source: gs2-extensions/parallel-building/trunk/src/perllib/inexport.pm@ 25401

Last change on this file since 25401 was 25401, checked in by jmt12, 12 years ago

Extend the code that 'pre-started' the GDBMServer to pre-start any infodbtype ending with *server (to support the new TDBServer)
File size: 41.6 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 # add collection's perllib dir into include path in
176 # case we have collection specific modules
177 my $collection_perllib_path = $ENV{'GSDLCOLLECTDIR'} . '/perllib';
178 my $inc_paths = join(':', @INC);
179 if ($inc_paths !~ /$collection_perllib_path/)
180 {
181 unshift (@INC, $collection_perllib_path); # [jmt12]
182 }
183
184 # check that we can open the faillog
185 my $faillog = $self->{'faillog'};
186 if ($faillog eq "") {
187 $faillog = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
188 }
189 open (FAILLOG, ">$faillog") ||
190 (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
191
192
193 my $faillogname = $faillog;
194 $faillog = 'inexport::FAILLOG';
195 $faillog->autoflush(1);
196 $self->{'faillog'} = $faillog;
197 $self->{'faillogname'} = $faillogname;
198
199 # Read in the collection configuration file.
200 my ($config_filename, $gs_mode) = &colcfg::get_collect_cfg_name($out);
201 my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
202
203 return ($config_filename,$collectcfg);
204}
205
206sub set_collection_options
207{
208 my $self = shift @_;
209 my ($collectcfg) = @_;
210
211 my $inexport_mode = $self->{'mode'};
212
213 my $verbosity = $self->{'verbosity'};
214 my $debug = $self->{'debug'};
215 my $importdir = $self->{'importdir'};
216 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'} || "";
217 my $out = $self->{'out'};
218
219 # If the infodbtype value wasn't defined in the collect.cfg file, use the default
220 if (!defined($collectcfg->{'infodbtype'}))
221 {
222 $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
223 }
224 if ($collectcfg->{'infodbtype'} eq "gdbm-txtgz") {
225 # we can't use the text version for archives dbs.
226 $collectcfg->{'infodbtype'} = "gdbm";
227 }
228 if (defined $collectcfg->{'importdir'} && $importdir eq "") {
229 $importdir = $collectcfg->{'importdir'};
230 }
231 if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
232 $archivedir = $collectcfg->{'archivedir'};
233 }
234 # fill in the default import and archives directories if none
235 # were supplied, turn all \ into / and remove trailing /
236 $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
237 $importdir =~ s/[\\\/]+/\//g;
238 $importdir =~ s/\/$//;
239 if (!-e $importdir) {
240 &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
241 die "\n";
242 }
243 $self->{'importdir'} = $importdir;
244
245 if ($archivedir eq "") {
246 if ($inexport_mode eq "import") {
247 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
248 }
249 elsif ($inexport_mode eq "export") {
250 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "export");
251 }
252 else {
253 print STDERR "Warning: Unrecognized import/export mode '$inexport_mode'\n";
254 print STDERR " Defaulting to 'archives' for file output\n";
255 $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
256 }
257 }
258
259 $archivedir =~ s/[\\\/]+/\//g;
260 $archivedir =~ s/\/$//;
261 $self->{'archivedir'} = $archivedir;
262
263 if ($verbosity !~ /\d+/) {
264 if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
265 $verbosity = $collectcfg->{'verbosity'};
266 } else {
267 $verbosity = 2; # the default
268 }
269 }
270 $self->{'verbosity'} = $verbosity;
271
272 if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
273 $self->{'manifest'} = $collectcfg->{'manifest'};
274 }
275
276 if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
277 if ($collectcfg->{'gzip'} =~ /^true$/i) {
278 $self->{'gzip'} = 1;
279 }
280 }
281
282 if ($self->{'maxdocs'} !~ /\-?\d+/) {
283 if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
284 $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
285 } else {
286 $self->{'maxdocs'} = -1; # the default
287 }
288 }
289
290 if ((defined $self->{'groupsize'}) && ($self->{'groupsize'} == 1)) {
291 if (defined $collectcfg->{'groupsize'} && $collectcfg->{'groupsize'} =~ /\d+/) {
292 $self->{'groupsize'} = $collectcfg->{'groupsize'};
293 }
294 }
295
296 if (!defined $self->{'OIDtype'}
297 || ($self->{'OIDtype'} !~ /^(hash|incremental|assigned|dirname)$/ )) {
