root/gs2-extensions/parallel-building/trunk/src/perllib/inexport.pm @ 30292

Revision 30292, 47.9 KB (checked in by jmt12, 5 years ago)

Removing reference to debugging module Devel::Peek

  • Property svn:executable set to *
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;
49use Scalar::Util 'blessed';
50
51my $oidtype_list =
52    [ { 'name' => "hash",
53        'desc' => "{import.OIDtype.hash}" },
54      { 'name' => "hash_on_full_filename",
55        'desc' => "{import.OIDtype.hash_on_full_filename}" },
56      { 'name' => "assigned",
57        'desc' => "{import.OIDtype.assigned}" },
58      { 'name' => "incremental",
59        'desc' => "{import.OIDtype.incremental}" },
60      { 'name' => "filename",
61        'desc' => "{import.OIDtype.filename}" },
62      { 'name' => "dirname",
63        'desc' => "{import.OIDtype.dirname}" },
64      { 'name' => "full_filename",
65        'desc' => "{import.OIDtype.full_filename}" } ];
66
67$inexport::directory_arguments =
68[
69      { 'name' => "importdir",
70    'desc' => "{import.importdir}",
71    'type' => "string",
72    'reqd' => "no",
73    'deft' => "import",
74        'hiddengli' => "yes" },
75      { 'name' => "collectdir",
76    'desc' => "{import.collectdir}",
77    'type' => "string",
78    # parsearg left "" as default
79    #'deft' => &FileUtils::filenameConcatenate($ENV{'GSDLHOME'}, "collect"),
80    'deft' => "",
81    'reqd' => "no",
82        'hiddengli' => "yes" },
83 
84];
85$inexport::arguments =
86[
87      # don't set the default to hash - want to allow this to come from
88      # entry in collect.cfg but want to override it here
89      { 'name' => "OIDtype",
90    'desc' => "{import.OIDtype}",
91    'type' => "enum",
92    'list' => $oidtype_list,
93    'deft' => "hash_on_full_filename",
94    'reqd' => "no",
95    'modegli' => "2" },
96      { 'name' => "OIDmetadata",
97    'desc' => "{import.OIDmetadata}",
98    'type' => "string",
99    'deft' => "dc.Identifier",
100    'reqd' => "no",
101    'modegli' => "2" },
102      { 'name' => "site",
103    'desc' => "{import.site}",
104    'type' => "string",
105    'deft' => "",
106    'reqd' => "no",
107        'hiddengli' => "yes" },
108      { 'name' => "manifest",
109    'desc' => "{import.manifest}",
110    'type' => "string",
111    'deft' => "",
112    'reqd' => "no",
113        'hiddengli' => "yes" } ,
114     { 'name' => "incremental",
115    'desc' => "{import.incremental}",
116    'type' => "flag",
117    'hiddengli' => "yes" },
118      { 'name' => "keepold",
119    'desc' => "{import.keepold}",
120    'type' => "flag",
121    'reqd' => "no",
122    'hiddengli' => "yes" },
123      { 'name' => "removeold",
124    'desc' => "{import.removeold}",
125    'type' => "flag",
126    'reqd' => "no",
127    'hiddengli' => "yes" },
128      { 'name' => "language",
129    'desc' => "{scripts.language}",
130    'type' => "string",
131    'reqd' => "no",
132    'hiddengli' => "yes" },
133      { 'name' => "maxdocs",
134    'desc' => "{import.maxdocs}",
135    'type' => "int",
136    'reqd' => "no",
137    'deft' => "-1",
138    'range' => "-1,",
139    'modegli' => "1" },
140       { 'name' => "debug",
141    'desc' => "{import.debug}",
142    'type' => "flag",
143    'reqd' => "no",
144        'hiddengli' => "yes" },
145      { 'name' => "faillog",
146    'desc' => "{import.faillog}",
147    'type' => "string",
148    # parsearg left "" as default
149    #'deft' => &FileUtils::filenameConcatenate("<collectdir>", "colname", "etc", "fail.log"),
150    'deft' => "",
151    'reqd' => "no",
152        'modegli' => "3" },
153       { 'name' => "out",
154    'desc' => "{import.out}",
155    'type' => "string",
156    'deft' => "STDERR",
157    'reqd' => "no",
158        'hiddengli' => "yes" },
159      { 'name' => "statsfile",
160    'desc' => "{import.statsfile}",
161    'type' => "string",
162    'deft' => "STDERR",
163    'reqd' => "no",
164        'hiddengli' => "yes" },
165      { 'name' => "verbosity",
166    'desc' => "{import.verbosity}",
167    'type' => "int",
168    'range' => "0,",
169    'deft' => "2",
170    'reqd' => "no",
171    'modegli' => "3" },
172      { 'name' => "gli",
173    'desc' => "{scripts.gli}",
174    'type' => "flag",
175    'reqd' => "no",
176    'hiddengli' => "yes" },
177      { 'name' => "xml",
178    'desc' => "{scripts.xml}",
179    'type' => "flag",
180    'reqd' => "no",
181    'hiddengli' => "yes" },
182
183];
184
185sub new
186{
187    my $class = shift (@_);
188    my ($mode,$argv,$options,$opt_listall_options) = @_;
189
190    my $self = { 'xml' => 0, 'mode' => $mode };
191
192    print "INFO: This inexport.pm supports version 2 manifest files\n";
193
194    # general options available to all plugins
195    my $arguments = $options->{'args'};
196    my $intArgLeftinAfterParsing = parse2::parse($argv,$arguments,$self,"allow_extra_options");
197    # Parse returns -1 if something has gone wrong
198    if ($intArgLeftinAfterParsing == -1)
199    {
200    &PrintUsage::print_txt_usage($options, "{import.params}",1);
201    print STDERR "Something went wrong during parsing the arguments. Scroll up for details.\n";
202    die "\n";
203    }
204
205    my $language = $self->{'language'};
206    # If $language has been specified, load the appropriate resource bundle
207    # (Otherwise, the default resource bundle will be loaded automatically)
208    if ($language && $language =~ /\S/) {
209    &gsprintf::load_language_specific_resource_bundle($language);
210    }
211
212    if ($self->{'listall'}) {
