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 |
|
---|
26 | package inexport;
|
---|
27 |
|
---|
28 | use strict;
|
---|
29 |
|
---|
30 | no strict 'refs'; # allow filehandles to be variables and vice versa
|
---|
31 | no strict 'subs'; # allow barewords (eg STDERR) as function arguments
|
---|
32 |
|
---|
33 | use arcinfo;
|
---|
34 | use colcfg;
|
---|
35 | use dbutil;
|
---|
36 | use doc;
|
---|
37 | use plugin;
|
---|
38 | use plugout;
|
---|
39 | use manifest;
|
---|
40 | use inexport;
|
---|
41 | use util;
|
---|
42 | use scriptutil;
|
---|
43 | use FileHandle;
|
---|
44 | use gsprintf 'gsprintf';
|
---|
45 | use printusage;
|
---|
46 | use parse2;
|
---|
47 |
|
---|
48 | use File::Basename;
|
---|
49 |
|
---|
50 | sub 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
|
---|
129 | sub 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 | }
|
---|
153 | sub get_collection
|
---|
154 | {
|
---|
155 | my $self = shift @_;
|
---|
156 |
|
---|
157 | return $self->{'collection'};
|
---|
158 | }
|
---|
159 |
|
---|
160 |
|
---|
161 | sub 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 |
|
---|
207 | sub 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|incremental|assigned|dirname)$/ )) {
|
---|
300 | if (defined $collectcfg->{'OIDtype'}
|
---|
301 | && $collectcfg->{'OIDtype'} =~ /^(hash|incremental|assigned|dirname)$/) {
|
---|
302 | $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
|
---|
303 | } else {
|
---|
304 | $self->{'OIDtype'} = "hash"; # the default
|
---|
305 | }
|
---|
306 | }
|
---|
307 |
|
---|
308 | if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
|
---|
309 | if (defined $collectcfg->{'OIDmetadata'}) {
|
---|
310 | $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
|
---|
311 | } else {
|
---|
312 | $self->{'OIDmetadata'} = "dc.Identifier"; # the default
|
---|
313 | }
|
---|
314 | }
|
---|
315 |
|
---|
316 | my $sortmeta = $self->{'sortmeta'};
|
---|
317 | if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
|
---|
318 | $sortmeta = $collectcfg->{'sortmeta'};
|
---|
319 | }
|
---|
320 | # sortmeta cannot be used with group size
|
---|
321 | $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
|
---|
322 | if (defined $sortmeta && $self->{'groupsize'} > 1) {
|
---|
323 | &gsprintf($out, "{import.cannot_sort}\n\n");
|
---|
324 | $sortmeta = undef;
|
---|
325 | }
|
---|
326 | $self->{'sortmeta'} = $sortmeta;
|
---|
327 |
|
---|
328 | if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
|
---|
329 | $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
|
---|
330 | }
|
---|
331 |
|
---|
332 | if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
|
---|
333 | $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
|
---|
334 | }
|
---|
335 | if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
|
---|
336 | $self->{'debug'} = 1;
|
---|
337 | }
|
---|
338 | if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
|
---|
339 | $self->{'gli'} = 1;
|
---|
340 | }
|
---|
341 | $self->{'gli'} = 0 unless defined $self->{'gli'};
|
---|
342 |
|
---|
343 | # check keepold and removeold
|
---|
344 | my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
|
---|
345 |
|
---|
346 | my ($removeold, $keepold, $incremental, $incremental_mode)
|
---|
347 | = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
|
---|
348 | $self->{'incremental'}, $checkdir,
|
---|
349 | $collectcfg);
|
---|
350 |
|
---|
351 | $self->{'removeold'} = $removeold;
|
---|
352 | $self->{'keepold'} = $keepold;
|
---|
353 | $self->{'incremental'} = $incremental;
|
---|
354 | $self->{'incremental_mode'} = $incremental_mode;
|
---|
355 | }
|
---|
356 |
|
---|
357 | sub process_files
|
---|
358 | {
|
---|
359 | my $self = shift @_;
|
---|
360 | my ($config_filename,$collectcfg) = @_;
|
---|
361 |
|
---|
362 | my $inexport_mode = $self->{'mode'};
|
---|
363 |
|
---|
364 | my $verbosity = $self->{'verbosity'};
|
---|
365 | my $debug = $self->{'debug'};
|
---|
366 |
|
---|
367 | my $importdir = $self->{'importdir'};
|
---|
368 | my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
|
---|
369 |
|
---|
370 | my $incremental = $self->{'incremental'};
|
---|
371 | my $incremental_mode = $self->{'incremental_mode'};
|
---|
372 |
|
---|
373 | my $gs_version = $self->{'gs_version'};
|
---|
374 |
|
---|
375 | my $removeold = $self->{'removeold'};
|
---|
376 | my $keepold = $self->{'keepold'};
|
---|
377 |
|
---|
378 | my $saveas = $self->{'saveas'};
|
---|
379 | my $OIDtype = $self->{'OIDtype'};
|
---|
380 | my $OIDmetadata = $self->{'OIDmetadata'};
|
---|
381 |
|
---|
382 | my $out = $self->{'out'};
|
---|
383 | my $faillog = $self->{'faillog'};
|
---|
384 |
|
---|
385 | my $maxdocs = $self->{'maxdocs'};
|
---|
386 | my $gzip = $self->{'gzip'};
|
---|
387 | my $groupsize = $self->{'groupsize'};
|
---|
388 | my $sortmeta = $self->{'sortmeta'};
|
---|
389 |
|
---|
390 | my $removeprefix = $self->{'removeprefix'};
|
---|
391 | my $removesuffix = $self->{'removesuffix'};
|
---|
392 |
|
---|
393 | my $gli = $self->{'gli'};
|
---|
394 |
|
---|
395 | my $jobs = $self->{'jobs'};
|
---|
396 | my $epoch = $self->{'epoch'};
|
---|
397 |
|
---|
398 | # related to export
|
---|
399 | my $xsltfile = $self->{'xsltfile'};
|
---|