298 if (defined $collectcfg->{'OIDtype'}
299 && $collectcfg->{'OIDtype'} =~ /^(hash|incremental|assigned|dirname)$/) {
300 $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
301 } else {
302 $self->{'OIDtype'} = "hash"; # the default
303 }
304 }
305
306 if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
307 if (defined $collectcfg->{'OIDmetadata'}) {
308 $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
309 } else {
310 $self->{'OIDmetadata'} = "dc.Identifier"; # the default
311 }
312 }
313
314 my $sortmeta = $self->{'sortmeta'};
315 if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
316 $sortmeta = $collectcfg->{'sortmeta'};
317 }
318 # sortmeta cannot be used with group size
319 $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
320 if (defined $sortmeta && $self->{'groupsize'} > 1) {
321 &gsprintf($out, "{import.cannot_sort}\n\n");
322 $sortmeta = undef;
323 }
324 $self->{'sortmeta'} = $sortmeta;
325
326 if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
327 $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
328 }
329
330 if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
331 $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
332 }
333 if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
334 $self->{'debug'} = 1;
335 }
336 if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
337 $self->{'gli'} = 1;
338 }
339 $self->{'gli'} = 0 unless defined $self->{'gli'};
340
341 # check keepold and removeold
342 my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
343
344 my ($removeold, $keepold, $incremental, $incremental_mode)
345 = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
346 $self->{'incremental'}, $checkdir,
347 $collectcfg);
348
349 $self->{'removeold'} = $removeold;
350 $self->{'keepold'} = $keepold;
351 $self->{'incremental'} = $incremental;
352 $self->{'incremental_mode'} = $incremental_mode;
353
354 # Since this wasted my morning, let's at least warn a user that manifest
355 # files now *only* work if keepold is set. [jmt12]
356 if ($self->{'manifest'} && !$self->{'keepold'})
357 {
358 print STDERR "Warning: -manifest flag should not be specified without also setting -keepold or -incremental. Ignoring.\n";
359 }
360}
361
362sub process_files
363{
364 my $self = shift @_;
365 my ($config_filename,$collectcfg) = @_;
366
367 my $inexport_mode = $self->{'mode'};
368
369 my $verbosity = $self->{'verbosity'};
370 my $debug = $self->{'debug'};
371
372 my $importdir = $self->{'importdir'};
373 my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
374
375 my $incremental = $self->{'incremental'};
376 my $incremental_mode = $self->{'incremental_mode'};
377
378 my $removeold = $self->{'removeold'};
379 my $keepold = $self->{'keepold'};
380
381 my $saveas = $self->{'saveas'};
382 my $OIDtype = $self->{'OIDtype'};
383 my $OIDmetadata = $self->{'OIDmetadata'};
384
385 my $out = $self->{'out'};
386 my $faillog = $self->{'faillog'};
387
388 my $maxdocs = $self->{'maxdocs'};
389 my $gzip = $self->{'gzip'};
390 my $groupsize = $self->{'groupsize'};
391 my $sortmeta = $self->{'sortmeta'};
392
393 my $removeprefix = $self->{'removeprefix'};
394 my $removesuffix = $self->{'removesuffix'};
395
396 my $gli = $self->{'gli'};
397
398 my $jobs = $self->{'jobs'};
399 my $epoch = $self->{'epoch'};
400
401 # related to export
402 my $xsltfile = $self->{'xsltfile'};
403 my $group_marc = $self->{'group_marc'};
404 my $mapping_file = $self->{'mapping_file'};
405 my $xslt_mets = $self->{'xslt_mets'};
406 my $xslt_txt = $self->{'xslt_txt'};
407 my $fedora_namespace = $self->{'fedora_namespace'};
408 my $metadata_prefix = $self->{'metadata_prefix'};
409
410 if ($inexport_mode eq "import") {
411 print STDERR "<Import>\n" if $gli;
412 }
413 else {
414 print STDERR "<export>\n" if $gli;
415 }
416
417 my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
418 if ($self->{'manifest'} ne "") {
419 my $manifest_filename = $self->{'manifest'};
420
421 if (!&util::filename_is_absolute($manifest_filename)) {
422 $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
423 }
424
425 $self->{'manifest'} =~ s/[\\\/]+/\//g;
426 $self->{'manifest'} =~ s/\/$//;
427
428 $manifest_lookup->parse($manifest_filename);
429 }
430
431 my $manifest = $self->{'manifest'};
432
433 # load all the plugins
434 my $plugins = [];
435 if (defined $collectcfg->{'plugin'}) {
436 $plugins = $collectcfg->{'plugin'};
437 }
438
439 my $plugin_incr_mode = $incremental_mode;