213    if ($self->{'xml'}) {
214        &PrintUsage::print_xml_usage($opt_listall_options);
215    }
216    else
217    {
218        &PrintUsage::print_txt_usage($opt_listall_options,"{export.params}");
219    }
220    die "\n";
221    }
222
223    if ($self->{'xml'}) {
224        &PrintUsage::print_xml_usage($options);
225    print "\n";
226    return bless $self, $class;
227    }
228
229    if ($self->{'gli'}) { # the gli wants strings to be in UTF-8
230    &gsprintf::output_strings_in_UTF8;
231    }
232
233    # If the user specified -h, then we output the usage
234    if (@$argv && $argv->[0] =~ /^\-+h/) {
235    &PrintUsage::print_txt_usage($options, "{import.params}");
236    die "\n";
237    }
238    # now check that we had exactly one leftover arg, which should be
239    # the collection name. We don't want to do this earlier, cos
240    # -xml arg doesn't need a collection name
241
242    if ($intArgLeftinAfterParsing != 1 )
243    {
244    &PrintUsage::print_txt_usage($options, "{import.params}", 1);
245    print STDERR "There should be one argument left after parsing the script args: the collection name.\n";
246    die "\n";
247    }
248
249    $self->{'close_out'} = 0;
250    my $out = $self->{'out'};
251    if ($out !~ /^(STDERR|STDOUT)$/i) {
252    open (OUT, ">$out") ||
253        (&gsprintf(STDERR, "{common.cannot_open_output_file}: $!\n", $out) && die);
254    $out = 'inexport::OUT';
255    $self->{'close_out'} = 1;
256    }
257    $out->autoflush(1);
258    $self->{'out'} = $out;
259
260    my $statsfile = $self->{'statsfile'};
261    if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
262    open (STATSFILE, ">$statsfile") ||
263        (&gsprintf(STDERR, "{common.cannot_open_output_file}: $!\n", $statsfile) && die);
264    $statsfile = 'inexport::STATSFILE';
265    $self->{'close_stats'} = 1;
266    }
267    $statsfile->autoflush(1);
268    $self->{'statsfile'} = $statsfile;
269
270    # @ARGV should be only one item, the name of the collection
271    $self->{'collection'} = shift @$argv;
272
273    # Unless otherwise stated all manifests are considered version 1---where
274    # they act more like an advanced process expression---as compared to newer
275    # manifest files that act as an explicit (and exhaustive) list of files to
276    # process [jmt12]
277    $self->{'manifest_version'} = 1;
278
279    return bless $self, $class;
280}
281
282# Simplified version of the contstructor for use with CGI scripts
283sub newCGI
284{
285    my $class = shift (@_);
286    my ($mode,$collect,$gsdl_cgi,$opt_site) = @_;
287
288    my $self = { 'xml' => 0, 'mode' => $mode };
289
290    $self->{'out'} = STDERR;
291   
292    if (defined $gsdl_cgi) {
293        $self->{'site'} = $opt_site;
294        my $collect_dir = $gsdl_cgi->get_collection_dir($opt_site);
295        $self->{'collectdir'} = $collect_dir;
296    }
297    else { 
298        $self->{'site'} = "";
299        $self->{'collectdir'} = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"collect");
300    }
301    $self->{'faillog'} = "";
302   
303    $self->{'collection'} = $collect;
304
305    return bless $self, $class;
306}
307sub get_collection
308{
309    my $self = shift @_;
310   
311    return $self->{'collection'};
312}
313
314
315sub read_collection_cfg
316{
317    my $self = shift @_;
318    my ($collection,$options) = @_;
319
320    my $collectdir = $self->{'collectdir'};
321    my $site       = $self->{'site'};
322    my $out        = $self->{'out'};
323     
324    if (($collection = &colcfg::use_collection($site, $collection, $collectdir)) eq "") {
325    #&PrintUsage::print_txt_usage($options, "{import.params}", 1);
326    die "\n";
327    }
328
329    # set gs_version 2/3
330    $self->{'gs_version'} = "2";
331    if ((defined $site) && ($site ne "")) {
332    # gs3
333    $self->{'gs_version'} = "3";
334    }
335
336    # add collection's perllib dir into include path in
337    # case we have collection specific modules
338    &util::augmentINC(&FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, 'perllib'));
339
340    # check that we can open the faillog
341    my $faillog = $self->{'faillog'};
342    if ($faillog eq "") {
343    $faillog = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
344    }
345    open (FAILLOG, ">$faillog") ||
346    (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
347
348   
349    my $faillogname = $faillog;
350    $faillog = 'inexport::FAILLOG';
351    $faillog->autoflush(1);
352    $self->{'faillog'} = $faillog;
353    $self->{'faillogname'} = $faillogname;
354    $self->{'close_faillog'} = 1;
355
356    # Read in the collection configuration file.
357    my $gs_mode = "gs".$self->{'gs_version'}; #gs2 or gs3
358    my $config_filename = &colcfg::get_collect_cfg_name($out, $gs_mode);
359    my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
360
361    return ($config_filename,$collectcfg);
362}
363
364sub set_collection_options
365{
366    my $self = shift @_;
367    my ($collectcfg) = @_;
368
369    my $inexport_mode = $self->{'mode'};
370
371    my $importdir  = $self->{'importdir'};
372    my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
373    my $out        = $self->{'out'};
374
375    # If the infodbtype value wasn't defined in the collect.cfg file, use the default
376    if (!defined($collectcfg->{'infodbtype'}))
377    {
378      $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
379    }
380    if ($collectcfg->{'infodbtype'} eq "gdbm-txtgz") {
381    # we can't use the text version for archives dbs.