400 | my $group_marc = $self->{'group_marc'};
|
---|
401 | my $mapping_file = $self->{'mapping_file'};
|
---|
402 | my $xslt_mets = $self->{'xslt_mets'};
|
---|
403 | my $xslt_txt = $self->{'xslt_txt'};
|
---|
404 | my $fedora_namespace = $self->{'fedora_namespace'};
|
---|
405 | my $metadata_prefix = $self->{'metadata_prefix'};
|
---|
406 |
|
---|
407 | if ($inexport_mode eq "import") {
|
---|
408 | print STDERR "<Import>\n" if $gli;
|
---|
409 | }
|
---|
410 | else {
|
---|
411 | print STDERR "<export>\n" if $gli;
|
---|
412 | }
|
---|
413 |
|
---|
414 | my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
|
---|
415 | if ($self->{'manifest'} ne "") {
|
---|
416 | my $manifest_filename = $self->{'manifest'};
|
---|
417 |
|
---|
418 | if (!&util::filename_is_absolute($manifest_filename)) {
|
---|
419 | $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
|
---|
420 | }
|
---|
421 |
|
---|
422 | $self->{'manifest'} =~ s/[\\\/]+/\//g;
|
---|
423 | $self->{'manifest'} =~ s/\/$//;
|
---|
424 |
|
---|
425 | $manifest_lookup->parse($manifest_filename);
|
---|
426 | }
|
---|
427 |
|
---|
428 | my $manifest = $self->{'manifest'};
|
---|
429 |
|
---|
430 | # load all the plugins
|
---|
431 | my $plugins = [];
|
---|
432 | if (defined $collectcfg->{'plugin'}) {
|
---|
433 | $plugins = $collectcfg->{'plugin'};
|
---|
434 | }
|
---|
435 |
|
---|
436 | my $plugin_incr_mode = $incremental_mode;
|
---|
437 | if ($manifest ne "") {
|
---|
438 | # if we have a manifest file, then we pretend we are fully incremental for plugins
|
---|
439 | $plugin_incr_mode = "all";
|
---|
440 | }
|
---|
441 | #some global options for the plugins
|
---|
442 | my @global_opts = ();
|
---|
443 |
|
---|
444 | my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode, $gs_version);
|
---|
445 | if (scalar(@$pluginfo) == 0) {
|
---|
446 | &gsprintf($out, "{import.no_plugins_loaded}\n");
|
---|
447 | die "\n";
|
---|
448 | }
|
---|
449 |
|
---|
450 | # remove the old contents of the archives directory (and tmp
|
---|
451 | # directory) if needed
|
---|
452 |
|
---|
453 | if ($removeold) {
|
---|
454 | if (-e $archivedir) {
|
---|
455 | &gsprintf($out, "{import.removing_archives}\n");
|
---|
456 | &util::rm_r ($archivedir);
|
---|
457 | }
|
---|
458 | my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
|
---|
459 | $tmpdir =~ s/[\\\/]+/\//g;
|
---|
460 | $tmpdir =~ s/\/$//;
|
---|
461 | if (-e $tmpdir) {
|
---|
462 | &gsprintf($out, "{import.removing_tmpdir}\n");
|
---|
463 | &util::rm_r ($tmpdir);
|
---|
464 | }
|
---|
465 | }
|
---|
466 |
|
---|
467 | # create the archives dir if needed
|
---|
468 | &util::mk_all_dir($archivedir);
|
---|
469 |
|
---|
470 | # read the archive information file
|
---|
471 |
|
---|
472 | # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
|
---|
473 | &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
|
---|
474 | &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
|
---|
475 |
|
---|
476 | my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir);
|
---|
477 | my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
|
---|
478 |
|
---|
479 | my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
|
---|
480 | $archive_info->load_info ($arcinfo_doc_filename);
|
---|
481 |
|
---|
482 | if ($manifest eq "") {
|
---|
483 | # Load in list of files in import folder from last import (if present)
|
---|
484 | $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
|
---|
485 | }
|
---|
486 |
|
---|
487 | ####Use Plugout####
|
---|
488 | my $plugout;
|
---|
489 |
|
---|
490 | if ($inexport_mode eq "import") {
|
---|
491 | if (defined $collectcfg->{'plugout'}) {
|
---|
492 | # If a plugout was specified in the collect.cfg file, assume it is sensible
|
---|
493 | # We can't check the name because it could be anything, if it is a custom plugout
|
---|
494 | $plugout = $collectcfg->{'plugout'};
|
---|
495 | }
|
---|
496 | else{
|
---|
497 | if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
|
---|
498 | push @$plugout,"GreenstoneXMLPlugout";
|
---|
499 | }
|
---|
500 | else{
|
---|
501 | push @$plugout,$saveas."Plugout";
|
---|
502 | }
|
---|
503 | }
|
---|
504 | }
|
---|
505 | else {
|
---|
506 | if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
|
---|
507 | $plugout = $collectcfg->{'plugout'};
|
---|
508 | }
|
---|
509 | else{
|
---|
510 | if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
|
---|
511 | push @$plugout,"GreenstoneMETSPlugout";
|
---|
512 | }
|
---|
513 | else{
|
---|
514 | push @$plugout,$saveas."Plugout";
|
---|
515 | }
|
---|
516 | }
|
---|
517 | }
|
---|
518 |
|
---|
519 | my $plugout_name = $plugout->[0];
|
---|
520 |
|
---|
521 | push @$plugout,("-output_info",$archive_info) if (defined $archive_info);
|
---|
522 | push @$plugout,("-verbosity",$verbosity) if (defined $verbosity);
|
---|
523 | push @$plugout,("-debug") if ($debug);
|
---|
524 | push @$plugout,("-group_size",$groupsize) if (defined $groupsize);
|
---|
525 | push @$plugout,("-gzip_output") if ($gzip);
|
---|
526 | push @$plugout,("-output_handle",$out) if (defined $out);
|
---|
527 |
|
---|
528 | push @$plugout,("-xslt_file",$xsltfile) if (defined $xsltfile && $xsltfile ne "");
|
---|
529 |
|
---|
530 | if ($plugout_name =~ m/^MARCXMLPlugout$/) {
|