440 if ($manifest ne "") {
441 # if we have a manifest file, then we pretend we are fully incremental for plugins
442 $plugin_incr_mode = "all";
443 }
444 #some global options for the plugins
445 my @global_opts = ();
446
447 my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode);
448 if (scalar(@$pluginfo) == 0) {
449 &gsprintf($out, "{import.no_plugins_loaded}\n");
450 die "\n";
451 }
452
453 # remove the old contents of the archives directory (and tmp
454 # directory) if needed
455
456 if ($removeold) {
457 if (-e $archivedir) {
458 &gsprintf($out, "{import.removing_archives}\n");
459 &util::rm_r ($archivedir);
460 }
461 my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
462 $tmpdir =~ s/[\\\/]+/\//g;
463 $tmpdir =~ s/\/$//;
464 if (-e $tmpdir) {
465 #&gsprintf($out, "{import.removing_tmpdir}\n");
466 #&util::rm_r ($tmpdir);
467 }
468 }
469
470 # create the archives dir if needed
471 &util::mk_all_dir($archivedir);
472
473 # read the archive information file
474
475 # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
476 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
477 &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
478
479 # Warning! Black magic follows. When the following functions are called on
480 # the GDBMServer class they will actually prompt the running of the Server
481 # and attach themselves as a listener (even though they don't do anything)
482 # This is done so that, in parallel importing, the server will persist
483 # until the top level import.pl (which will be the first this that calls
484 # this function) completes. [jmt12]
485 my $create_server = 0;
486 # - infodb's of type *server need to be started on the same machine that
487 # runs parallel-import.sh, especially when distributed processing
488 # gets involved.
489 if ($collectcfg->{'infodbtype'} =~ /server$/)
490 {
491 $create_server = 1;
492 }
493 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir, $create_server);
494 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir, $create_server);
495
496 my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
497 $archive_info->load_info ($arcinfo_doc_filename);
498
499 if ($manifest eq "") {
500 # Load in list of files in import folder from last import (if present)
501 $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
502 }
503
504 ####Use Plugout####
505 my $plugout;
506
507 if ($inexport_mode eq "import") {
508 if (defined $collectcfg->{'plugout'}) {
509 # If a plugout was specified in the collect.cfg file, assume it is sensible
510 # We can't check the name because it could be anything, if it is a custom plugout
511 $plugout = $collectcfg->{'plugout'};
512 }
513 else{
514 if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
515 push @$plugout,"GreenstoneXMLPlugout";
516 }
517 else{
518 push @$plugout,$saveas."Plugout";
519 }
520 }
521 }
522 else {
523 if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
524 $plugout = $collectcfg->{'plugout'};
525 }
526 else{
527 if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
528 push @$plugout,"GreenstoneMETSPlugout";
529 }
530 else{
531 push @$plugout,$saveas."Plugout";
532 }
533 }
534 }
535
536 my $plugout_name = $plugout->[0];
537
538 push @$plugout,("-output_info",$archive_info) if (defined $archive_info);
539 push @$plugout,("-verbosity",$verbosity) if (defined $verbosity);
540 push @$plugout,("-debug") if ($debug);
541 push @$plugout,("-group_size",$groupsize) if (defined $groupsize);
542 push @$plugout,("-gzip_output") if ($gzip);
543 push @$plugout,("-output_handle",$out) if (defined $out);
544
545 push @$plugout,("-xslt_file",$xsltfile) if (defined $xsltfile && $xsltfile ne "");
546
547 if ($plugout_name =~ m/^MARCXMLPlugout$/) {
548 push @$plugout,("-group") if ($group_marc);
549 push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
550 }
551 if ($plugout_name =~ m/^.*METSPlugout$/) {
552 push @$plugout,("-xslt_mets",$xslt_mets) if (defined $xslt_mets && $xslt_mets ne "");
553 push @$plugout,("-xslt_txt",$xslt_txt) if (defined $xslt_txt && $xslt_txt ne "");
554 }
555
556 if ($plugout_name eq "FedoraMETSPlugout") {
557 push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
558 }
559
560 if ($plugout_name eq "DSpacePlugout") {
561 push @$plugout,("-metadata_prefix",$metadata_prefix) if (defined $metadata_prefix && $metadata_prefix ne "");
562 }
563
564 my $processor = &plugout::load_plugout($plugout);