382    $collectcfg->{'infodbtype'} = "gdbm";
383    }
384
385    if (defined $self->{'default_importdir'} && defined $collectcfg->{'importdir'}) {
386    $importdir = $collectcfg->{'importdir'};
387    }
388
389    if ($inexport_mode eq "import") {
390    if ( defined $self->{'default_archivedir'} && defined $collectcfg->{'archivedir'}) {
391        $archivedir = $collectcfg->{'archivedir'};
392    }
393    }
394    elsif ($inexport_mode eq "export") {
395    if (defined $self->{'default_exportdir'} && defined $collectcfg->{'exportdir'}) {
396        $archivedir = $collectcfg->{'exportdir'};
397    }
398    }
399    # fill in the default import and archives directories if none
400    # were supplied, turn all \ into / and remove trailing /
401    if (!&FileUtils::isFilenameAbsolute($importdir))
402    {
403      $importdir = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, $importdir);
404    }
405    else
406    {
407      # Don't do this - it kills protocol prefixes
408      #$importdir =~ s/[\\\/]+/\//g;
409      #$importdir =~ s/\/$//;
410      # Do this instead
411      &FileUtils::sanitizePath($importdir);
412    }
413    if (!&FileUtils::directoryExists($importdir))
414    {
415      &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
416      die "\n";
417    }
418    $self->{'importdir'} = $importdir;
419
420    if (!&FileUtils::isFilenameAbsolute($archivedir)) {
421    $archivedir = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, $archivedir);
422    }
423    else {
424   
425    $archivedir = &FileUtils::sanitizePath($archivedir);
426    }
427    $self->{'archivedir'} = $archivedir;
428
429    if (defined $self->{'default_verbosity'}) {
430    if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
431        $self->{'verbosity'} = $collectcfg->{'verbosity'};
432    }
433    }
434 
435    if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
436    $self->{'manifest'} = $collectcfg->{'manifest'};
437    }
438
439    if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
440    if ($collectcfg->{'gzip'} =~ /^true$/i) {
441        $self->{'gzip'} = 1;
442    }
443    }
444
445    if (defined $self->{'default_maxdocs'}) {
446    if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
447        $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
448    }
449    }
450
451   
452
453    if (defined $self->{'default_OIDtype'} ) {
454    if (defined $collectcfg->{'OIDtype'}
455        && $collectcfg->{'OIDtype'} =~ /^(hash|hash_on_full_filename|incremental|assigned|filename|dirname|full_filename)$/) {
456        $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
457    }
458    }
459
460    if (defined $self->{'default_OIDmetadata'}) {
461    if (defined $collectcfg->{'OIDmetadata'}) {
462        $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
463    }
464    }
465
466    if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
467    $self->{'debug'} = 1;
468    }
469    if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
470    $self->{'gli'} = 1;
471    }
472    $self->{'gli'} = 0 unless defined $self->{'gli'};
473       
474    # check keepold and removeold
475    my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
476
477    my ($removeold, $keepold, $incremental, $incremental_mode)
478    = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
479                           $self->{'incremental'}, $checkdir,
480                           $collectcfg);
481
482    $self->{'removeold'}        = $removeold;
483    $self->{'keepold'}          = $keepold;
484    $self->{'incremental'}      = $incremental;
485    $self->{'incremental_mode'} = $incremental_mode;
486
487    # We'll need direct access to this plugin to support v2 manifests
488    $self->{'directoryplugin'} = 0;
489}
490
491sub process_files
492{
493    my $self = shift @_;
494    my ($config_filename,$collectcfg) = @_;
495
496    my $inexport_mode = $self->{'mode'};
497
498    my $verbosity   = $self->{'verbosity'};
499    my $debug       = $self->{'debug'};
500
501    my $importdir   = $self->{'importdir'};
502    my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
503
504    my $incremental = $self->{'incremental'};
505    my $incremental_mode = $self->{'incremental_mode'};
506
507    my $gs_version = $self->{'gs_version'};
508
509    my $removeold   = $self->{'removeold'};
510    my $keepold     = $self->{'keepold'};
511
512    my $saveas      = $self->{'saveas'};
513    my $saveas_options = $self->{'saveas_options'};
514    my $OIDtype     = $self->{'OIDtype'};
515    my $OIDmetadata = $self->{'OIDmetadata'};
516
517    my $out         = $self->{'out'};
518    my $faillog     = $self->{'faillog'};
519
520    my $maxdocs     = $self->{'maxdocs'};
521    my $gzip        = $self->{'gzip'};
522    my $groupsize   = $self->{'groupsize'};
523    my $sortmeta    = $self->{'sortmeta'};
524
525    my $removeprefix = $self->{'removeprefix'};
526    my $removesuffix = $self->{'removesuffix'};
527
528    my $gli          = $self->{'gli'};
529
530    # related to export
531    my $xsltfile         = $self->{'xsltfile'};
532    my $group_marc       = $self->{'group_marc'};
533    my $mapping_file     = $self->{'mapping_file'};
534    my $xslt_mets        = $self->{'xslt_mets'};
535    my $xslt_txt         = $self->{'xslt_txt'};
536    my $fedora_namespace = $self->{'fedora_namespace'};
537    my $metadata_prefix  = $self->{'metadata_prefix'};
538
539    if ($inexport_mode eq "import") {
540    print STDERR "<Import>\n" if $gli;
541    }
542    else {
543    print STDERR "<export>\n" if $gli;
544    }
545
546    my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
547    if ($self->{'manifest'} ne "") {
548    my $manifest_filename = $self->{'manifest'};
549
550    if (!&FileUtils::isFilenameAbsolute($manifest_filename)) {
551        $manifest_filename = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
552    }
553        $self->{'manifest'} = &FileUtils::sanitizePath($self->{'manifest'});
554    #$self->{'manifest'} =~ s/[\\\/]+/\//g;
555    #$self->{'manifest'} =~ s/\/$//;
556
557    $manifest_lookup->parse($manifest_filename);
558
559        # manifests may now include a version number [jmt12]
560        $self->{'manifest_version'} = $manifest_lookup->get_version();
561    }
562
563    my $manifest = $self->{'manifest'};
564
565    # load all the plugins
566    my $plugins = [];
567    if (defined $collectcfg->{'plugin'}) {
568    $plugins = $collectcfg->{'plugin'};
569    }
570
571    my $plugin_incr_mode = $incremental_mode;
