = 1;
964
965 }
966 }
967 }
968 else {
969 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
970 # the new file is the special sort of file greenstone uses
971 # to attach metadata to src documents
972 # i.e metadata.xml
973 # (but note, the filename used is not constrained in
974 # Greenstone to always be this)
975
976 print STDERR "***** Detected new metadata file: $full_curr_file\n" if $verbosity > 2;
977 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
978 }
979 else {
980 $block_hash->{'new_files'}->{$full_curr_file} = 1;
981 }
982 }
983
984
985 delete $block_hash->{'all_files'}->{$curr_file};
986 }
987
988
989
990
991 # Deal with complication of new or modified metadata files by forcing
992 # everything from this point down in the file hierarchy to
993 # be freshly imported.
994 #
995 # This may mean files that have not changed are reindexed, but does
996 # guarantee by the end of processing all new metadata is correctly
997 # associated with the relevant document(s).
998
999 foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
1000 my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
1001
1002 $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
1003 $situated_dir =~ s/\\/\\\\/g; # need to protect windows slash \ in regular expression
1004
1005 # Go through existing_files, and mark anything that is contained
1006 # within 'situated_dir' to be reindexed (in case some of the metadata
1007 # attaches to one of these files)
1008
1009 my $reindex_files = [];
1010
1011 foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
1012
1013 if ($existing_f =~ m/^$situated_dir/) {
1014 push(@$reindex_files,$existing_f);
1015 $block_hash->{'reindex_files'}->{$existing_f} = 1;
1016 delete $block_hash->{'existing_files'}->{$existing_f};
1017
1018 }
1019 }
1020
1021 # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
1022 # (or equivalent)
1023 $block_hash->{'new_files'}->{$new_mdf} = 1;
1024
1025 }
1026
1027 # go through remaining existing files and work out what has changed and needs to be reindexed.
1028 my @existing_files = sort keys %{$block_hash->{'existing_files'}};
1029
1030 my $reindex_files = [];
1031
1032 foreach my $existing_filename (@existing_files) {
1033 if (-M $existing_filename < $archiveinf_timestamp) {
1034 # file is newer than last build
1035
1036 my $existing_file = $existing_filename;
1037 #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
1038
1039 #my $collectdir_resafe = &util::filename_to_regex($collectdir);
1040 #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
1041
1042 print STDERR "**** Reindexing existing file: $existing_file\n";
1043
1044 push(@$reindex_files,$existing_file);
1045 $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1046 }
1047
1048 }
1049
1050
1051 # By this point full_prev_all_files contains the files
1052 # mentioned in archiveinf-src.db but are not in the 'import'
1053 # folder (or whatever was specified through -importdir ...)
1054
1055 # This list can contain files that were created in the 'tmp' or
1056 # 'cache' areas (such as screen-size and thumbnail images).
1057 #
1058 # In building the final list of files to delete, we test to see if
1059 # it exists on the filesystem and if it does (unusual for a "normal"
1060 # file in import, but possible in the case of 'tmp' files),
1061 # supress it from going into the final list
1062
1063 my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1064
1065 my @deleted_files = values %$full_prev_all_files;
1066 map { my $curr_file = $_;
1067 my $full_curr_file = $curr_file;
1068
1069 if (!&util::filename_is_absolute($curr_file)) {
1070 # add in import dir to make absolute
1071
1072 $full_curr_file = &util::filename_cat($collectdir,$curr_file);
1073 }
1074
1075
1076 if (!-e $full_curr_file) {
1077 $block_hash->{'deleted_files'}->{$curr_file} = 1;
1078 }
1079 } @deleted_files;
1080
1081
1082
1083}
1084
1085
1086# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1087# $mode is 'delete' or 'reindex'
1088sub mark_docs_for_deletion
1089{
1090 my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1091
1092 my $mode_text = "deleted from index";
1093 if ($mode eq "reindex") {
1094 $mode_text = "reindexed";
1095 }
1096
1097 # Get the infodbtype value for this collection from the arcinfo object
1098 my $infodbtype = $archive_info->{'infodbtype'};
1099
1100 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1101 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1102
1103
1104 # record files marked for deletion in arcinfo
1105 foreach my $file (@$deleted_files) {
1106 # use 'archiveinf-src' info database file to look up all the OIDs
1107 # that this file is used in (note in most cases, it's just one OID)
1108
1109 my $src_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
1110 my $src_rec = &dbutil::convert_infodb_string_to_hash($src_rec_string);
1111 my $oids = $src_rec->{'oid'};
1112 my $file_record_deleted = 0;
1113
1114 # delete the src record
1115 my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1116 &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
1117 &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1118
1119
1120 foreach my $oid (@$oids) {
1121
1122 # find the source doc (the primary file that becomes this oid)
1123 my $doc_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1124 my $doc_rec = &dbutil::convert_infodb_string_to_hash($doc_rec_string);
1125 my $doc_source_file = $doc_rec->{'src-file'}->[0];
1126 if (!&util::filename_is_absolute($doc_source_file)) {
1127 $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1128 }
1129
1130 if ($doc_source_file ne $file) {
1131 # its an associated or metadata file
1132
1133 # mark source doc for reimport as one of its assoc files has changed or deleted
1134 $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1135
1136 }
1137 my $curr_status = $archive_info->get_status_info($oid);
1138 if (defined($curr_status) && (($curr_status ne "D"))) {
1139 if ($verbosity>1) {
1140 print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1141 }
1142 # mark oid for deletion (it will be deleted or reimported)
1143 $archive_info->set_status_info($oid,"D");
1144 my $val = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1145 $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1146
1147 my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1148 my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1149
1150 &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1151 &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1152 }
1153 }
1154
1155 }
1156 # now go through and check that we haven't marked any primary files for reindex (because their associated files have changed/deleted) when they have been deleted themselves. only in delete mode.
1157 if ($mode eq "delete") {
1158 foreach my $file (@$deleted_files) {
1159 if (defined $block_hash->{'reindex_files'}->{$file}) {
1160 delete $block_hash->{'reindex_files'}->{$file};
1161 }
1162 }
1163 }
1164
1165
1166}
1167
1168sub add_dir_contents_to_list {
1169
1170 my ($dirname, $list) = @_;
1171
1172 # Recur over directory contents.
1173 my (@dir, $subfile);
1174
1175 # find all the files in the directory
1176 if (!opendir (DIR, $dirname)) {
1177 print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
1178 return -1; # error in processing
1179 }
1180 @dir = readdir (DIR);
1181 closedir (DIR);
1182
1183 for (my $i = 0; $i < scalar(@dir); $i++) {
1184 my $subfile = $dir[$i];
1185 next if ($subfile =~ m/^\.\.?$/);
1186 next if ($subfile =~ /^\.svn$/);
1187 my $full_file = &util::filename_cat($dirname, $subfile);
1188 if (-d $full_file) {
1189 &add_dir_contents_to_list($full_file, $list);
1190 } else {
1191 push (@$list, $full_file);
1192 }
1193 }
1194
1195}
1196
1197
11981;
Note: See TracBrowser for help on using the repository browser.