565 $processor->setoutputdir ($archivedir);
566 $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
567 $processor->set_OIDtype ($OIDtype, $OIDmetadata);
568
569 &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
570
571 if ($removeold) {
572 # occasionally, plugins may want to do something on remove
573 # old, eg pharos image indexing
574 &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
575 }
576
577 # process the import directory
578 my $block_hash = {};
579 $block_hash->{'new_files'} = {};
580 $block_hash->{'reindex_files'} = {};
581 # All of these are set somewhere else, so it's kinda nice to define them
582 # here. [jmt12]
583 $block_hash->{'all_files'} = {};
584 $block_hash->{'deleted_files'} = {};
585 $block_hash->{'file_blocks'} = {};
586 $block_hash->{'metadata_files'} = {};
587 $block_hash->{'shared_fileroot'} = '';
588 # My new flag so we can tell we had a manifest way down in the plugins
589 # [jmt12]
590 $block_hash->{'manifest'} = 'false';
591 my $metadata = {};
592
593 # global blocking pass may set up some metadata
594 # - when we have a manifest file we don't do this -unless- the collection
595 # configuration indicates this collection contains complex (inherited)
596 # metadata. [jmt12]
597 if ($manifest eq '' || (defined $collectcfg->{'complexmeta'} && $collectcfg->{'complexmeta'} eq 'true'))
598 {
599 #print "Temporarily skipping global file scan as it is slow as molasses\n";
600 &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
601 }
602 else
603 {
604 print "Skipping global file scan due to manifest and complexmeta configuration\n";
605 }
606
607 if ($manifest ne "") {
608
609 $block_hash->{'manifest'} = 'true';
610
611 #
612 # 1. Process delete files first
613 #
614 my @deleted_files = keys %{$manifest_lookup->{'delete'}};
615 my @full_deleted_files = ();
616
617 # ensure all filenames are absolute
618 foreach my $df (@deleted_files) {
619 my $full_df =
620 (&util::filename_is_absolute($df))
621 ? $df
622 : &util::filename_cat($importdir,$df);
623
624 if (-d $full_df) {
625 &add_dir_contents_to_list($full_df, \@full_deleted_files);
626 } else {
627 push(@full_deleted_files,$full_df);
628 }
629 }
630
631 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
632 mark_docs_for_deletion($archive_info,{},
633 \@full_deleted_files,
634 $archivedir, $verbosity, "delete");
635
636
637 #
638 # 2. Now files for reindexing
639 #
640
641 my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
642 my @full_reindex_files = ();
643 # ensure all filenames are absolute
644 foreach my $rf (@reindex_files) {
645 my $full_rf =
646 (&util::filename_is_absolute($rf))
647 ? $rf
648 : &util::filename_cat($importdir,$rf);
649
650 if (-d $full_rf) {
651 &add_dir_contents_to_list($full_rf, \@full_reindex_files);
652 } else {
653 push(@full_reindex_files,$full_rf);
654 }
655 }
656
657 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
658 mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
659
660 # And now to ensure the new version of the file processed by
661 # appropriate plugin, we need to add it to block_hash reindex list
662 foreach my $full_rf (@full_reindex_files) {
663 $block_hash->{'reindex_files'}->{$full_rf} = 1;
664 }
665
666
667 #
668 # 3. Now finally any new files - add to block_hash new_files list
669 #
670
671 my @new_files = keys %{$manifest_lookup->{'index'}};
672 my @full_new_files = ();
673
674 foreach my $nf (@new_files) {
675 # ensure filename is absolute
676 my $full_nf =
677 (&util::filename_is_absolute($nf))
678 ? $nf
679 : &util::filename_cat($importdir,$nf);
680
681 if (-d $full_nf) {
682 &add_dir_contents_to_list($full_nf, \@full_new_files);
683 } else {
684 push(@full_new_files,$full_nf);
685 }
686 }
687
688 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
689 # need to check this file exists before trying to read it. [jmt12]
690 if (-e $arcinfo_src_filename)
691 {
692 my $arcinfodb_map = {};
693 &dbutil::read_infodb_file($collectcfg->{'infodbtype'}, $arcinfo_src_filename, $arcinfodb_map);
694 foreach my $f (@full_new_files) {
695 # check that we haven't seen it already
696 if (defined $arcinfodb_map->{$f}) {
697 # TODO make better warning
698 print STDERR "Warning: $f already in src archive, \n";
699 } else {
700 $block_hash->{'new_files'}->{$f} = 1;
701 }
702 }
703 undef $arcinfodb_map;
704 }
705 # no existing files - so we can just add all the ones that need adding.