572    if ($manifest ne "") {
573    # if we have a manifest file, then we pretend we are fully incremental for plugins
574    $plugin_incr_mode = "all";
575    }
576    #some global options for the plugins
577    my @global_opts = ();
578
579    my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode, $gs_version);
580    if (scalar(@$pluginfo) == 0) {
581    &gsprintf($out, "{import.no_plugins_loaded}\n");
582    die "\n";
583    }
584    # Store a reference to the DirectoryPlugin
585    foreach my $a_plugin (@{$pluginfo})
586    {
587    if (blessed ($a_plugin) eq 'DirectoryPlugin')
588    {
589        $self->{'directoryplugin'} = $a_plugin;
590    }
591    }
592    # No directory plugin - no v2 manifest support
593    if ($self->{'directoryplugin'} == 0)
594    {
595    print STDERR "WARNING: DirectoryPlugin not loaded: metadata.xml files not supported.\n";
596    }
597
598    # remove the old contents of the archives directory (and tmp
599    # directory) if needed
600
601    if ($removeold) {
602    if (&FileUtils::directoryExists($archivedir)) {
603        &gsprintf($out, "{import.removing_archives}\n");
604        &FileUtils::removeFilesRecursive($archivedir);
605    }
606    my $tmpdir = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, "tmp");
607    $tmpdir =~ s/[\\\/]+/\//g;
608    $tmpdir =~ s/\/$//;
609    if (&FileUtils::directoryExists($tmpdir)) {
610        &gsprintf($out, "{import.removing_tmpdir}\n");
611        &FileUtils::removeFilesRecursive($tmpdir);
612    }
613    }
614
615    # create the archives dir if needed
616    &FileUtils::makeAllDirectories($archivedir);
617
618    # read the archive information file
619
620    # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
621    &util::rename_ldb_or_bdb_file(&FileUtils::filenameConcatenate($archivedir, "archiveinf-doc"));
622    &util::rename_ldb_or_bdb_file(&FileUtils::filenameConcatenate($archivedir, "archiveinf-src"));
623
624    # When we make these initial calls to determine the archive information doc
625    # and src databases we pass through a '1' to indicate this is the first
626    # time we are referring to these databases. When using dynamic dbutils
627    # (available in extensions) this indicates to some database types (for
628    # example, persistent servers) that this is a good time to perform any
629    # one time initialization. The argument has no effect on vanilla dbutils
630    # [jmt12]
631    my $perform_firsttime_init = 1;
632    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir, $perform_firsttime_init);
633    my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir, $perform_firsttime_init);
634
635    my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
636    $archive_info->load_info ($arcinfo_doc_filename);
637
638    if ($manifest eq "") {
639    # Load in list of files in import folder from last import (if present)
640    $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
641    }
642
643    ####Use Plugout####
644    my $plugout;
645
646    my $generate_auxiliary_files = 0;
647    if ($inexport_mode eq "import") {
648    $generate_auxiliary_files = 1;
649    }
650    elsif ($self->{'include_auxiliary_database_files'}) {
651    $generate_auxiliary_files = 1;
652    }
653    $self->{'generate_auxiliary_files'} = $generate_auxiliary_files;
654
655    # Option to use user defined plugout
656    if ($inexport_mode eq "import") {
657    if (defined $collectcfg->{'plugout'}) {
658        # If a plugout was specified in the collect.cfg file, assume it is sensible
659        # We can't check the name because it could be anything, if it is a custom plugout
660        print STDERR "Using plugout specified in collect.cfg: ".join(' ', @{$collectcfg->{'plugout'}})."\n";
661        $plugout = $collectcfg->{'plugout'};
662    }
663    else {
664        push @$plugout,$saveas."Plugout";
665    }
666
667    }
668    else {
669    if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(GreenstoneXML|.*METS|DSpace|MARCXML)Plugout/) {
670        $plugout = $collectcfg->{'plugout'};
671        print STDERR "Using plugout specified in collect.cfg: $collectcfg->{'plugout'}\n";
672    }
673    else {
674        push @$plugout,$saveas."Plugout";
675    }
676    }
677
678    my $plugout_name = $plugout->[0];
679
680    if ($inexport_mode eq "export" && defined $saveas_options) {
681    my @user_plugout_options = split(" ", $saveas_options);
682    push @$plugout, @user_plugout_options;
683    }
684    push @$plugout,("-output_info",$archive_info)  if (defined $archive_info);
685    push @$plugout,("-verbosity",$verbosity)       if (defined $verbosity);
686    push @$plugout,("-debug")                      if ($debug);
687    push @$plugout,("-gzip_output")                if ($gzip);
688    push @$plugout,("-output_handle",$out)         if (defined $out);
689
690    push @$plugout,("-xslt_file",$xsltfile)        if (defined $xsltfile && $xsltfile ne "");
691    push @$plugout, ("-no_auxiliary_databases") if ($generate_auxiliary_files == 0);
692    if ($inexport_mode eq "import") {
693    if ($plugout_name =~ m/^GreenstoneXMLPlugout$/) {
694        push @$plugout,("-group_size",$groupsize)      if (defined $groupsize);
695    }
696    }
697    my $processor = &plugout::load_plugout($plugout);
698    $processor->setoutputdir ($archivedir);
699    $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
700    $processor->set_OIDtype ($OIDtype, $OIDmetadata);
701    $processor->begin();
702    &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
703   
704    if ($removeold) {
705        # occasionally, plugins may want to do something on remove
706        # old, eg pharos image indexing
707    &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
708    }
709
710    # process the import directory
711    my $block_hash = {};
712    $block_hash->{'new_files'} = {};
713    $block_hash->{'reindex_files'} = {};
714    # all of these are set somewhere else, so it's more readable to define them
715    # here [jmt12]
716    $block_hash->{'all_files'} = {};
717    $block_hash->{'deleted_files'} = {};
718    $block_hash->{'file_blocks'} = {};
719    $block_hash->{'metadata_files'} = {};
720    $block_hash->{'shared_fileroot'} = '';
721    # a new flag so we can tell we had a manifest way down in the plugins
722    # [jmt12]
723    $block_hash->{'manifest'} = 'false';
724    my $metadata = {};
725   
726    # global blocking pass may set up some metadata
727    # does this set up metadata?????