---|
531 | push @$plugout,("-group") if ($group_marc);
|
---|
532 | push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
|
---|
533 | }
|
---|
534 | if ($plugout_name =~ m/^.*METSPlugout$/) {
|
---|
535 | push @$plugout,("-xslt_mets",$xslt_mets) if (defined $xslt_mets && $xslt_mets ne "");
|
---|
536 | push @$plugout,("-xslt_txt",$xslt_txt) if (defined $xslt_txt && $xslt_txt ne "");
|
---|
537 | }
|
---|
538 |
|
---|
539 | if ($plugout_name eq "FedoraMETSPlugout") {
|
---|
540 | push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
|
---|
541 | }
|
---|
542 |
|
---|
543 | if ($plugout_name eq "DSpacePlugout") {
|
---|
544 | push @$plugout,("-metadata_prefix",$metadata_prefix) if (defined $metadata_prefix && $metadata_prefix ne "");
|
---|
545 | }
|
---|
546 |
|
---|
547 | my $processor = &plugout::load_plugout($plugout);
|
---|
548 | $processor->setoutputdir ($archivedir);
|
---|
549 | $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
|
---|
550 | $processor->set_OIDtype ($OIDtype, $OIDmetadata);
|
---|
551 |
|
---|
552 | &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
|
---|
553 |
|
---|
554 | if ($removeold) {
|
---|
555 | # occasionally, plugins may want to do something on remove
|
---|
556 | # old, eg pharos image indexing
|
---|
557 | &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
|
---|
558 | }
|
---|
559 |
|
---|
560 | # process the import directory
|
---|
561 | my $block_hash = {};
|
---|
562 | $block_hash->{'new_files'} = {};
|
---|
563 | $block_hash->{'reindex_files'} = {};
|
---|
564 | my $metadata = {};
|
---|
565 |
|
---|
566 | # global blocking pass may set up some metadata
|
---|
567 | &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
|
---|
568 |
|
---|
569 | if ($manifest ne "") {
|
---|
570 | #
|
---|
571 | # 1. Process delete files first
|
---|
572 | #
|
---|
573 | my @deleted_files = keys %{$manifest_lookup->{'delete'}};
|
---|
574 | my @full_deleted_files = ();
|
---|
575 |
|
---|
576 | # ensure all filenames are absolute
|
---|
577 | foreach my $df (@deleted_files) {
|
---|
578 | my $full_df =
|
---|
579 | (&util::filename_is_absolute($df))
|
---|
580 | ? $df
|
---|
581 | : &util::filename_cat($importdir,$df);
|
---|
582 |
|
---|
583 | if (-d $full_df) {
|
---|
584 | &add_dir_contents_to_list($full_df, \@full_deleted_files);
|
---|
585 | } else {
|
---|
586 | push(@full_deleted_files,$full_df);
|
---|
587 | }
|
---|
588 | }
|
---|
589 |
|
---|
590 | &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
|
---|
591 | mark_docs_for_deletion($archive_info,{},
|
---|
592 | \@full_deleted_files,
|
---|
593 | $archivedir, $verbosity, "delete");
|
---|
594 |
|
---|
595 |
|
---|
596 | #
|
---|
597 | # 2. Now files for reindexing
|
---|
598 | #
|
---|
599 |
|
---|
600 | my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
|
---|
601 | my @full_reindex_files = ();
|
---|
602 | # ensure all filenames are absolute
|
---|
603 | foreach my $rf (@reindex_files) {
|
---|
604 | my $full_rf =
|
---|
605 | (&util::filename_is_absolute($rf))
|
---|
606 | ? $rf
|
---|
607 | : &util::filename_cat($importdir,$rf);
|
---|
608 |
|
---|
609 | if (-d $full_rf) {
|
---|
610 | &add_dir_contents_to_list($full_rf, \@full_reindex_files);
|
---|
611 | } else {
|
---|
612 | push(@full_reindex_files,$full_rf);
|
---|
613 | }
|
---|
614 | }
|
---|
615 |
|
---|
616 | &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
|
---|
617 | mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
|
---|
618 |
|
---|
619 | # And now to ensure the new version of the file processed by
|
---|
620 | # appropriate plugin, we need to add it to block_hash reindex list
|
---|
621 | foreach my $full_rf (@full_reindex_files) {
|
---|
622 | $block_hash->{'reindex_files'}->{$full_rf} = 1;
|
---|
623 | }
|
---|
624 |
|
---|
625 |
|
---|
626 | #
|
---|
627 | # 3. Now finally any new files - add to block_hash new_files list
|
---|
628 | #
|
---|
629 |
|
---|
630 | my @new_files = keys %{$manifest_lookup->{'index'}};
|
---|
631 | my @full_new_files = ();
|
---|
632 |
|
---|
633 | foreach my $nf (@new_files) {
|
---|
634 | # ensure filename is absolute
|
---|
635 | my $full_nf =
|
---|
636 | (&util::filename_is_absolute($nf))
|
---|
637 | ? $nf
|
---|
638 | : &util::filename_cat($importdir,$nf);
|
---|
639 |
|
---|
640 | if (-d $full_nf) {
|
---|
641 | &add_dir_contents_to_list($full_nf, \@full_new_files);
|
---|
642 | } else {
|
---|
643 | push(@full_new_files,$full_nf);
|
---|
644 | }
|
---|
645 | }
|
---|
646 |
|
---|
647 | my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
|
---|
648 | my $arcinfodb_map = {};
|
---|
649 | &dbutil::read_infodb_file($collectcfg->{'infodbtype'}, $arcinfo_src_filename, $arcinfodb_map);
|
---|
650 | foreach my $f (@full_new_files) {
|
---|
651 | # check that we haven't seen it already
|
---|
652 | if (defined $arcinfodb_map->{$f}) {
|
---|
653 | # TODO make better warning
|
---|
654 | print STDERR "Warning: $f already in src archive, \n";
|
---|
655 | } else {
|
---|
656 | $block_hash->{'new_files'}->{$f} = 1;
|
---|
657 | }
|
---|
658 | }
|
---|
659 |
|
---|
660 | undef $arcinfodb_map;
|
---|
661 | }
|
---|
662 | else {
|
---|
663 | # if incremental, we read through the import folder to see whats changed.
