source: main/trunk/greenstone2/perllib/inexport.pm@ 26536

Last change on this file since 26536 was 26536, checked in by davidb, 11 years ago

Introduction of two new OIDtype values (hash_on_full_filename and full_filename) designed to help provide more stable document IDs for collections that are rebuilt over time, including rebuilt after the Greenstone install has been upgraded
  • Property svn:executable set to *
File size: 38.6 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,$collect,$gsdl_cgi,$opt_site) = @_;
133
134 my $self = { 'xml' => 0, 'mode' => $mode };
135
136 $self->{'out'} = STDERR;
137
138 if (defined $gsdl_cgi) {
139 $self->{'site'} = $opt_site;
140 my $collect_dir = $gsdl_cgi->get_collection_dir($opt_site);
141 $self->{'collectdir'} = $collect_dir;
142 }
143 else {
144 $self->{'site'} = "";
145 $self->{'collectdir'} = &util::filename_cat($ENV{'GSDLHOME'},"collect");
146 }
147 $self->{'faillog'} = "";
148
149 $self->{'collection'} = $collect;
150
151 return bless $self, $class;
152}
153sub get_collection
154{
155 my $self = shift @_;
156
157 return $self->{'collection'};
158}
159
160
161sub read_collection_cfg
162{
163 my $self = shift @_;
164 my ($collection,$options) = @_;
165
166 my $collectdir = $self->{'collectdir'};
167 my $site = $self->{'site'};
168 my $out = $self->{'out'};
169
170 if (($collection = &colcfg::use_collection($site, $collection, $collectdir)) eq "") {
171 &PrintUsage::print_txt_usage($options, "{import.params}");
172 die "\n";
173 }
174
175 # set gs_verison 2/3
176 $self->{'gs_version'} = "2";
177 if ((defined $site) && ($site ne "")) {
178 # gs3
179 $self->{'gs_version'} = "3";
180 }
181 # add collection's perllib dir into include path in
182 # case we have collection specific modules
183 unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
184
185 # check that we can open the faillog
186 my $faillog = $self->{'faillog'};
187 if ($faillog eq "") {
188 $faillog = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
189 }
190 open (FAILLOG, ">$faillog") ||
191 (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
192
193
194 my $faillogname = $faillog;
195 $faillog = 'inexport::FAILLOG';
196 $faillog->autoflush(1);
197 $self->{'faillog'} = $faillog;
198 $self->{'faillogname'} = $faillogname;
199
200 # Read in the collection configuration file.
201 my ($config_filename, $gs_mode) = &colcfg::get_collect_cfg_name($out);
202 my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
203
204 return ($config_filename,$collectcfg);
205}
206
207sub set_collection_options
208{
209 my $self = shift @_;
210 my ($collectcfg) = @_;
211
212 my $inexport_mode = $self->{'mode'};
213
214 my $verbosity = $self->{'verbosity'};
215 my $debug = $self->{'debug'};
216 my $importdir = $self->{'importdir'};
217 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'} || "";
218 my $out = $self->{'out'};
219
220 # If the infodbtype value wasn't defined in the collect.cfg file, use the default
221 if (!defined($collectcfg->{'infodbtype'}))
222 {
223 $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
224 }
225 if ($collectcfg->{'infodbtype'} eq "gdbm-txtgz") {
226 # we can't use the text version for archives dbs.