728    # - when we have a newer manifest file we don't do this -unless- the
729    #   collection configuration indicates this collection contains complex
730    #   (inherited) metadata [jmt12]
731    if ($manifest eq '' || (defined $collectcfg->{'complexmeta'} && $collectcfg->{'complexmeta'} eq 'true'))
732    {
733      &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
734    }
735    else
736    {
737      print STDERR "Skipping global file scan due to manifest and complexmeta configuration\n";
738    }
739
740    if ($manifest ne "") {
741
742      # mark that we are using a manifest - information that might be needed
743      # down in plugins (for instance DirectoryPlugin)
744      $block_hash->{'manifest'} = $self->{'manifest_version'};
745
746    #
747    # 1. Process delete files first
748    #
749    my @deleted_files = keys %{$manifest_lookup->{'delete'}};
750    my @full_deleted_files = ();
751
752    # ensure all filenames are absolute
753    foreach my $df (@deleted_files) {
754        my $full_df =
755        (&FileUtils::isFilenameAbsolute($df))
756        ? $df
757        : &FileUtils::filenameConcatenate($importdir,$df);
758
759        if (-d $full_df) {
760        &add_dir_contents_to_list($full_df, \@full_deleted_files);
761        } else {
762        push(@full_deleted_files,$full_df);
763        }
764    }
765   
766    &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
767    mark_docs_for_deletion($archive_info,{},
768                   \@full_deleted_files,
769                   $archivedir, $verbosity, "delete");
770
771
772    #
773    # 2. Now files for reindexing
774    #
775
776    my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
777    my @full_reindex_files = ();
778    # ensure all filenames are absolute
779    foreach my $rf (@reindex_files) {       
780        my $full_rf =
781        (&FileUtils::isFilenameAbsolute($rf))
782        ? $rf
783        : &FileUtils::filenameConcatenate($importdir,$rf);
784
785        if (-d $full_rf) {
786        &add_dir_contents_to_list($full_rf, \@full_reindex_files);
787        } else {
788        push(@full_reindex_files,$full_rf);
789        }
790    }
791   
792    &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
793    mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
794
795    # And now to ensure the new version of the file processed by
796    # appropriate plugin, we need to add it to block_hash reindex list
797    foreach my $full_rf (@full_reindex_files) {
798        $block_hash->{'reindex_files'}->{$full_rf} = 1;
799    }
800
801
802    #
803    # 3. Now finally any new files - add to block_hash new_files list
804    #
805
806    my @new_files = keys %{$manifest_lookup->{'index'}};
807    my @full_new_files = ();
808
809    foreach my $nf (@new_files) {
810        # ensure filename is absolute
811        my $full_nf =
812        (&FileUtils::isFilenameAbsolute($nf))
813        ? $nf
814        : &FileUtils::filenameConcatenate($importdir,$nf);
815
816        if (-d $full_nf) {
817        &add_dir_contents_to_list($full_nf, \@full_new_files);
818        } else {
819        push(@full_new_files,$full_nf);
820        }
821    }
822
823    my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
824      # need to check this file exists before trying to read it - in the past
825      # it wasn't possible to have a manifest unless keepold was also set so
826      # you were pretty much guaranteed arcinfo existed
827      # [jmt12]
828      # @todo &FileUtils::fileExists($arcinfo_src_filename) [jmt12]
829      if (-e $arcinfo_src_filename)
830      {
831    my $arcinfodb_map = {};
832    &dbutil::read_infodb_file($collectcfg->{'infodbtype'}, $arcinfo_src_filename, $arcinfodb_map);
833    foreach my $f (@full_new_files) {
834        my $rel_f = &util::abspath_to_placeholders($f);
835
836        # check that we haven't seen it already
837        if (defined $arcinfodb_map->{$rel_f}) {
838        # TODO make better warning
839        print STDERR "Warning: $f ($rel_f) already in src archive, \n";
840        } else {
841        $block_hash->{'new_files'}->{$f} = 1;
842        }
843    }
844
845    undef $arcinfodb_map;
846      }
847      # no existing files - so we can just add all the files [jmt12]
848      else
849      {
850        foreach my $f (@full_new_files)
851        {
852          $block_hash->{'new_files'}->{$f} = 1;
853        }
854      }
855
856      # If we are not using complex inherited metadata (and thus have skipped
857      # the global file scan) we need to at least check for a matching
858      # metadata.xml for the files being indexed/reindexed
859      # - unless we are using the newer version of Manifests, which are treated
860      #   verbatim, and should have a metadata element for metadata files (so
861      #   we can explicitly process metadata files other than metadata.xml)
862      # [jmt12]
863      if ($self->{'manifest_version'} == 1 && (!defined $collectcfg->{'complexmeta'} || $collectcfg->{'complexmeta'} ne 'true'))
864      {
865        my @all_files_to_import = (keys %{$block_hash->{'reindex_files'}}, keys %{$block_hash->{'new_files'}});
866        foreach my $file_to_import (@all_files_to_import)
867        {
868          my $metadata_xml_path = $file_to_import;
869          $metadata_xml_path =~ s/[^\\\/]*$/metadata.xml/;
870          if (&FileUtils::fileExists($metadata_xml_path))
871          {
872            &plugin::file_block_read($pluginfo, '', $metadata_xml_path, $block_hash, $metadata, $gli);
873          }
874        }
875      }
876
877      # new version manifest files explicitly list files to be processed and
878      # only support 'simplemeta' format (ignoring complexmeta if set) in that
879      # each document can be accompanied by a metadata.xml file in the same
880      # directory. The metadata.xml can only apply to the fileset ".*".
881      # [jmt12]
882      if ($self->{'manifest_version'} > 1)
883      {
884        # Process metadata files
885        foreach my $file_to_import (keys %{$block_hash->{'reindex_files'}}, keys %{$block_hash->{'new_files'}})
886        {
887        $self->{'directoryplugin'}->read_for_manifest_v2($pluginfo, $file_to_import, $block_hash, $processor, $gli);
888        }
889      }
890    }
891    else {
892    # if incremental, we read through the import folder to see whats changed.
893
894    if ($incremental || $incremental_mode eq "onlyadd") {
895        prime_doc_oid_count($archivedir);
896
897        # Can now work out which files were new, already existed, and have
898        # been deleted
899       
900        new_vs_old_import_diff($archive_info,$block_hash,$importdir,
901                   $archivedir,$verbosity,$incremental_mode);
902       
903        my @new_files = sort keys %{$block_hash->{'new_files'}};
904        if (scalar(@new_files>0)) {
905        print STDERR "New files and modified metadata files since last import:\n  ";
906        print STDERR join("\n  ",@new_files), "\n";
907        }
908
909        if ($incremental) {
910               # only look for deletions if we are truely incremental
911        my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
912        # Filter out any in gsdl/tmp area
913        my @filtered_deleted_files = ();
914        my $gsdl_tmp_area = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'}, "tmp");
915        my $collect_tmp_area = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'}, "tmp");
916        $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
917        $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
918                 
919        foreach my $df (@deleted_files) {
920            next if ($df =~ m/^$gsdl_tmp_area/);
921            next if ($df =~ m/^$collect_tmp_area/);
922           
923            push(@filtered_deleted_files,$df);
924        }       
925       
926
927        @deleted_files = @filtered_deleted_files;
928       
929        if (scalar(@deleted_files)>0) {
930            print STDERR "Files deleted since last import:\n  ";
931            print STDERR join("\n  ",@deleted_files), "\n";
932       
933       
934            &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
935           
936            mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
937        }
938       
939        my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
940       
941        if (scalar(@reindex_files)>0) {
942            print STDERR "Files to reindex since last import:\n  ";
943            print STDERR join("\n  ",@reindex_files), "\n";
944            &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
945            mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
946        }
947               
948        }       
949    }
950    }
951
952    # Check for existence of the file that's to contain earliestDateStamp in archivesdir
953    # Do nothing if the file already exists (file exists on incremental build).