|
---|
664 |
|
---|
665 | if ($incremental || $incremental_mode eq "onlyadd") {
|
---|
666 | prime_doc_oid_count($archivedir);
|
---|
667 |
|
---|
668 | # Can now work out which files were new, already existed, and have
|
---|
669 | # been deleted
|
---|
670 |
|
---|
671 | new_vs_old_import_diff($archive_info,$block_hash,$importdir,
|
---|
672 | $archivedir,$verbosity,$incremental_mode);
|
---|
673 |
|
---|
674 | my @new_files = sort keys %{$block_hash->{'new_files'}};
|
---|
675 | if (scalar(@new_files>0)) {
|
---|
676 | print STDERR "New files and modified metadata files since last import:\n ";
|
---|
677 | print STDERR join("\n ",@new_files), "\n";
|
---|
678 | }
|
---|
679 |
|
---|
680 | if ($incremental) {
|
---|
681 | # only look for deletions if we are truely incremental
|
---|
682 | my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
|
---|
683 | # Filter out any in gsdl/tmp area
|
---|
684 | my @filtered_deleted_files = ();
|
---|
685 | my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
|
---|
686 | my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
|
---|
687 | $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
|
---|
688 | $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
|
---|
689 |
|
---|
690 | foreach my $df (@deleted_files) {
|
---|
691 | next if ($df =~ m/^$gsdl_tmp_area/);
|
---|
692 | next if ($df =~ m/^$collect_tmp_area/);
|
---|
693 |
|
---|
694 | push(@filtered_deleted_files,$df);
|
---|
695 | }
|
---|
696 |
|
---|
697 |
|
---|
698 | @deleted_files = @filtered_deleted_files;
|
---|
699 |
|
---|
700 | if (scalar(@deleted_files)>0) {
|
---|
701 | print STDERR "Files deleted since last import:\n ";
|
---|
702 | print STDERR join("\n ",@deleted_files), "\n";
|
---|
703 |
|
---|
704 |
|
---|
705 | &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
|
---|
706 |
|
---|
707 | mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
|
---|
708 | }
|
---|
709 |
|
---|
710 | my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
|
---|
711 |
|
---|
712 | if (scalar(@reindex_files)>0) {
|
---|
713 | print STDERR "Files to reindex since last import:\n ";
|
---|
714 | print STDERR join("\n ",@reindex_files), "\n";
|
---|
715 | &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
|
---|
716 | mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
|
---|
717 | }
|
---|
718 |
|
---|
719 | }
|
---|
720 | }
|
---|
721 | }
|
---|
722 |
|
---|
723 | # Check for existence of the file that's to contain earliestDateStamp in archivesdir
|
---|
724 | # Do nothing if the file already exists (file exists on incremental build).