227 $collectcfg->{'infodbtype'} = "gdbm";
228 }
229
230 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
231 $importdir = $collectcfg->{'importdir'};
232 }
233 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
234 $archivedir = $collectcfg->{'archivedir'};
235 }
236 # fill in the default import and archives directories if none
237 # were supplied, turn all \ into / and remove trailing /
238 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
239 $importdir =~ s/[\\\/]+/\//g;
240 $importdir =~ s/\/$//;
241 if (!-e $importdir) {
242 &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
243 die "\n";
244 }
245 $self->{'importdir'} = $importdir;
246
247 if ($archivedir eq "") {
248 if ($inexport_mode eq "import") {
249 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
250 }
251 elsif ($inexport_mode eq "export") {
252 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "export");
253 }
254 else {
255 print STDERR "Warning: Unrecognized import/export mode '$inexport_mode'\n";
256 print STDERR " Defaulting to 'archives' for file output\n";
257 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
258 }
259 }
260
261 $archivedir =~ s/[\\\/]+/\//g;
262 $archivedir =~ s/\/$//;
263 $self->{'archivedir'} = $archivedir;
264
265 if ($verbosity !~ /\d+/) {
266 if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
267 $verbosity = $collectcfg->{'verbosity'};
268 } else {
269 $verbosity = 2; # the default
270 }
271 }
272 $self->{'verbosity'} = $verbosity;
273
274 if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
275 $self->{'manifest'} = $collectcfg->{'manifest'};
276 }
277
278 if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
279 if ($collectcfg->{'gzip'} =~ /^true$/i) {
280 $self->{'gzip'} = 1;
281 }
282 }
283
284 if ($self->{'maxdocs'} !~ /\-?\d+/) {
285 if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
286 $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
287 } else {
288 $self->{'maxdocs'} = -1; # the default
289 }
290 }
291
292 if ((defined $self->{'groupsize'}) && ($self->{'groupsize'} == 1)) {
293 if (defined $collectcfg->{'groupsize'} && $collectcfg->{'groupsize'} =~ /\d+/) {
294 $self->{'groupsize'} = $collectcfg->{'groupsize'};
295 }
296 }
297
298 if (!defined $self->{'OIDtype'}
299 || ($self->{'OIDtype'} !~ /^(hash|hash_on_full_filename|incremental|assigned|dirname|full_filename)$/ )) {
300 # OIDtype was either not defined on the command-line, or if it was not one of the recognized values
301 if (defined $collectcfg->{'OIDtype'}
302 && $collectcfg->{'OIDtype'} =~ /^(hash|hash_on_full_filename|incremental|assigned|dirname|full_filename)$/) {
303 $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
304 } else {
305 $self->{'OIDtype'} = "hash"; # the default
306 }
307 }
308
309 if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
310 if (defined $collectcfg->{'OIDmetadata'}) {
311 $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
312 } else {
313 $self->{'OIDmetadata'} = "dc.Identifier"; # the default
314 }
315 }
316
317 my $sortmeta = $self->{'sortmeta'};
318 if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
319 $sortmeta = $collectcfg->{'sortmeta'};
320 }
321 # sortmeta cannot be used with group size
322 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
323 if (defined $sortmeta && $self->{'groupsize'} > 1) {
324 &gsprintf($out, "{import.cannot_sort}\n\n");
325 $sortmeta = undef;
326 }
327 $self->{'sortmeta'} = $sortmeta;
328
329 if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
330 $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
331 }
332
333 if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
334 $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
335 }
336 if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
337 $self->{'debug'} = 1;
338 }
339 if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
340 $self->{'gli'} = 1;
341 }
342 $self->{'gli'} = 0 unless defined $self->{'gli'};
343
344 # check keepold and removeold
345 my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
346
347 my ($removeold, $keepold, $incremental, $incremental_mode)
348 = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
349 $self->{'incremental'}, $checkdir,
350 $collectcfg);
351
352 $self->{'removeold'} = $removeold;
353 $self->{'keepold'} = $keepold;
354 $self->{'incremental'} = $incremental;
355 $self->{'incremental_mode'} = $incremental_mode;
356}
357
358sub process_files
359{
360 my $self = shift @_;
361 my ($config_filename,$collectcfg) = @_;
362
363 my $inexport_mode = $self->{'mode'};
364
365 my $verbosity = $self->{'verbosity'};