954    # If the file doesn't exist, as happens on full build, create it and write out the current datestamp into it
955    # In buildcol, read the file's contents and set the earliestdateStamp in GS2's build.cfg / GS3's buildconfig.xml
956    # In doc.pm have set_oaiLastModified similar to set_lastmodified, and create the doc fields
957    # oailastmodified and oailastmodifieddate
958    my $earliestDatestampFile = &FileUtils::filenameConcatenate($archivedir, "earliestDatestamp");
959    if ($self->{'generate_auxiliary_files'}) {
960    if (!-f $earliestDatestampFile && -d $archivedir) {
961    my $current_time_in_seconds = time; # in seconds
962
963    if(open(FOUT, ">$earliestDatestampFile")) {
964        # || (&gsprintf(STDERR, "{common.cannot_open}: $!\n", $earliestDatestampFile) && die);
965        print FOUT $current_time_in_seconds;
966        close(FOUT);
967    }
968    else {
969        &gsprintf(STDERR, "{import.cannot_write_earliestdatestamp}\n", $earliestDatestampFile);
970    }
971
972    }
973    }
974
975    if ($self->{'manifest_version'} != 2)
976    {
977    $self->perform_process_files($manifest, $pluginfo, $importdir, '', $block_hash, $metadata, $processor, $maxdocs);
978    }
979   
980    if ($saveas eq "FedoraMETS") {
981    # create collection "doc obj" for Fedora that contains
982    # collection-level metadata
983   
984    my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
985    $doc_obj->set_OID("collection");
986   
987    my $col_name = undef;
988    my $col_meta = $collectcfg->{'collectionmeta'};
989   
990    if (defined $col_meta) {       
991        store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
992        store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description     
993    }
994    $processor->process($doc_obj);
995    }
996
997    &plugin::end($pluginfo, $processor);
998
999    &plugin::deinit($pluginfo, $processor);
1000
1001    # Store the value of OIDCount (used in doc.pm) so it can be
1002    # restored correctly to this value on an incremental build
1003    # - this OIDcount file should only be generated for numerical oids [jmt12]
1004    if ($self->{'OIDtype'} eq 'incremental')
1005    {
1006      store_doc_oid_count($archivedir);
1007    }
1008
1009    # signal to the processor (plugout) that we have finished processing - if we are group processing, then the final output file needs closing.
1010    $processor->close_group_output() if $processor->is_group();
1011
1012#    if ($inexport_mode eq "import") {
1013    if ($self->{'generate_auxiliary_files'}) {
1014    # write out the archive information file
1015    # for backwards compatability with archvies.inf file
1016    if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
1017        $archive_info->save_info($arcinfo_doc_filename);
1018    }
1019    else {
1020        $archive_info->save_revinfo_db($arcinfo_src_filename);
1021    }
1022    }
1023    return $pluginfo;
1024}
1025
1026# @function perform_process_files()
1027# while process_files() above prepares the system to import files this is the
1028# function that actually initiates the plugin pipeline to process the files.
1029# This function the therefore be overridden in subclasses of inexport.pm should
1030# they wish to do different or further processing
1031# @author jmt12
1032sub perform_process_files
1033{
1034  my $self = shift(@_);
1035  my ($manifest, $pluginfo, $importdir, $file_to_import, $block_hash, $metadata, $processor, $maxdocs) = @_;
1036  my $gli = $self->{'gli'};
1037  # specific file to process - via manifest version 2+
1038  if ($file_to_import ne '')
1039  {
1040    &plugin::read ($pluginfo, '', $file_to_import, $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
1041  }
1042  # global file scan - if we are using a new version manifest, files would have
1043  # been read above. Older manifests use extra settings in the $block_hash to
1044  # control what is imported, while non-manifest imports use a regular
1045  # $block_hash (so obeying process_exp and block_exp) [jmt12]
1046  elsif ($manifest eq '' || $self->{'manifest_version'} == 1)
1047  {
1048    &plugin::read ($pluginfo, $importdir, '', $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
1049  }
1050  else
1051  {
1052    print STDERR "Skipping perform_process_files() due to manifest presence and version\n";
1053  }
1054}
1055# perform_process_files()
1056
1057# @function generate_statistics()
1058sub generate_statistics
1059{
1060  my $self = shift @_;
1061  my ($pluginfo) = @_;
1062
1063  my $inexport_mode = $self->{'mode'};
1064  my $out           = $self->{'out'};
1065  my $faillogname   = $self->{'faillogname'};
1066  my $statsfile     = $self->{'statsfile'};
1067  my $gli           = $self->{'gli'};
1068
1069  &gsprintf($out, "\n");
1070  &gsprintf($out, "*********************************************\n");
1071  &gsprintf($out, "{$inexport_mode.complete}\n");
1072  &gsprintf($out, "*********************************************\n");
1073
1074  &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
1075}
1076# generate_statistics()
1077
1078
1079# @function deinit()
1080# Close down any file handles that we opened (and hence are responsible for
1081# closing
1082sub deinit
1083{
1084  my $self = shift(@_);
1085  close OUT if $self->{'close_out'};
1086  close FAILLOG if $self->{'close_faillog'};
1087  close STATSFILE if $self->{'close_statsfile'};
1088}
1089# deinit()
1090
1091
1092sub store_collectionmeta
1093{
1094    my ($collectionmeta,$field,$doc_obj) = @_;
1095   
1096    my $section = $doc_obj->get_top_section();
1097   
1098    my $field_hash = $collectionmeta->{$field};
1099   
1100    foreach my $k (keys %$field_hash)
1101    {
1102    my $val = $field_hash->{$k};
1103   
1104    ### print STDERR "*** $k = $field_hash->{$k}\n";
1105   
1106    my $md_label = "ex.$field";
1107   
1108   
1109    if ($k =~ m/^\[l=(.*?)\]$/)
1110    {
1111       
1112        my $md_suffix = $1;
1113        $md_label .= "^$md_suffix";
1114    }
1115   
1116   
1117    $doc_obj->add_utf8_metadata($section,$md_label, $val);
1118   
1119    # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
1120    # while "collectionname" in GS2 is called "name" in GS3.