|
---|
725 | # If the file doesn't exist, as happens on full build, create it and write out the current datestamp into it
|
---|
726 | # In buildcol, read the file's contents and set the earliestdateStamp in GS2's build.cfg / GS3's buildconfig.xml
|
---|
727 | # In doc.pm have set_oaiLastModified similar to set_lastmodified, and create the doc fields
|
---|
728 | # oailastmodified and oailastmodifieddate
|
---|
729 | my $earliestDatestampFile = &util::filename_cat($archivedir, "earliestDatestamp");
|
---|
730 | if (!-f $earliestDatestampFile && -d $archivedir) {
|
---|
731 | my $current_time_in_seconds = time; # in seconds
|
---|
732 |
|
---|
733 | if(open(FOUT, ">$earliestDatestampFile")) {
|
---|
734 | # || (&gsprintf(STDERR, "{common.cannot_open}: $!\n", $earliestDatestampFile) && die);
|
---|
735 | print FOUT $current_time_in_seconds;
|
---|
736 | close(FOUT);
|
---|
737 | }
|
---|
738 | else {
|
---|
739 | &gsprintf(STDERR, "{import.cannot_write_earliestdatestamp}\n", $earliestDatestampFile);
|
---|
740 | }
|
---|
741 |
|
---|
742 | }
|
---|
743 |
|
---|
744 | # now, whichever mode we are in, we can process the entire import folder
|
---|
745 | if ((defined $jobs) && ($jobs > 1))
|
---|
746 | {
|
---|
747 | # if jobs are set to >1, run in parallel using MPI helper
|
---|
748 | # [hs, 1 july 2010]
|
---|
749 | &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
|
---|
750 | $self->{'collection'}, $self->{'site'});
|
---|
751 | }
|
---|
752 | else
|
---|
753 | {
|
---|
754 | &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
|
---|
755 | }
|
---|
756 |
|
---|
757 |
|
---|
758 | if ($saveas eq "FedoraMETS") {
|
---|
759 | # create collection "doc obj" for Fedora that contains
|
---|
760 | # collection-level metadata
|
---|
761 |
|
---|
762 | my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
|
---|
763 | $doc_obj->set_OID("collection");
|
---|
764 |
|
---|
765 | my $col_name = undef;
|
---|
766 | my $col_meta = $collectcfg->{'collectionmeta'};
|
---|
767 |
|
---|
768 | if (defined $col_meta) {
|
---|
769 | store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
|
---|
770 | store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description
|
---|
771 | }
|
---|
772 | $processor->process($doc_obj);
|
---|
773 | }
|
---|
774 |
|
---|
775 | &plugin::end($pluginfo, $processor);
|
---|
776 |
|
---|
777 | &plugin::deinit($pluginfo, $processor);
|
---|
778 |
|
---|
779 | # Store the value of OIDCount (used in doc.pm) so it can be
|
---|
780 | # restored correctly to this value on an incremental build
|
---|
781 | store_doc_oid_count($archivedir);
|
---|
782 |
|
---|
783 | # write out the archive information file
|
---|
784 | $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
|
---|
785 | $processor->close_group_output() if $processor->is_group();
|
---|
786 |
|
---|
787 | # for backwards compatability with archvies.inf file
|
---|
788 | if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
|
---|
789 | $archive_info->save_info($arcinfo_doc_filename);
|
---|
790 | }
|
---|
791 | else {
|
---|
792 | $archive_info->save_revinfo_db($arcinfo_src_filename);
|
---|
793 | }
|
---|
794 |
|
---|
795 | return $pluginfo;
|
---|
796 | }
|
---|
797 |
|
---|
798 |
|
---|
799 | sub generate_statistics
|
---|
800 | {
|
---|
801 | my $self = shift @_;
|
---|
802 | my ($pluginfo) = @_;
|
---|
803 |
|
---|
804 | my $inexport_mode = $self->{'mode'};
|
---|
805 |
|
---|
806 | my $statsfile = $self->{'statsfile'};
|
---|
807 | my $out = $self->{'out'};
|
---|
808 | my $faillogname = $self->{'faillogname'};
|
---|
809 | my $gli = $self->{'gli'};
|
---|
810 | my $jobs = $self->{'jobs'};
|
---|
811 |
|
---|
812 | # write out import stats
|
---|
813 |
|
---|
814 | if ((!defined $jobs) || ($jobs == 1))
|
---|
815 | {
|
---|
816 | # only output statistics if there are multiple jobs
|
---|
817 | # [hs, 1 july 2010]
|
---|
818 |
|
---|
819 | my $close_stats = 0;
|
---|
820 | if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
|
---|
821 | if (open (STATS, ">$statsfile")) {
|
---|
822 | $statsfile = 'inexport::STATS';
|
---|
823 | $close_stats = 1;
|
---|
824 | } else {
|
---|
825 | &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
|
---|
826 | &gsprintf($out, "{import.stats_backup}\n");
|
---|
827 | $statsfile = 'STDERR';
|
---|
828 | }
|
---|
829 | }
|
---|
830 |
|
---|
831 | &gsprintf($out, "\n");
|
---|
832 | &gsprintf($out, "*********************************************\n");
|
---|
833 | &gsprintf($out, "{$inexport_mode.complete}\n");
|
---|
834 | &gsprintf($out, "*********************************************\n");
|
---|
835 |
|
---|
836 | &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
|
---|
837 | if ($close_stats) {
|
---|
838 | close STATS;
|
---|
839 | }
|
---|
840 | }
|
---|
841 |
|
---|
842 | close OUT if $self->{'close_out'};
|
---|
843 | close FAILLOG;
|
---|
844 | }
|
---|
845 |
|
---|
846 |
|
---|
847 | sub store_collectionmeta
|
---|
848 | {
|
---|
849 | my ($collectionmeta,$field,$doc_obj) = @_;
|
---|
850 |
|
---|
851 | my $section = $doc_obj->get_top_section();
|
---|
852 |
|
---|
853 | my $field_hash = $collectionmeta->{$field};
|
---|
854 |
|
---|
855 | foreach my $k (keys %$field_hash)
|
---|
856 | {
|
---|
857 | my $val = $field_hash->{$k};
|
---|
858 |
|
---|
859 | ### print STDERR "*** $k = $field_hash->{$k}\n";
|
---|
860 |
|
---|
861 | my $md_label = "ex.$field";
|
---|
862 |
|
---|
863 |
|
---|
864 | if ($k =~ m/^\[l=(.*?)\]$/)
|
---|
865 | {
|
---|
866 |
|
---|
867 | my $md_suffix = $1;
|
---|
868 | $md_label .= "^$md_suffix";
|
---|
869 | }
|
---|
870 |
|
---|
871 |
|
---|
872 | $doc_obj->add_utf8_metadata($section,$md_label, $val);
|
---|
873 |
|
---|
874 | # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
|
---|
875 | # while "collectionname" in GS2 is called "name" in GS3.