366 my $debug = $self->{'debug'};
367
368 my $importdir = $self->{'importdir'};
369 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
370
371 my $incremental = $self->{'incremental'};
372 my $incremental_mode = $self->{'incremental_mode'};
373
374 my $gs_version = $self->{'gs_version'};
375
376 my $removeold = $self->{'removeold'};
377 my $keepold = $self->{'keepold'};
378
379 my $saveas = $self->{'saveas'};
380 my $OIDtype = $self->{'OIDtype'};
381 my $OIDmetadata = $self->{'OIDmetadata'};
382
383 my $out = $self->{'out'};
384 my $faillog = $self->{'faillog'};
385
386 my $maxdocs = $self->{'maxdocs'};
387 my $gzip = $self->{'gzip'};
388 my $groupsize = $self->{'groupsize'};
389 my $sortmeta = $self->{'sortmeta'};
390
391 my $removeprefix = $self->{'removeprefix'};
392 my $removesuffix = $self->{'removesuffix'};
393
394 my $gli = $self->{'gli'};
395
396 my $jobs = $self->{'jobs'};
397 my $epoch = $self->{'epoch'};
398
399 # related to export
400 my $xsltfile = $self->{'xsltfile'};
401 my $group_marc = $self->{'group_marc'};
402 my $mapping_file = $self->{'mapping_file'};
403 my $xslt_mets = $self->{'xslt_mets'};
404 my $xslt_txt = $self->{'xslt_txt'};
405 my $fedora_namespace = $self->{'fedora_namespace'};
406 my $metadata_prefix = $self->{'metadata_prefix'};
407
408 if ($inexport_mode eq "import") {
409 print STDERR "<Import>\n" if $gli;
410 }
411 else {
412 print STDERR "<export>\n" if $gli;
413 }
414
415 my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
416 if ($self->{'manifest'} ne "") {
417 my $manifest_filename = $self->{'manifest'};
418
419 if (!&util::filename_is_absolute($manifest_filename)) {
420 $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
421 }
422
423 $self->{'manifest'} =~ s/[\\\/]+/\//g;
424 $self->{'manifest'} =~ s/\/$//;
425
426 $manifest_lookup->parse($manifest_filename);
427 }
428
429 my $manifest = $self->{'manifest'};
430
431 # load all the plugins
432 my $plugins = [];
433 if (defined $collectcfg->{'plugin'}) {
434 $plugins = $collectcfg->{'plugin'};
435 }
436
437 my $plugin_incr_mode = $incremental_mode;
438 if ($manifest ne "") {
439 # if we have a manifest file, then we pretend we are fully incremental for plugins
440 $plugin_incr_mode = "all";
441 }
442 #some global options for the plugins
443 my @global_opts = ();
444
445 my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode, $gs_version);
446 if (scalar(@$pluginfo) == 0) {
447 &gsprintf($out, "{import.no_plugins_loaded}\n");
448 die "\n";
449 }
450
451 # remove the old contents of the archives directory (and tmp
452 # directory) if needed
453
454 if ($removeold) {
455 if (-e $archivedir) {
456 &gsprintf($out, "{import.removing_archives}\n");
457 &util::rm_r ($archivedir);
458 }
459 my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
460 $tmpdir =~ s/[\\\/]+/\//g;
461 $tmpdir =~ s/\/$//;
462 if (-e $tmpdir) {
463 &gsprintf($out, "{import.removing_tmpdir}\n");
464 &util::rm_r ($tmpdir);
465 }
466 }
467
468 # create the archives dir if needed
469 &util::mk_all_dir($archivedir);
470
471 # read the archive information file
472
473 # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
474 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
475 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
476
477 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir);
478 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
479
480 my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
481 $archive_info->load_info ($arcinfo_doc_filename);
482
483 if ($manifest eq "") {
484 # Load in list of files in import folder from last import (if present)
485 $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
486 }
487
488 ####Use Plugout####
489 my $plugout;
490
491 if ($inexport_mode eq "import") {
492 if (defined $collectcfg->{'plugout'}) {
493 # If a plugout was specified in the collect.cfg file, assume it is sensible
494 # We can't check the name because it could be anything, if it is a custom plugout
495 $plugout = $collectcfg->{'plugout'};
496 }
497 else{
498 if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
499 push @$plugout,"GreenstoneXMLPlugout";
500 }
501 else{
502 push @$plugout,$saveas."Plugout";
503 }
504 }
505 }
506 else {
507 if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
508 $plugout = $collectcfg->{'plugout'};
509 }
510 else{
511 if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
512 push @$plugout,"GreenstoneMETSPlugout";
513 }
514 else{
515 push @$plugout,$saveas."Plugout";
516 }
517 }
518 }
519
520 my $plugout_name = $plugout->[0];