1121    # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
1122    if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
1123    {
1124        $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
1125    }
1126   
1127    }
1128}
1129
1130
1131sub oid_count_file {
1132    my ($archivedir) = @_;
1133    return &FileUtils::filenameConcatenate($archivedir, "OIDcount");
1134}
1135
1136
1137sub prime_doc_oid_count
1138{
1139    my ($archivedir) = @_;
1140    my $oid_count_filename = &oid_count_file($archivedir);
1141
1142    if (-e $oid_count_filename) {
1143    if (open(OIDIN,"<$oid_count_filename")) {
1144        my $OIDcount = <OIDIN>;
1145        chomp $OIDcount;       
1146        close(OIDIN);
1147
1148        $doc::OIDcount = $OIDcount;     
1149    }
1150    else {     
1151        &gsprintf(STDERR, "{import.cannot_read_OIDcount}\n", $oid_count_filename);
1152    }
1153    }
1154   
1155}
1156
1157sub store_doc_oid_count
1158{
1159    # Use the file "OIDcount" in the archives directory to record
1160    # what value doc.pm got up to
1161
1162    my ($archivedir) = @_;
1163    my $oid_count_filename = &oid_count_file($archivedir);
1164
1165    # @todo $oidout = &FileUtils::openFileDescriptor($oid_count_filename, 'w') [jmt12]
1166    if (open(OIDOUT,">$oid_count_filename")) {
1167    print OIDOUT $doc::OIDcount, "\n";
1168       
1169    close(OIDOUT);
1170    }
1171    else {
1172    &gsprintf(STDERR, "{import.cannot_write_OIDcount}\n", $oid_count_filename);
1173    }
1174}
1175
1176
1177
1178sub new_vs_old_import_diff
1179{
1180    my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
1181
1182    # Get the infodbtype value for this collection from the arcinfo object
1183    my $infodbtype = $archive_info->{'infodbtype'};
1184
1185    # in this method, we want to know if metadata files are modified or not.
1186    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1187
1188    my $archiveinf_timestamp = -M $arcinfo_doc_filename;
1189
1190    # First convert all files to absolute form
1191    # This is to support the situation where the import folder is not
1192    # the default
1193   
1194    my $prev_all_files = $archive_info->{'prev_import_filelist'};
1195    my $full_prev_all_files = {};
1196
1197    foreach my $prev_file (keys %$prev_all_files) {
1198
1199    if (!&FileUtils::isFilenameAbsolute($prev_file)) {
1200        my $full_prev_file = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'},$prev_file);
1201        $full_prev_all_files->{$full_prev_file} = $prev_file;
1202    }
1203    else {
1204        $full_prev_all_files->{$prev_file} = $prev_file;
1205    }
1206    }
1207
1208
1209    # Figure out which are the new files, existing files and so
1210    # by implication the files from the previous import that are not
1211    # there any more => mark them for deletion
1212    foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
1213   
1214    my $full_curr_file = $curr_file;
1215
1216    # entry in 'all_files' is moved to either 'existing_files',
1217    # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
1218
1219    if (!&FileUtils::isFilenameAbsolute($curr_file)) {
1220        # add in import dir to make absolute
1221        $full_curr_file = &FileUtils::filenameConcatenate($importdir,$curr_file);
1222    }
1223
1224    # figure out if new file or not
1225    if (defined $full_prev_all_files->{$full_curr_file}) {
1226        # delete it so that only files that need deleting are left
1227        delete $full_prev_all_files->{$full_curr_file};
1228       
1229        # had it before. is it a metadata file?
1230        if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
1231       
1232        # is it modified??
1233        if (-M $full_curr_file < $archiveinf_timestamp) {
1234            print STDERR "*** Detected a *modified metadata* file: $full_curr_file\n" if $verbosity >= 2;
1235            # its newer than last build
1236            $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
1237        }
1238        }
1239        else {
1240        if ($incremental_mode eq "all") {
1241           
1242            # had it before
1243            $block_hash->{'existing_files'}->{$full_curr_file} = 1;
1244           
1245        }
1246        else {
1247            # Warning in "onlyadd" mode, but had it before!
1248            print STDERR "Warning: File $full_curr_file previously imported.\n";
1249            print STDERR "         Treating as new file\n";
1250           
1251            $block_hash->{'new_files'}->{$full_curr_file} = 1;
1252           
1253        }
1254        }
1255    }
1256    else {
1257        if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
1258        # the new file is the special sort of file greenstone uses
1259        # to attach metadata to src documents
1260        # i.e metadata.xml
1261        # (but note, the filename used is not constrained in
1262        # Greenstone to always be this)
1263
1264        print STDERR "*** Detected *new* metadata file: $full_curr_file\n" if $verbosity >= 2;
1265        $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
1266        }
1267        else {
1268        $block_hash->{'new_files'}->{$full_curr_file} = 1;
1269        }
1270    }
1271
1272   
1273    delete $block_hash->{'all_files'}->{$curr_file};
1274    }
1275
1276
1277
1278
1279    # Deal with complication of new or modified metadata files by forcing
1280    # everything from this point down in the file hierarchy to
1281    # be freshly imported. 
1282    #
1283    # This may mean files that have not changed are reindexed, but does
1284    # guarantee by the end of processing all new metadata is correctly
1285    # associated with the relevant document(s).