706 # [jmt12]
707 else
708 {
709 foreach my $f (@full_new_files)
710 {
711 $block_hash->{'new_files'}->{$f} = 1;
712 }
713 }
714
715 # If we are not using complex inherited metadata (and thus have skipped
716 # the global file scan) we need to at least scan the directory of the
717 # files being indexed/reindexed. [jmt12]
718 if (!defined $collectcfg->{'complexmeta'} || $collectcfg->{'complexmeta'} ne 'true')
719 {
720 my @all_files_to_import = (keys %{$block_hash->{'reindex_files'}}, keys %{$block_hash->{'new_files'}});
721 foreach my $file_to_import (@all_files_to_import)
722 {
723 my $dir_to_import = $file_to_import;
724 $dir_to_import =~ s/[^\\\/]*$//;
725 # - one day we may need to manually scan this directory for child
726 # directories and somehow explicitly block them from being
727 # recursed.
728 if (-d $dir_to_import)
729 {
730 &plugin::file_block_read($pluginfo, $dir_to_import, '', $block_hash, $metadata, $gli);
731 }
732 }
733 }
734 }
735 else {
736 # if incremental, we read through the import folder to see whats changed.
737
738 if ($incremental || $incremental_mode eq "onlyadd") {
739 prime_doc_oid_count($archivedir);
740
741 # Can now work out which files were new, already existed, and have
742 # been deleted
743
744 new_vs_old_import_diff($archive_info,$block_hash,$importdir,
745 $archivedir,$verbosity,$incremental_mode);
746
747 my @new_files = sort keys %{$block_hash->{'new_files'}};
748 if (scalar(@new_files>0)) {
749 print STDERR "New files and modified metadata files since last import:\n ";
750 print STDERR join("\n ",@new_files), "\n";
751 }
752
753 if ($incremental) {
754 # only look for deletions if we are truely incremental
755 my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
756 # Filter out any in gsdl/tmp area
757 my @filtered_deleted_files = ();
758 my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
759 my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
760 $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
761 $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
762
763 foreach my $df (@deleted_files) {
764 next if ($df =~ m/^$gsdl_tmp_area/);
765 next if ($df =~ m/^$collect_tmp_area/);
766
767 push(@filtered_deleted_files,$df);
768 }
769
770
771 @deleted_files = @filtered_deleted_files;
772
773 if (scalar(@deleted_files)>0) {
774 print STDERR "Files deleted since last import:\n ";
775 print STDERR join("\n ",@deleted_files), "\n";
776
777
778 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
779
780 mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
781 }
782
783 my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
784
785 if (scalar(@reindex_files)>0) {
786 print STDERR "Files to reindex since last import:\n ";
787 print STDERR join("\n ",@reindex_files), "\n";
788 &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
789 mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
790 }
791
792 }
793 }
794 }
795
796 # Check for existence of the file that's to contain earliestDateStamp in archivesdir
797 # Do nothing if the file already exists (file exists on incremental build).
798 # If the file doesn't exist, as happens on full build, create it and write out the current datestamp into it
799 # In buildcol, read the file's contents and set the earliestdateStamp in GS2's build.cfg / GS3's buildconfig.xml
800 # In doc.pm have set_oaiLastModified similar to set_lastmodified, and create the doc fields
801 # oailastmodified and oailastmodifieddate
802 my $earliestDatestampFile = &util::filename_cat($archivedir, "earliestDatestamp");
803 if (!-f $earliestDatestampFile && -d $archivedir) {
804 my $current_time_in_seconds = time; # in seconds
805
806 if(open(FOUT, ">$earliestDatestampFile")) {
807 # || (&gsprintf(STDERR, "{common.cannot_open}: $!\n", $earliestDatestampFile) && die);
808 print FOUT $current_time_in_seconds;
809 close(FOUT);
810 }
811 else {
812 &gsprintf(STDERR, "{import.cannot_write_earliestdatestamp}\n", $earliestDatestampFile);
813 }
814
815 }
816
817 # now, whichever mode we are in, we can process the entire import folder
818 if ((defined $jobs) && ($jobs > 1))
819 {
820 # if jobs are set to >1, run in parallel using MPI helper
821 # [hs, 1 july 2010]
822 &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
823 $self->{'collection'}, $self->{'site'});
824 }
825 else
826 {
827 &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
828 }
829
830
831 if ($saveas eq "FedoraMETS") {
832 # create collection "doc obj" for Fedora that contains
833 # collection-level metadata
834