521
522 push @$plugout,("-output_info",$archive_info) if (defined $archive_info);
523 push @$plugout,("-verbosity",$verbosity) if (defined $verbosity);
524 push @$plugout,("-debug") if ($debug);
525 push @$plugout,("-group_size",$groupsize) if (defined $groupsize);
526 push @$plugout,("-gzip_output") if ($gzip);
527 push @$plugout,("-output_handle",$out) if (defined $out);
528
529 push @$plugout,("-xslt_file",$xsltfile) if (defined $xsltfile && $xsltfile ne "");
530
531 if ($plugout_name =~ m/^MARCXMLPlugout$/) {
532 push @$plugout,("-group") if ($group_marc);
533 push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
534 }
535 if ($plugout_name =~ m/^.*METSPlugout$/) {
536 push @$plugout,("-xslt_mets",$xslt_mets) if (defined $xslt_mets && $xslt_mets ne "");
537 push @$plugout,("-xslt_txt",$xslt_txt) if (defined $xslt_txt && $xslt_txt ne "");
538 }
539
540 if ($plugout_name eq "FedoraMETSPlugout") {
541 push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
542 }
543
544 if ($plugout_name eq "DSpacePlugout") {
545 push @$plugout,("-metadata_prefix",$metadata_prefix) if (defined $metadata_prefix && $metadata_prefix ne "");
546 }
547
548 my $processor = &plugout::load_plugout($plugout);
549 $processor->setoutputdir ($archivedir);
550 $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
551
552 $processor->set_OIDtype ($OIDtype, $OIDmetadata);
553
554 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
555
556 if ($removeold) {
557 # occasionally, plugins may want to do something on remove
558 # old, eg pharos image indexing
559 &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
560 }
561
562 # process the import directory
563 my $block_hash = {};
564 $block_hash->{'new_files'} = {};
565 $block_hash->{'reindex_files'} = {};
566 my $metadata = {};
567
568 # global blocking pass may set up some metadata
569 &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
570
571 if ($manifest ne "") {
572 #
573 # 1. Process delete files first
574 #
575 my @deleted_files = keys %{$manifest_lookup->{'delete'}};
576 my @full_deleted_files = ();
577
578 # ensure all filenames are absolute
579 foreach my $df (@deleted_files) {
580 my $full_df =
581 (&util::filename_is_absolute($df))
582 ? $df
583 : &util::filename_cat($importdir,$df);
584
585 if (-d $full_df) {
586 &add_dir_contents_to_list($full_df, \@full_deleted_files);
587 } else {
588 push(@full_deleted_files,$full_df);
589 }
590 }
591
592 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
593 mark_docs_for_deletion($archive_info,{},
594 \@full_deleted_files,
595 $archivedir, $verbosity, "delete");
596
597
598 #
599 # 2. Now files for reindexing
600 #
601
602 my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
603 my @full_reindex_files = ();
604 # ensure all filenames are absolute
605 foreach my $rf (@reindex_files) {
606 my $full_rf =
607 (&util::filename_is_absolute($rf))
608 ? $rf
609 : &util::filename_cat($importdir,$rf);
610
611 if (-d $full_rf) {
612 &add_dir_contents_to_list($full_rf, \@full_reindex_files);
613 } else {
614 push(@full_reindex_files,$full_rf);
615 }
616 }
617
618 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
619 mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
620
621 # And now to ensure the new version of the file processed by
622 # appropriate plugin, we need to add it to block_hash reindex list
623 foreach my $full_rf (@full_reindex_files) {
624 $block_hash->{'reindex_files'}->{$full_rf} = 1;
625 }
626
627
628 #
629 # 3. Now finally any new files - add to block_hash new_files list
630 #
631
632 my @new_files = keys %{$manifest_lookup->{'index'}};
633 my @full_new_files = ();
634
635 foreach my $nf (@new_files) {
636 # ensure filename is absolute
637 my $full_nf =
638 (&util::filename_is_absolute($nf))
639 ? $nf
640 : &util::filename_cat($importdir,$nf);
641
642 if (-d $full_nf) {
643 &add_dir_contents_to_list($full_nf, \@full_new_files);
644 } else {
645 push(@full_new_files,$full_nf);
646 }
647 }
648
649 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
650 my $arcinfodb_map = {};
651 &dbutil::read_infodb_file($collectcfg->{'infodbtype'}, $arcinfo_src_filename, $arcinfodb_map);
652 foreach my $f (@full_new_files) {
653 # check that we haven't seen it already
654 if (defined $arcinfodb_map->{$f}) {
655 # TODO make better warning
656 print STDERR "Warning: $f already in src archive, \n";
657 } else {
658 $block_hash->{'new_files'}->{$f} = 1;
659 }
660 }
661
662 undef $arcinfodb_map;
663 }
664 else {
665 # if incremental, we read through the import folder to see whats changed.