1286
1287    foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
1288    my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
1289
1290    $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
1291    $situated_dir = &util::filename_to_regex($situated_dir); # need to escape windows slash \ and brackets in regular expression
1292   
1293    # Go through existing_files, and mark anything that is contained
1294    # within 'situated_dir' to be reindexed (in case some of the metadata
1295    # attaches to one of these files)
1296
1297    my $reindex_files = [];
1298
1299    foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
1300   
1301        if ($existing_f =~ m/^$situated_dir/) {
1302
1303        print STDERR "**** Existing file $existing_f\nis located within\n$situated_dir\n";
1304
1305        push(@$reindex_files,$existing_f);
1306        $block_hash->{'reindex_files'}->{$existing_f} = 1;
1307        delete $block_hash->{'existing_files'}->{$existing_f};
1308
1309        }
1310    }
1311   
1312    # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
1313    # (or equivalent)
1314    $block_hash->{'new_files'}->{$new_mdf} = 1;
1315
1316    }
1317
1318    # go through remaining existing files and work out what has changed and needs to be reindexed.
1319    my @existing_files = sort keys %{$block_hash->{'existing_files'}};
1320
1321    my $reindex_files = [];
1322
1323    foreach my $existing_filename (@existing_files) {
1324    if (-M $existing_filename < $archiveinf_timestamp) {
1325        # file is newer than last build
1326       
1327        my $existing_file = $existing_filename;
1328        #my $collectdir = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'});
1329
1330        #my $collectdir_resafe = &util::filename_to_regex($collectdir);
1331        #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
1332       
1333        print STDERR "**** Reindexing existing file: $existing_file\n";
1334
1335        push(@$reindex_files,$existing_file);
1336        $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1337    }
1338
1339    }
1340
1341   
1342    # By this point full_prev_all_files contains the files
1343    # mentioned in archiveinf-src.db but are not in the 'import'
1344    # folder (or whatever was specified through -importdir ...)
1345
1346    # This list can contain files that were created in the 'tmp' or
1347    # 'cache' areas (such as screen-size and thumbnail images).
1348    #
1349    # In building the final list of files to delete, we test to see if
1350    # it exists on the filesystem and if it does (unusual for a "normal"
1351    # file in import, but possible in the case of 'tmp' files),
1352    # supress it from going into the final list
1353
1354    my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1355
1356    my @deleted_files = values %$full_prev_all_files;
1357    map { my $curr_file = $_;
1358      my $full_curr_file = $curr_file;
1359
1360      if (!&FileUtils::isFilenameAbsolute($curr_file)) {
1361          # add in import dir to make absolute
1362
1363          $full_curr_file = &FileUtils::filenameConcatenate($collectdir,$curr_file);
1364      }
1365
1366
1367      if (!-e $full_curr_file) {
1368          $block_hash->{'deleted_files'}->{$curr_file} = 1;
1369      }
1370      } @deleted_files;
1371
1372
1373
1374}
1375
1376
1377# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1378# $mode is 'delete' or 'reindex'
1379sub mark_docs_for_deletion
1380{
1381    my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1382
1383    my $mode_text = "deleted from index";
1384    if ($mode eq "reindex") {
1385    $mode_text = "reindexed";
1386    }
1387
1388    # Get the infodbtype value for this collection from the arcinfo object
1389    my $infodbtype = $archive_info->{'infodbtype'};
1390
1391    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1392    my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1393
1394
1395    # record files marked for deletion in arcinfo
1396    foreach my $file (@$deleted_files) {
1397    # use 'archiveinf-src' info database file to look up all the OIDs
1398    # that this file is used in (note in most cases, it's just one OID)
1399   
1400    my $relfile = &util::abspath_to_placeholders($file);
1401
1402    my $src_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $relfile);
1403    my $oids = $src_rec->{'oid'};
1404    my $file_record_deleted = 0;
1405
1406    # delete the src record
1407    my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1408    &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $relfile);
1409    &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1410
1411
1412    foreach my $oid (@$oids) {
1413
1414        # find the source doc (the primary file that becomes this oid)
1415        my $doc_rec = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1416        my $doc_source_file = $doc_rec->{'src-file'}->[0];
1417        $doc_source_file = &util::placeholders_to_abspath($doc_source_file);
1418
1419        if (!&FileUtils::isFilenameAbsolute($doc_source_file)) {
1420        $doc_source_file = &FileUtils::filenameConcatenate($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1421        }
1422
1423        if ($doc_source_file ne $file) {
1424        # its an associated or metadata file
1425       
1426        # mark source doc for reimport as one of its assoc files has changed or deleted
1427        $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1428       
1429        }
1430        my $curr_status = $archive_info->get_status_info($oid);
1431        if (defined($curr_status) && (($curr_status ne "D"))) {
1432        if ($verbosity>1) {
1433            print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1434        }
1435        # mark oid for deletion (it will be deleted or reimported)
1436        $archive_info->set_status_info($oid,"D");
1437        my $val = &dbutil::read_infodb_rawentry($infodbtype, $arcinfo_doc_filename, $oid);
1438        $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1439
1440        my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1441        my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1442
1443        &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1444        &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1445        }
1446    }
1447   
1448    }
1449
1450    # now go through and check that we haven't marked any primary
1451    # files for reindex (because their associated files have
1452    # changed/deleted) when they have been deleted themselves. only in
1453    # delete mode.
1454
1455    if ($mode eq "delete") {
1456    foreach my $file (@$deleted_files) {
1457        if (defined $block_hash->{'reindex_files'}->{$file}) {
1458        delete $block_hash->{'reindex_files'}->{$file};
1459        }
1460    }
1461    }
1462
1463
1464}
1465
1466sub add_dir_contents_to_list {
1467
1468    my ($dirname, $list) = @_;
1469 
1470    # Recur over directory contents.
1471    my (@dir, $subfile);
1472   
1473    # find all the files in the directory
1474    if (!opendir (DIR, $dirname)) {
1475    print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
1476    return -1; # error in processing
1477    }
1478    @dir = readdir (DIR);
1479    closedir (DIR);
1480   
1481    for (my $i = 0; $i < scalar(@dir); $i++) {
1482    my $subfile = $dir[$i];
1483    next if ($subfile =~ m/^\.\.?$/);
1484    next if ($subfile =~ /^\.svn$/);
1485    my $full_file = &FileUtils::filenameConcatenate($dirname, $subfile);
1486    if (-d $full_file) {
1487        &add_dir_contents_to_list($full_file, $list);
1488    } else {
1489        push (@$list, $full_file);
1490    }
1491    }
1492   
1493}
1494
1495   
14961;
Note: See TracBrowser for help on using the browser.