|
---|
876 | # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
|
---|
877 | if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
|
---|
878 | {
|
---|
879 | $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
|
---|
880 | }
|
---|
881 |
|
---|
882 | }
|
---|
883 | }
|
---|
884 |
|
---|
885 |
|
---|
886 | sub oid_count_file {
|
---|
887 | my ($archivedir) = @_;
|
---|
888 | return &util::filename_cat ($archivedir, "OIDcount");
|
---|
889 | }
|
---|
890 |
|
---|
891 |
|
---|
892 | sub prime_doc_oid_count
|
---|
893 | {
|
---|
894 | my ($archivedir) = @_;
|
---|
895 | my $oid_count_filename = &oid_count_file($archivedir);
|
---|
896 |
|
---|
897 | if (-e $oid_count_filename) {
|
---|
898 | if (open(OIDIN,"<$oid_count_filename")) {
|
---|
899 | my $OIDcount = <OIDIN>;
|
---|
900 | chomp $OIDcount;
|
---|
901 | close(OIDIN);
|
---|
902 |
|
---|
903 | $doc::OIDcount = $OIDcount;
|
---|
904 | }
|
---|
905 | else {
|
---|
906 | &gsprintf(STDERR, "{import.cannot_read_OIDcount}\n", $oid_count_filename);
|
---|
907 | }
|
---|
908 | }
|
---|
909 |
|
---|
910 | }
|
---|
911 |
|
---|
912 | sub store_doc_oid_count
|
---|
913 | {
|
---|
914 | # Use the file "OIDcount" in the archives directory to record
|
---|
915 | # what value doc.pm got up to
|
---|
916 |
|
---|
917 | my ($archivedir) = @_;
|
---|
918 | my $oid_count_filename = &oid_count_file($archivedir);
|
---|
919 |
|
---|
920 |
|
---|
921 | if (open(OIDOUT,">$oid_count_filename")) {
|
---|
922 | print OIDOUT $doc::OIDcount, "\n";
|
---|
923 |
|
---|
924 | close(OIDOUT);
|
---|
925 | }
|
---|
926 | else {
|
---|
927 | &gsprintf(STDERR, "{import.cannot_write_OIDcount}\n", $oid_count_filename);
|
---|
928 | }
|
---|
929 | }
|
---|
930 |
|
---|
931 |
|
---|
932 |
|
---|
933 | sub new_vs_old_import_diff
|
---|
934 | {
|
---|
935 | my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
|
---|
936 |
|
---|
937 | # Get the infodbtype value for this collection from the arcinfo object
|
---|
938 | my $infodbtype = $archive_info->{'infodbtype'};
|
---|
939 |
|
---|
940 | # in this method, we want to know if metadata files are modified or not.
|
---|
941 | my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
|
---|
942 |
|
---|
943 | my $archiveinf_timestamp = -M $arcinfo_doc_filename;
|
---|
944 |
|
---|
945 | # First convert all files to absolute form
|
---|
946 | # This is to support the situation where the import folder is not
|
---|
947 | # the default
|
---|
948 |
|
---|
949 | my $prev_all_files = $archive_info->{'prev_import_filelist'};
|
---|
950 | my $full_prev_all_files = {};
|
---|
951 |
|
---|
952 | foreach my $prev_file (keys %$prev_all_files) {
|
---|
953 |
|
---|
954 | if (!&util::filename_is_absolute($prev_file)) {
|
---|
955 | my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
|
---|
956 | $full_prev_all_files->{$full_prev_file} = $prev_file;
|
---|
957 | }
|
---|
958 | else {
|
---|
959 | $full_prev_all_files->{$prev_file} = $prev_file;
|
---|
960 | }
|
---|
961 | }
|
---|
962 |
|
---|
963 |
|
---|
964 | # Figure out which are the new files, existing files and so
|
---|
965 | # by implication the files from the previous import that are not
|
---|
966 | # there any more => mark them for deletion
|
---|
967 | foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
|
---|
968 |
|
---|
969 | my $full_curr_file = $curr_file;
|
---|
970 |
|
---|
971 | # entry in 'all_files' is moved to either 'existing_files',
|
---|
972 | # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
|
---|
973 |
|
---|
974 | if (!&util::filename_is_absolute($curr_file)) {
|
---|
975 | # add in import dir to make absolute
|
---|
976 | $full_curr_file = &util::filename_cat($importdir,$curr_file);
|
---|
977 | }
|
---|
978 |
|
---|
979 | # figure out if new file or not
|
---|
980 | if (defined $full_prev_all_files->{$full_curr_file}) {
|
---|
981 | # delete it so that only files that need deleting are left
|
---|
982 | delete $full_prev_all_files->{$full_curr_file};
|
---|
983 |
|
---|
984 | # had it before. is it a metadata file?
|
---|
985 | if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
|
---|
986 |
|
---|
987 | # is it modified??
|
---|
988 | if (-M $full_curr_file < $archiveinf_timestamp) {
|
---|
989 | print STDERR "*** Detected a *modified metadata* file: $full_curr_file\n" if $verbosity >= 2;
|
---|
990 | # its newer than last build
|
---|
991 | $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
|
---|
992 | }
|
---|
993 | }
|
---|
994 | else {
|
---|
995 | if ($incremental_mode eq "all") {
|
---|
996 |
|
---|
997 | # had it before
|
---|
998 | $block_hash->{'existing_files'}->{$full_curr_file} = 1;
|
---|
999 |
|
---|
1000 | }
|
---|
1001 | else {
|
---|
1002 | # Warning in "onlyadd" mode, but had it before!