666
667 if ($incremental || $incremental_mode eq "onlyadd") {
668 prime_doc_oid_count($archivedir);
669
670 # Can now work out which files were new, already existed, and have
671 # been deleted
672
673 new_vs_old_import_diff($archive_info,$block_hash,$importdir,
674 $archivedir,$verbosity,$incremental_mode);
675
676 my @new_files = sort keys %{$block_hash->{'new_files'}};
677 if (scalar(@new_files>0)) {
678 print STDERR "New files and modified metadata files since last import:\n ";
679 print STDERR join("\n ",@new_files), "\n";
680 }
681
682 if ($incremental) {
683 # only look for deletions if we are truely incremental
684 my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
685 # Filter out any in gsdl/tmp area
686 my @filtered_deleted_files = ();
687 my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
688 my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
689 $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
690 $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
691
692 foreach my $df (@deleted_files) {
693 next if ($df =~ m/^$gsdl_tmp_area/);
694 next if ($df =~ m/^$collect_tmp_area/);
695
696 push(@filtered_deleted_files,$df);
697 }
698
699
700 @deleted_files = @filtered_deleted_files;
701
702 if (scalar(@deleted_files)>0) {
703 print STDERR "Files deleted since last import:\n ";
704 print STDERR join("\n ",@deleted_files), "\n";
705
706
707 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
708
709 mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
710 }
711
712 my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
713
714 if (scalar(@reindex_files)>0) {
715 print STDERR "Files to reindex since last import:\n ";
716 print STDERR join("\n ",@reindex_files), "\n";
717 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
718 mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
719 }
720
721 }
722 }
723 }
724
725 # Check for existence of the file that's to contain earliestDateStamp in archivesdir
726 # Do nothing if the file already exists (file exists on incremental build).
727 # If the file doesn't exist, as happens on full build, create it and write out the current datestamp into it
728 # In buildcol, read the file's contents and set the earliestdateStamp in GS2's build.cfg / GS3's buildconfig.xml
729 # In doc.pm have set_oaiLastModified similar to set_lastmodified, and create the doc fields
730 # oailastmodified and oailastmodifieddate
731 my $earliestDatestampFile = &util::filename_cat($archivedir, "earliestDatestamp");
732 if (!-f $earliestDatestampFile && -d $archivedir) {
733 my $current_time_in_seconds = time; # in seconds
734
735 if(open(FOUT, ">$earliestDatestampFile")) {
736 # || (&gsprintf(STDERR, "{common.cannot_open}: $!\n", $earliestDatestampFile) && die);
737 print FOUT $current_time_in_seconds;
738 close(FOUT);
739 }
740 else {
741 &gsprintf(STDERR, "{import.cannot_write_earliestdatestamp}\n", $earliestDatestampFile);
742 }
743
744 }
745
746 # now, whichever mode we are in, we can process the entire import folder
747 if ((defined $jobs) && ($jobs > 1))
748 {
749 # if jobs are set to >1, run in parallel using MPI helper
750 # [hs, 1 july 2010]
751 &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
752 $self->{'collection'}, $self->{'site'});
753 }
754 else
755 {
756 &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
757 }
758
759
760 if ($saveas eq "FedoraMETS") {
761 # create collection "doc obj" for Fedora that contains
762 # collection-level metadata
763
764 my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
765 $doc_obj->set_OID("collection");
766
767 my $col_name = undef;
768 my $col_meta = $collectcfg->{'collectionmeta'};
769
770 if (defined $col_meta) {
771 store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
772 store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description
773 }
774 $processor->process($doc_obj);
775 }
776
777 &plugin::end($pluginfo, $processor);
778
779 &plugin::deinit($pluginfo, $processor);
780
781 # Store the value of OIDCount (used in doc.pm) so it can be
782 # restored correctly to this value on an incremental build
783 store_doc_oid_count($archivedir);
784
785 # write out the archive information file
786 $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
787 $processor->close_group_output() if $processor->is_group();
788
789 # for backwards compatability with archvies.inf file
790 if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
791 $archive_info->save_info($arcinfo_doc_filename);
792 }
793 else {
794 $archive_info->save_revinfo_db($arcinfo_src_filename);
795 }
796
797 return $pluginfo;
798}
799
800
801sub generate_statistics
802{
803 my $self = shift @_;
804 my ($pluginfo) = @_;
805
806 my $inexport_mode = $self->{'mode'};
807
808 my $statsfile = $self->{'statsfile'};