835 my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
836 $doc_obj->set_OID("collection");
837
838 my $col_name = undef;
839 my $col_meta = $collectcfg->{'collectionmeta'};
840
841 if (defined $col_meta) {
842 store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
843 store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description
844 }
845 $processor->process($doc_obj);
846 }
847
848 &plugin::end($pluginfo, $processor);
849
850 &plugin::deinit($pluginfo, $processor);
851
852 # Store the value of OIDCount (used in doc.pm) so it can be
853 # restored correctly to this value on an incremental build
854 store_doc_oid_count($archivedir);
855
856 # write out the archive information file
857 $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
858 $processor->close_group_output() if $processor->is_group();
859
860 # for backwards compatability with archvies.inf file
861 if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
862 $archive_info->save_info($arcinfo_doc_filename);
863 }
864 else {
865 $archive_info->save_revinfo_db($arcinfo_src_filename);
866 }
867
868 return $pluginfo;
869}
870
871
872sub generate_statistics
873{
874 my $self = shift @_;
875 my ($pluginfo) = @_;
876
877 my $inexport_mode = $self->{'mode'};
878
879 my $statsfile = $self->{'statsfile'};
880 my $out = $self->{'out'};
881 my $faillogname = $self->{'faillogname'};
882 my $gli = $self->{'gli'};
883 my $jobs = $self->{'jobs'};
884
885 # write out import stats
886
887 if ((!defined $jobs) || ($jobs == 1))
888 {
889 # only output statistics if there are multiple jobs
890 # [hs, 1 july 2010]
891
892 my $close_stats = 0;
893 if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
894 if (open (STATS, ">$statsfile")) {
895 $statsfile = 'inexport::STATS';
896 $close_stats = 1;
897 } else {
898 &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
899 &gsprintf($out, "{import.stats_backup}\n");
900 $statsfile = 'STDERR';
901 }
902 }
903
904 &gsprintf($out, "\n");
905 &gsprintf($out, "*********************************************\n");
906 &gsprintf($out, "{$inexport_mode.complete}\n");
907 &gsprintf($out, "*********************************************\n");
908
909 &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
910 if ($close_stats) {
911 close STATS;
912 }
913 }
914
915 close OUT if $self->{'close_out'};
916 close FAILLOG;
917}
918
919
920sub store_collectionmeta
921{
922 my ($collectionmeta,$field,$doc_obj) = @_;
923
924 my $section = $doc_obj->get_top_section();
925
926 my $field_hash = $collectionmeta->{$field};
927
928 foreach my $k (keys %$field_hash)
929 {
930 my $val = $field_hash->{$k};
931
932 ### print STDERR "*** $k = $field_hash->{$k}\n";
933
934 my $md_label = "ex.$field";
935
936
937 if ($k =~ m/^\[l=(.*?)\]$/)
938 {
939
940 my $md_suffix = $1;
941 $md_label .= "^$md_suffix";
942 }
943
944
945 $doc_obj->add_utf8_metadata($section,$md_label, $val);
946
947 # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
948 # while "collectionname" in GS2 is called "name" in GS3.
949 # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
950 if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
951 {
952 $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
953 }
954
955 }
956}
957
958
959sub oid_count_file {
960 my ($archivedir) = @_;
961 return &util::filename_cat ($archivedir, "OIDcount");
962}
963
964
965sub prime_doc_oid_count
966{
967 my ($archivedir) = @_;
968 my $oid_count_filename = &oid_count_file($archivedir);
969
970 if (-e $oid_count_filename) {
971 if (open(OIDIN,"<$oid_count_filename")) {
972 my $OIDcount = <OIDIN>;
973 chomp $OIDcount;
974 close(OIDIN);
975
976 $doc::OIDcount = $OIDcount;
977 }
978 else {
979 &gsprintf(STDERR, "{import.cannot_read_OIDcount}\n", $oid_count_filename);
980 }
981 }
982
983}
984
985sub store_doc_oid_count
986{
987 # Use the file "OIDcount" in the archives directory to record
988 # what value doc.pm got up to
989
990 my ($archivedir) = @_;
991 my $oid_count_filename = &oid_count_file($archivedir);
992
993
994 if (open(OIDOUT,">$oid_count_filename")) {
995 print OIDOUT $doc::OIDcount, "\n";
996
997 close(OIDOUT);
998 }
999 else {
1000 &gsprintf(STDERR, "{import.cannot_write_OIDcount}\n", $oid_count_filename);
1001 }
1002}
1003
1004
1005
1006sub new_vs_old_import_diff
1007{
1008 my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
1009
1010 # Get the infodbtype value for this collection from the arcinfo object
1011 my $infodbtype = $archive_info->{'infodbtype'};
1012
1013 # in this method, we want to know if metadata files are modified or not.