|
---|
1003 | print STDERR "Warning: File $full_curr_file previously imported.\n";
|
---|
1004 | print STDERR " Treating as new file\n";
|
---|
1005 |
|
---|
1006 | $block_hash->{'new_files'}->{$full_curr_file} = 1;
|
---|
1007 |
|
---|
1008 | }
|
---|
1009 | }
|
---|
1010 | }
|
---|
1011 | else {
|
---|
1012 | if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
|
---|
1013 | # the new file is the special sort of file greenstone uses
|
---|
1014 | # to attach metadata to src documents
|
---|
1015 | # i.e metadata.xml
|
---|
1016 | # (but note, the filename used is not constrained in
|
---|
1017 | # Greenstone to always be this)
|
---|
1018 |
|
---|
1019 | print STDERR "*** Detected *new* metadata file: $full_curr_file\n" if $verbosity >= 2;
|
---|
1020 | $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
|
---|
1021 | }
|
---|
1022 | else {
|
---|
1023 | $block_hash->{'new_files'}->{$full_curr_file} = 1;
|
---|
1024 | }
|
---|
1025 | }
|
---|
1026 |
|
---|
1027 |
|
---|
1028 | delete $block_hash->{'all_files'}->{$curr_file};
|
---|
1029 | }
|
---|
1030 |
|
---|
1031 |
|
---|
1032 |
|
---|
1033 |
|
---|
1034 | # Deal with complication of new or modified metadata files by forcing
|
---|
1035 | # everything from this point down in the file hierarchy to
|
---|
1036 | # be freshly imported.
|
---|
1037 | #
|
---|
1038 | # This may mean files that have not changed are reindexed, but does
|
---|
1039 | # guarantee by the end of processing all new metadata is correctly
|
---|
1040 | # associated with the relevant document(s).
|
---|
1041 |
|
---|
1042 | foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
|
---|
1043 | my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
|
---|
1044 |
|
---|
1045 | $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
|
---|
1046 | $situated_dir = &util::filename_to_regex($situated_dir); # need to escape windows slash \ and brackets in regular expression
|
---|
1047 |
|
---|
1048 | # Go through existing_files, and mark anything that is contained
|
---|
1049 | # within 'situated_dir' to be reindexed (in case some of the metadata
|
---|
1050 | # attaches to one of these files)
|
---|
1051 |
|
---|
1052 | my $reindex_files = [];
|
---|
1053 |
|
---|
1054 | foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
|
---|
1055 |
|
---|
1056 | if ($existing_f =~ m/^$situated_dir/) {
|
---|
1057 |
|
---|
1058 | print STDERR "**** Existing file $existing_f\nis located within\n$situated_dir\n";
|
---|
1059 |
|
---|
1060 | push(@$reindex_files,$existing_f);
|
---|
1061 | $block_hash->{'reindex_files'}->{$existing_f} = 1;
|
---|
1062 | delete $block_hash->{'existing_files'}->{$existing_f};
|
---|
1063 |
|
---|
1064 | }
|
---|
1065 | }
|
---|
1066 |
|
---|
1067 | # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
|
---|
1068 | # (or equivalent)
|
---|
1069 | $block_hash->{'new_files'}->{$new_mdf} = 1;
|
---|
1070 |
|
---|
1071 | }
|
---|
1072 |
|
---|
1073 | # go through remaining existing files and work out what has changed and needs to be reindexed.
|
---|
1074 | my @existing_files = sort keys %{$block_hash->{'existing_files'}};
|
---|
1075 |
|
---|
1076 | my $reindex_files = [];
|
---|
1077 |
|
---|
1078 | foreach my $existing_filename (@existing_files) {
|
---|
1079 | if (-M $existing_filename < $archiveinf_timestamp) {
|
---|
1080 | # file is newer than last build
|
---|
1081 |
|
---|
1082 | my $existing_file = $existing_filename;
|
---|
1083 | #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
|
---|
1084 |
|
---|
1085 | #my $collectdir_resafe = &util::filename_to_regex($collectdir);
|
---|
1086 | #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
|
---|
1087 |
|
---|
1088 | print STDERR "**** Reindexing existing file: $existing_file\n";
|
---|
1089 |
|
---|
1090 | push(@$reindex_files,$existing_file);
|
---|
1091 | $block_hash->{'reindex_files'}->{$existing_filename} = 1;
|
---|
1092 | }
|
---|
1093 |
|
---|
1094 | }
|
---|
1095 |
|
---|
1096 |
|
---|
1097 | # By this point full_prev_all_files contains the files
|
---|
1098 | # mentioned in archiveinf-src.db but are not in the 'import'
|
---|
1099 | # folder (or whatever was specified through -importdir ...)
|
---|
1100 |
|
---|
1101 | # This list can contain files that were created in the 'tmp' or
|
---|
1102 | # 'cache' areas (such as screen-size and thumbnail images).