809 my $out = $self->{'out'};
810 my $faillogname = $self->{'faillogname'};
811 my $gli = $self->{'gli'};
812 my $jobs = $self->{'jobs'};
813
814 # write out import stats
815
816 if ((!defined $jobs) || ($jobs == 1))
817 {
818 # only output statistics if there are multiple jobs
819 # [hs, 1 july 2010]
820
821 my $close_stats = 0;
822 if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
823 if (open (STATS, ">$statsfile")) {
824 $statsfile = 'inexport::STATS';
825 $close_stats = 1;
826 } else {
827 &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
828 &gsprintf($out, "{import.stats_backup}\n");
829 $statsfile = 'STDERR';
830 }
831 }
832
833 &gsprintf($out, "\n");
834 &gsprintf($out, "*********************************************\n");
835 &gsprintf($out, "{$inexport_mode.complete}\n");
836 &gsprintf($out, "*********************************************\n");
837
838 &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
839 if ($close_stats) {
840 close STATS;
841 }
842 }
843
844 close OUT if $self->{'close_out'};
845 close FAILLOG;
846}
847
848
849sub store_collectionmeta
850{
851 my ($collectionmeta,$field,$doc_obj) = @_;
852
853 my $section = $doc_obj->get_top_section();
854
855 my $field_hash = $collectionmeta->{$field};
856
857 foreach my $k (keys %$field_hash)
858 {
859 my $val = $field_hash->{$k};
860
861 ### print STDERR "*** $k = $field_hash->{$k}\n";
862
863 my $md_label = "ex.$field";
864
865
866 if ($k =~ m/^\[l=(.*?)\]$/)
867 {
868
869 my $md_suffix = $1;
870 $md_label .= "^$md_suffix";
871 }
872
873
874 $doc_obj->add_utf8_metadata($section,$md_label, $val);
875
876 # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
877 # while "collectionname" in GS2 is called "name" in GS3.
878 # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
879 if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
880 {
881 $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
882 }
883
884 }
885}
886
887
888sub oid_count_file {
889 my ($archivedir) = @_;
890 return &util::filename_cat ($archivedir, "OIDcount");
891}
892
893
894sub prime_doc_oid_count
895{
896 my ($archivedir) = @_;
897 my $oid_count_filename = &oid_count_file($archivedir);
898
899 if (-e $oid_count_filename) {
900 if (open(OIDIN,"<$oid_count_filename")) {
901 my $OIDcount = <OIDIN>;
902 chomp $OIDcount;
903 close(OIDIN);
904
905 $doc::OIDcount = $OIDcount;
906 }
907 else {
908 &gsprintf(STDERR, "{import.cannot_read_OIDcount}\n", $oid_count_filename);
909 }
910 }
911
912}
913
914sub store_doc_oid_count
915{
916 # Use the file "OIDcount" in the archives directory to record
917 # what value doc.pm got up to
918
919 my ($archivedir) = @_;
920 my $oid_count_filename = &oid_count_file($archivedir);
921
922
923 if (open(OIDOUT,">$oid_count_filename")) {
924 print OIDOUT $doc::OIDcount, "\n";
925
926 close(OIDOUT);
927 }
928 else {
929 &gsprintf(STDERR, "{import.cannot_write_OIDcount}\n", $oid_count_filename);
930 }
931}
932
933
934
935sub new_vs_old_import_diff
936{
937 my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
938
939 # Get the infodbtype value for this collection from the arcinfo object
940 my $infodbtype = $archive_info->{'infodbtype'};
941
942 # in this method, we want to know if metadata files are modified or not.
943 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
944
945 my $archiveinf_timestamp = -M $arcinfo_doc_filename;
946
947 # First convert all files to absolute form
948 # This is to support the situation where the import folder is not
949 # the default
950
951 my $prev_all_files = $archive_info->{'prev_import_filelist'};
952 my $full_prev_all_files = {};
953
954 foreach my $prev_file (keys %$prev_all_files) {
955
956 if (!&util::filename_is_absolute($prev_file)) {
957 my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
958 $full_prev_all_files->{$full_prev_file} = $prev_file;
959 }
960 else {
961 $full_prev_all_files->{$prev_file} = $prev_file;
962 }
963 }
964
965
966 # Figure out which are the new files, existing files and so
967 # by implication the files from the previous import that are not
968 # there any more => mark them for deletion
969 foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
970
971 my $full_curr_file = $curr_file;
972
973 # entry in 'all_files' is moved to either 'existing_files',
974 # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
975
976 if (!&util::filename_is_absolute($curr_file)) {
977 # add in import dir to make absolute
978 $full_curr_file = &util::filename_cat($importdir,$curr_file);
979 }
980
981 # figure out if new file or not
982 if (defined $full_prev_all_files->{$full_curr_file}) {
983 # delete it so that only files that need deleting are left
984 delete $full_prev_all_files->{$full_curr_file};
985
986 # had it before. is it a metadata file?