1014 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1015
1016 my $archiveinf_timestamp = -M $arcinfo_doc_filename;
1017
1018 # First convert all files to absolute form
1019 # This is to support the situation where the import folder is not
1020 # the default
1021
1022 my $prev_all_files = $archive_info->{'prev_import_filelist'};
1023 my $full_prev_all_files = {};
1024
1025 foreach my $prev_file (keys %$prev_all_files) {
1026
1027 if (!&util::filename_is_absolute($prev_file)) {
1028 my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
1029 $full_prev_all_files->{$full_prev_file} = $prev_file;
1030 }
1031 else {
1032 $full_prev_all_files->{$prev_file} = $prev_file;
1033 }
1034 }
1035
1036
1037 # Figure out which are the new files, existing files and so
1038 # by implication the files from the previous import that are not
1039 # there any more => mark them for deletion
1040 foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
1041
1042 my $full_curr_file = $curr_file;
1043
1044 # entry in 'all_files' is moved to either 'existing_files',
1045 # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
1046
1047 if (!&util::filename_is_absolute($curr_file)) {
1048 # add in import dir to make absolute
1049 $full_curr_file = &util::filename_cat($importdir,$curr_file);
1050 }
1051
1052 # figure out if new file or not
1053 if (defined $full_prev_all_files->{$full_curr_file}) {
1054 # delete it so that only files that need deleting are left
1055 delete $full_prev_all_files->{$full_curr_file};
1056
1057 # had it before. is it a metadata file?
1058 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
1059
1060 # is it modified??
1061 if (-M $full_curr_file < $archiveinf_timestamp) {
1062 print STDERR "*** Detected a *modified metadata* file: $full_curr_file\n" if $verbosity >= 2;
1063 # its newer than last build
1064 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
1065 }
1066 }
1067 else {
1068 if ($incremental_mode eq "all") {
1069
1070 # had it before
1071 $block_hash->{'existing_files'}->{$full_curr_file} = 1;
1072
1073 }
1074 else {
1075 # Warning in "onlyadd" mode, but had it before!
1076 print STDERR "Warning: File $full_curr_file previously imported.\n";
1077 print STDERR " Treating as new file\n";
1078
1079 $block_hash->{'new_files'}->{$full_curr_file} = 1;
1080
1081 }
1082 }
1083 }
1084 else {
1085 if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
1086 # the new file is the special sort of file greenstone uses
1087 # to attach metadata to src documents
1088 # i.e metadata.xml
1089 # (but note, the filename used is not constrained in
1090 # Greenstone to always be this)
1091
1092 print STDERR "*** Detected *new* metadata file: $full_curr_file\n" if $verbosity >= 2;
1093 $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
1094 }
1095 else {
1096 $block_hash->{'new_files'}->{$full_curr_file} = 1;
1097 }
1098 }
1099
1100
1101 delete $block_hash->{'all_files'}->{$curr_file};
1102 }
1103
1104
1105
1106
1107 # Deal with complication of new or modified metadata files by forcing
1108 # everything from this point down in the file hierarchy to
1109 # be freshly imported.
1110 #
1111 # This may mean files that have not changed are reindexed, but does
1112 # guarantee by the end of processing all new metadata is correctly
1113 # associated with the relevant document(s).
1114
1115 foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
1116 my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
1117
1118 $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
1119 $situated_dir =~ s/\\/\\\\/g; # need to protect windows slash \ in regular expression
1120
1121 # Go through existing_files, and mark anything that is contained
1122 # within 'situated_dir' to be reindexed (in case some of the metadata
1123 # attaches to one of these files)
1124
1125 my $reindex_files = [];
1126
1127 foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
1128
1129 if ($existing_f =~ m/^$situated_dir/) {
1130
1131 print STDERR "**** Existing file $existing_f\nis located within\n$situated_dir\n";
1132
1133 push(@$reindex_files,$existing_f);
1134 $block_hash->{'reindex_files'}->{$existing_f} = 1;
1135 delete $block_hash->{'existing_files'}->{$existing_f};
1136
1137 }
1138 }
1139
1140 # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
1141 # (or equivalent)
1142 $block_hash->{'new_files'}->{$new_mdf} = 1;
1143
1144 }
1145
1146 # go through remaining existing files and work out what has changed and needs to be reindexed.
1147 my @existing_files = sort keys %{$block_hash->{'existing_files'}};
1148
1149 my $reindex_files = [];
1150
1151 foreach my $existing_filename (@existing_files) {
1152 if (-M $existing_filename < $archiveinf_timestamp) {
1153 # file is newer than last build
1154
1155 my $existing_file = $existing_filename;
1156 #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
1157
1158 #my $collectdir_resafe = &util::filename_to_regex($collectdir);
1159 #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
1160
1161 print STDERR "**** Reindexing existing file: $existing_file\n";
1162
1163 push(@$reindex_files,$existing_file);
1164 $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1165 }
1166
1167 }
1168
1169
1170 # By this point full_prev_all_files contains the files
1171 # mentioned in archiveinf-src.db but are not in the 'import'
1172 # folder (or whatever was specified through -importdir ...)