|
---|
1103 | #
|
---|
1104 | # In building the final list of files to delete, we test to see if
|
---|
1105 | # it exists on the filesystem and if it does (unusual for a "normal"
|
---|
1106 | # file in import, but possible in the case of 'tmp' files),
|
---|
1107 | # supress it from going into the final list
|
---|
1108 |
|
---|
1109 | my $collectdir = $ENV{'GSDLCOLLECTDIR'};
|
---|
1110 |
|
---|
1111 | my @deleted_files = values %$full_prev_all_files;
|
---|
1112 | map { my $curr_file = $_;
|
---|
1113 | my $full_curr_file = $curr_file;
|
---|
1114 |
|
---|
1115 | if (!&util::filename_is_absolute($curr_file)) {
|
---|
1116 | # add in import dir to make absolute
|
---|
1117 |
|
---|
1118 | $full_curr_file = &util::filename_cat($collectdir,$curr_file);
|
---|
1119 | }
|
---|
1120 |
|
---|
1121 |
|
---|
1122 | if (!-e $full_curr_file) {
|
---|
1123 | $block_hash->{'deleted_files'}->{$curr_file} = 1;
|
---|
1124 | }
|
---|
1125 | } @deleted_files;
|
---|
1126 |
|
---|
1127 |
|
---|
1128 |
|
---|
1129 | }
|
---|
1130 |
|
---|
1131 |
|
---|
1132 | # this is used to delete "deleted" docs, and to remove old versions of "changed" docs
|
---|
1133 | # $mode is 'delete' or 'reindex'
|
---|
1134 | sub mark_docs_for_deletion
|
---|
1135 | {
|
---|
1136 | my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
|
---|
1137 |
|
---|
1138 | my $mode_text = "deleted from index";
|
---|
1139 | if ($mode eq "reindex") {
|
---|
1140 | $mode_text = "reindexed";
|
---|
1141 | }
|
---|
1142 |
|
---|
1143 | # Get the infodbtype value for this collection from the arcinfo object
|
---|
1144 | my $infodbtype = $archive_info->{'infodbtype'};
|
---|
1145 |
|
---|
1146 | my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
|
---|
1147 | my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
|
---|
1148 |
|
---|
1149 |
|
---|
1150 | # record files marked for deletion in arcinfo
|
---|
1151 | foreach my $file (@$deleted_files) {
|
---|
1152 | # use 'archiveinf-src' info database file to look up all the OIDs
|
---|
1153 | # that this file is used in (note in most cases, it's just one OID)
|
---|
1154 |
|
---|
1155 | my $src_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
|
---|
1156 | my $oids = $src_rec->{'oid'};
|
---|
1157 | my $file_record_deleted = 0;
|
---|
1158 |
|
---|
1159 | # delete the src record
|
---|
1160 | my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
|
---|
1161 | &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
|
---|
1162 | &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
|
---|
1163 |
|
---|
1164 |
|
---|
1165 | foreach my $oid (@$oids) {
|
---|
1166 |
|
---|
1167 | # find the source doc (the primary file that becomes this oid)
|
---|
1168 | my $doc_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
|
---|
1169 | my $doc_source_file = $doc_rec->{'src-file'}->[0];
|
---|
1170 | if (!&util::filename_is_absolute($doc_source_file)) {
|
---|
1171 | $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
|
---|
1172 | }
|
---|
1173 |
|
---|
1174 | if ($doc_source_file ne $file) {
|
---|
1175 | # its an associated or metadata file
|
---|
1176 |
|
---|
1177 | # mark source doc for reimport as one of its assoc files has changed or deleted
|
---|
1178 | $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
|
---|
1179 |
|
---|
1180 | }
|
---|
1181 | my $curr_status = $archive_info->get_status_info($oid);
|
---|
1182 | if (defined($curr_status) && (($curr_status ne "D"))) {
|
---|
1183 | if ($verbosity>1) {
|
---|
1184 | print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
|
---|
1185 | }
|
---|
1186 | # mark oid for deletion (it will be deleted or reimported)
|
---|
1187 | $archive_info->set_status_info($oid,"D");
|
---|
1188 | my $val = &dbutil::read_infodb_rawentry($infodbtype, $arcinfo_doc_filename, $oid);
|
---|
1189 | $val =~ s/^<index-status>(.*)$/<index-status>D/m;
|
---|
1190 |
|
---|
1191 | my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
|
---|
1192 | my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
|
---|
1193 |
|
---|
1194 | &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
|
---|
1195 | &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
|
---|
1196 | }
|
---|
1197 | }
|
---|
1198 |
|
---|
1199 | }
|
---|
1200 |
|
---|
1201 | # now go through and check that we haven't marked any primary
|
---|
1202 | # files for reindex (because their associated files have
|
---|
1203 | # changed/deleted) when they have been deleted themselves. only in
|
---|
1204 | # delete mode.
|
---|
1205 |
|
---|
1206 | if ($mode eq "delete") {
|
---|
1207 | foreach my $file (@$deleted_files) {
|
---|
1208 | if (defined $block_hash->{'reindex_files'}->{$file}) {
|
---|
1209 | delete $block_hash->{'reindex_files'}->{$file};
|
---|
1210 | }
|
---|
1211 | }
|
---|
1212 | }
|
---|
1213 |
|
---|
1214 |
|
---|
1215 | }
|
---|
1216 |
|
---|
1217 | sub add_dir_contents_to_list {
|
---|
1218 |
|
---|
1219 | my ($dirname, $list) = @_;
|
---|
1220 |
|
---|
1221 | # Recur over directory contents.
|
---|
1222 | my (@dir, $subfile);
|
---|
1223 |
|
---|
1224 | # find all the files in the directory
|
---|
1225 | if (!opendir (DIR, $dirname)) {
|
---|
1226 | print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
|
---|
1227 | return -1; # error in processing
|
---|
1228 | }
|
---|
1229 | @dir = readdir (DIR);
|
---|
1230 | closedir (DIR);
|
---|
1231 |
|
---|
1232 | for (my $i = 0; $i < scalar(@dir); $i++) {
|
---|
1233 | my $subfile = $dir[$i];
|
---|
1234 | next if ($subfile =~ m/^\.\.?$/);
|
---|
1235 | next if ($subfile =~ /^\.svn$/);
|
---|
1236 | my $full_file = &util::filename_cat($dirname, $subfile);
|
---|
1237 | if (-d $full_file) {
|
---|
1238 | &add_dir_contents_to_list($full_file, $list);
|
---|
1239 | } else {
|
---|
1240 | push (@$list, $full_file);
|
---|
1241 | }
|
---|
1242 | }
|
---|
1243 |
|
---|
1244 | }
|
---|
1245 |
|
---|
1246 |
|
---|
1247 | 1;
|
---|