987 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
988
989 # is it modified??
990 if (-M $full_curr_file < $archiveinf_timestamp) {
991 print STDERR "*** Detected a *modified metadata* file: $full_curr_file\n" if $verbosity >= 2;
992 # its newer than last build
993 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
994 }
995 }
996 else {
997 if ($incremental_mode eq "all") {
998
999 # had it before
1000 $block_hash->{'existing_files'}->{$full_curr_file} = 1;
1001
1002 }
1003 else {
1004 # Warning in "onlyadd" mode, but had it before!
1005 print STDERR "Warning: File $full_curr_file previously imported.\n";
1006 print STDERR " Treating as new file\n";
1007
1008 $block_hash->{'new_files'}->{$full_curr_file} = 1;
1009
1010 }
1011 }
1012 }
1013 else {
1014 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
1015 # the new file is the special sort of file greenstone uses
1016 # to attach metadata to src documents
1017 # i.e metadata.xml
1018 # (but note, the filename used is not constrained in
1019 # Greenstone to always be this)
1020
1021 print STDERR "*** Detected *new* metadata file: $full_curr_file\n" if $verbosity >= 2;
1022 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
1023 }
1024 else {
1025 $block_hash->{'new_files'}->{$full_curr_file} = 1;
1026 }
1027 }
1028
1029
1030 delete $block_hash->{'all_files'}->{$curr_file};
1031 }
1032
1033
1034
1035
1036 # Deal with complication of new or modified metadata files by forcing
1037 # everything from this point down in the file hierarchy to
1038 # be freshly imported.
1039 #
1040 # This may mean files that have not changed are reindexed, but does
1041 # guarantee by the end of processing all new metadata is correctly
1042 # associated with the relevant document(s).
1043
1044 foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
1045 my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
1046
1047 $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
1048 $situated_dir = &util::filename_to_regex($situated_dir); # need to escape windows slash \ and brackets in regular expression
1049
1050 # Go through existing_files, and mark anything that is contained
1051 # within 'situated_dir' to be reindexed (in case some of the metadata
1052 # attaches to one of these files)
1053
1054 my $reindex_files = [];
1055
1056 foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
1057
1058 if ($existing_f =~ m/^$situated_dir/) {
1059
1060 print STDERR "**** Existing file $existing_f\nis located within\n$situated_dir\n";
1061
1062 push(@$reindex_files,$existing_f);
1063 $block_hash->{'reindex_files'}->{$existing_f} = 1;
1064 delete $block_hash->{'existing_files'}->{$existing_f};
1065
1066 }
1067 }
1068
1069 # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
1070 # (or equivalent)
1071 $block_hash->{'new_files'}->{$new_mdf} = 1;
1072
1073 }
1074
1075 # go through remaining existing files and work out what has changed and needs to be reindexed.
1076 my @existing_files = sort keys %{$block_hash->{'existing_files'}};
1077
1078 my $reindex_files = [];
1079
1080 foreach my $existing_filename (@existing_files) {
1081 if (-M $existing_filename < $archiveinf_timestamp) {
1082 # file is newer than last build
1083
1084 my $existing_file = $existing_filename;
1085 #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
1086
1087 #my $collectdir_resafe = &util::filename_to_regex($collectdir);
1088 #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
1089
1090 print STDERR "**** Reindexing existing file: $existing_file\n";
1091
1092 push(@$reindex_files,$existing_file);
1093 $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1094 }
1095
1096 }
1097
1098
1099 # By this point full_prev_all_files contains the files
1100 # mentioned in archiveinf-src.db but are not in the 'import'
1101 # folder (or whatever was specified through -importdir ...)
1102
1103 # This list can contain files that were created in the 'tmp' or
1104 # 'cache' areas (such as screen-size and thumbnail images).