1173
1174 # This list can contain files that were created in the 'tmp' or
1175 # 'cache' areas (such as screen-size and thumbnail images).
1176 #
1177 # In building the final list of files to delete, we test to see if
1178 # it exists on the filesystem and if it does (unusual for a "normal"
1179 # file in import, but possible in the case of 'tmp' files),
1180 # supress it from going into the final list
1181
1182 my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1183
1184 my @deleted_files = values %$full_prev_all_files;
1185 map { my $curr_file = $_;
1186 my $full_curr_file = $curr_file;
1187
1188 if (!&util::filename_is_absolute($curr_file)) {
1189 # add in import dir to make absolute
1190
1191 $full_curr_file = &util::filename_cat($collectdir,$curr_file);
1192 }
1193
1194
1195 if (!-e $full_curr_file) {
1196 $block_hash->{'deleted_files'}->{$curr_file} = 1;
1197 }
1198 } @deleted_files;
1199
1200
1201
1202}
1203
1204
1205# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1206# $mode is 'delete' or 'reindex'
1207sub mark_docs_for_deletion
1208{
1209 my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1210
1211 my $mode_text = "deleted from index";
1212 if ($mode eq "reindex") {
1213 $mode_text = "reindexed";
1214 }
1215
1216 # Get the infodbtype value for this collection from the arcinfo object
1217 my $infodbtype = $archive_info->{'infodbtype'};
1218
1219 my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1220 my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1221
1222
1223 # record files marked for deletion in arcinfo
1224 foreach my $file (@$deleted_files) {
1225 # use 'archiveinf-src' info database file to look up all the OIDs
1226 # that this file is used in (note in most cases, it's just one OID)
1227
1228 my $src_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
1229 my $oids = $src_rec->{'oid'};
1230 my $file_record_deleted = 0;
1231
1232 # delete the src record
1233 my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1234 &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
1235 &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1236
1237
1238 foreach my $oid (@$oids) {
1239
1240 # find the source doc (the primary file that becomes this oid)
1241 my $doc_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1242 my $doc_source_file = $doc_rec->{'src-file'}->[0];
1243 if (!&util::filename_is_absolute($doc_source_file)) {
1244 $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1245 }
1246
1247 if ($doc_source_file ne $file) {
1248 # its an associated or metadata file
1249
1250 # mark source doc for reimport as one of its assoc files has changed or deleted
1251 $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1252
1253 }
1254 my $curr_status = $archive_info->get_status_info($oid);
1255 if (defined($curr_status) && (($curr_status ne "D"))) {
1256 if ($verbosity>1) {
1257 print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1258 }
1259 # mark oid for deletion (it will be deleted or reimported)
1260 $archive_info->set_status_info($oid,"D");
1261 my $val = &dbutil::read_infodb_rawentry($infodbtype, $arcinfo_doc_filename, $oid);
1262 $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1263
1264 my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1265 my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1266
1267 &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1268 &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1269 }
1270 }
1271
1272 }
1273
1274 # now go through and check that we haven't marked any primary
1275 # files for reindex (because their associated files have
1276 # changed/deleted) when they have been deleted themselves. only in
1277 # delete mode.
1278
1279 if ($mode eq "delete") {
1280 foreach my $file (@$deleted_files) {
1281 if (defined $block_hash->{'reindex_files'}->{$file}) {
1282 delete $block_hash->{'reindex_files'}->{$file};
1283 }
1284 }
1285 }
1286
1287
1288}
1289
1290sub add_dir_contents_to_list {
1291
1292 my ($dirname, $list) = @_;
1293
1294 # Recur over directory contents.
1295 my (@dir, $subfile);
1296
1297 # find all the files in the directory
1298 if (!opendir (DIR, $dirname)) {
1299 print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
1300 return -1; # error in processing
1301 }
1302 @dir = readdir (DIR);
1303 closedir (DIR);
1304
1305 for (my $i = 0; $i < scalar(@dir); $i++) {
1306 my $subfile = $dir[$i];
1307 next if ($subfile =~ m/^\.\.?$/);
1308 next if ($subfile =~ /^\.svn$/);
1309 my $full_file = &util::filename_cat($dirname, $subfile);
1310 if (-d $full_file) {
1311 &add_dir_contents_to_list($full_file, $list);
1312 } else {
1313 push (@$list, $full_file);
1314 }
1315 }
1316
1317}
1318
1319
13201;
Note: See TracBrowser for help on using the repository browser.