1105 #
1106 # In building the final list of files to delete, we test to see if
1107 # it exists on the filesystem and if it does (unusual for a "normal"
1108 # file in import, but possible in the case of 'tmp' files),
1109 # supress it from going into the final list
1110
1111 my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1112
1113 my @deleted_files = values %$full_prev_all_files;
1114 map { my $curr_file = $_;
1115 my $full_curr_file = $curr_file;
1116
1117 if (!&util::filename_is_absolute($curr_file)) {
1118 # add in import dir to make absolute
1119
1120 $full_curr_file = &util::filename_cat($collectdir,$curr_file);
1121 }
1122
1123
1124 if (!-e $full_curr_file) {
1125 $block_hash->{'deleted_files'}->{$curr_file} = 1;
1126 }
1127 } @deleted_files;
1128
1129
1130
1131}
1132
1133
1134# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1135# $mode is 'delete' or 'reindex'
1136sub mark_docs_for_deletion
1137{
1138 my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1139
1140 my $mode_text = "deleted from index";
1141 if ($mode eq "reindex") {
1142 $mode_text = "reindexed";
1143 }
1144
1145 # Get the infodbtype value for this collection from the arcinfo object
1146 my $infodbtype = $archive_info->{'infodbtype'};
1147
1148 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1149 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1150
1151
1152 # record files marked for deletion in arcinfo
1153 foreach my $file (@$deleted_files) {
1154 # use 'archiveinf-src' info database file to look up all the OIDs
1155 # that this file is used in (note in most cases, it's just one OID)
1156
1157 my $src_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
1158 my $oids = $src_rec->{'oid'};
1159 my $file_record_deleted = 0;
1160
1161 # delete the src record
1162 my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1163 &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
1164 &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1165
1166
1167 foreach my $oid (@$oids) {
1168
1169 # find the source doc (the primary file that becomes this oid)
1170 my $doc_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1171 my $doc_source_file = $doc_rec->{'src-file'}->[0];
1172 if (!&util::filename_is_absolute($doc_source_file)) {
1173 $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1174 }
1175
1176 if ($doc_source_file ne $file) {
1177 # its an associated or metadata file
1178
1179 # mark source doc for reimport as one of its assoc files has changed or deleted
1180 $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1181
1182 }
1183 my $curr_status = $archive_info->get_status_info($oid);
1184 if (defined($curr_status) && (($curr_status ne "D"))) {
1185 if ($verbosity>1) {
1186 print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1187 }
1188 # mark oid for deletion (it will be deleted or reimported)
1189 $archive_info->set_status_info($oid,"D");
1190 my $val = &dbutil::read_infodb_rawentry($infodbtype, $arcinfo_doc_filename, $oid);
1191 $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1192
1193 my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1194 my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1195
1196 &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1197 &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1198 }
1199 }
1200
1201 }
1202
1203 # now go through and check that we haven't marked any primary
1204 # files for reindex (because their associated files have
1205 # changed/deleted) when they have been deleted themselves. only in
1206 # delete mode.
1207
1208 if ($mode eq "delete") {
1209 foreach my $file (@$deleted_files) {
1210 if (defined $block_hash->{'reindex_files'}->{$file}) {
1211 delete $block_hash->{'reindex_files'}->{$file};
1212 }
1213 }
1214 }
1215
1216
1217}
1218
1219sub add_dir_contents_to_list {
1220
1221 my ($dirname, $list) = @_;
1222
1223 # Recur over directory contents.
1224 my (@dir, $subfile);
1225
1226 # find all the files in the directory
1227 if (!opendir (DIR, $dirname)) {
1228 print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
1229 return -1; # error in processing
1230 }
1231 @dir = readdir (DIR);
1232 closedir (DIR);
1233
1234 for (my $i = 0; $i < scalar(@dir); $i++) {
1235 my $subfile = $dir[$i];
1236 next if ($subfile =~ m/^\.\.?$/);
1237 next if ($subfile =~ /^\.svn$/);
1238 my $full_file = &util::filename_cat($dirname, $subfile);
1239 if (-d $full_file) {
1240 &add_dir_contents_to_list($full_file, $list);
1241 } else {
1242 push (@$list, $full_file);
1243 }
1244 }
1245
1246}
1247
1248
12491;
Note: See TracBrowser for help on using the repository browser.