root/main/trunk/greenstone2/perllib/inexport.pm @ 23053

Revision 23053, 36.0 KB (checked in by kjdon, 9 years ago)

reworking of manifest stuff. Now, instead of calling plugin::read on each file in the reindex/new_files lists, we add all the filenames into the standard block_hash. This is the equivalent of what the code for incremental build does, except that instead of inspecting the import folder, we just use what is listed in manifest file. If a directory is listed, then we look through the directory and add each file in there to the appropriate list. Once block_hash is set up, then we tell the plugins we are incremental, and just process the import folder as per usual. Only those docs in new_files and reindex lists in block_hash will be processed. But this gets a full metadata_read done, which we need in case there is metadata specified up a level for the files that are changed/new.

  • 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;
49
50sub new
51{
52    my $class = shift (@_);
53    my ($mode,$argv,$options,$opt_listall_options) = @_;
54
55    my $self = { 'xml' => 0, 'mode' => $mode };
56
57    # general options available to all plugins
58    my $arguments = $options->{'args'};
59    my $intArgLeftinAfterParsing = parse2::parse($argv,$arguments,$self,"allow_extra_options");
60    # Parse returns -1 if something has gone wrong
61    if ($intArgLeftinAfterParsing == -1)
62    {
63    &PrintUsage::print_txt_usage($options, "{import.params}");
64    die "\n";
65    }
66   
67    my $language = $self->{'language'};
68    # If $language has been specified, load the appropriate resource bundle
69    # (Otherwise, the default resource bundle will be loaded automatically)
70    if ($language && $language =~ /\S/) {
71    &gsprintf::load_language_specific_resource_bundle($language);
72    }
73
74    if ($self->{'listall'}) {
75    if ($self->{'xml'}) {
76        &PrintUsage::print_xml_usage($opt_listall_options);
77    }
78    else
79    {
80        &PrintUsage::print_txt_usage($opt_listall_options,"{export.params}");
81    }
82    die "\n";
83    }
84
85
86    if ($self->{'xml'}) {
87        &PrintUsage::print_xml_usage($options);
88    print "\n";
89    return bless $self, $class;
90    }
91
92    if ($self->{'gli'}) { # the gli wants strings to be in UTF-8
93    &gsprintf::output_strings_in_UTF8;
94    }
95   
96    # now check that we had exactly one leftover arg, which should be
97    # the collection name. We don't want to do this earlier, cos
98    # -xml arg doesn't need a collection name
99    # Or if the user specified -h, then we output the usage also
100
101    if ($intArgLeftinAfterParsing != 1 || (@$argv && $argv->[0] =~ /^\-+h/))
102    {
103    &PrintUsage::print_txt_usage($options, "{import.params}");
104    die "\n";
105    }
106
107    $self->{'close_out'} = 0;
108    my $out = $self->{'out'};
109    if ($out !~ /^(STDERR|STDOUT)$/i) {
110    open (OUT, ">$out") ||
111        (&gsprintf(STDERR, "{common.cannot_open_output_file}: $!\n", $out) && die);
112    $out = 'inexport::OUT';
113    $self->{'close_out'} = 1;
114    }
115    $out->autoflush(1);
116    $self->{'out'} = $out;
117
118    # @ARGV should be only one item, the name of the collection
119    $self->{'collection'} = shift @$argv;
120
121    if ((defined $self->{'jobs'}) && ($self->{'jobs'}>1)) {
122    require ParallelInexport;
123    }
124
125    return bless $self, $class;
126}
127
128sub get_collection
129{
130    my $self = shift @_;
131   
132    return $self->{'collection'};
133}
134
135
136sub read_collection_cfg
137{
138    my $self = shift @_;
139    my ($collection,$options) = @_;
140
141    my $collectdir = $self->{'collectdir'};
142    my $site       = $self->{'site'};
143    my $out        = $self->{'out'};
144 
145    if (($collection = &colcfg::use_collection($site, $collection, $collectdir)) eq "") {
146    &PrintUsage::print_txt_usage($options, "{import.params}");
147    die "\n";
148    }
149
150    # add collection's perllib dir  into include path in
151    # case we have collection specific modules
152    unshift (@INC, "$ENV{'GSDLCOLLECTDIR'}/perllib");
153
154    # check that we can open the faillog
155    my $faillog = $self->{'faillog'};
156    if ($faillog eq "") {
157    $faillog = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "etc", "fail.log");
158    }
159    open (FAILLOG, ">$faillog") ||
160    (&gsprintf(STDERR, "{import.cannot_open_fail_log}\n", $faillog) && die);
161
162   
163    my $faillogname = $faillog;
164    $faillog = 'inexport::FAILLOG';
165    $faillog->autoflush(1);
166    $self->{'faillog'} = $faillog;
167    $self->{'faillogname'} = $faillogname;
168
169    # Read in the collection configuration file.
170    my ($config_filename, $gs_mode) = &colcfg::get_collect_cfg_name($out);
171    my $collectcfg = &colcfg::read_collection_cfg ($config_filename, $gs_mode);
172
173    return ($config_filename,$collectcfg);
174}
175
176sub set_collection_options
177{
178    my $self = shift @_;
179    my ($collectcfg) = @_;
180
181    my $inexport_mode = $self->{'mode'};
182
183    my $verbosity  = $self->{'verbosity'};
184    my $debug      = $self->{'debug'};
185    my $importdir  = $self->{'importdir'};
186    my $archivedir = $self->{'archivedir'} || $self->{'exportdir'} || "";
187    my $out        = $self->{'out'};
188
189    # If the infodbtype value wasn't defined in the collect.cfg file, use the default
190    if (!defined($collectcfg->{'infodbtype'}))
191    {
192      $collectcfg->{'infodbtype'} = &dbutil::get_default_infodb_type();
193    }
194
195    if (defined $collectcfg->{'importdir'} && $importdir eq "") {
196    $importdir = $collectcfg->{'importdir'};
197    }
198    if (defined $collectcfg->{'archivedir'} && $archivedir eq "") {
199    $archivedir = $collectcfg->{'archivedir'};
200    }
201    # fill in the default import and archives directories if none
202    # were supplied, turn all \ into / and remove trailing /
203    $importdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "import") if $importdir eq "";
204    $importdir =~ s/[\\\/]+/\//g;
205    $importdir =~ s/\/$//;
206    if (!-e $importdir) {
207    &gsprintf($out, "{import.no_import_dir}\n\n", $importdir);
208    die "\n";
209    }
210    $self->{'importdir'} = $importdir;
211
212    if ($archivedir eq "") {
213    if ($inexport_mode eq "import") {
214        $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
215    }
216    elsif ($inexport_mode eq "export") {
217        $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "export");
218    }
219    else {
220        print STDERR "Warning: Unrecognized import/export mode '$inexport_mode'\n";
221        print STDERR "         Defaulting to 'archives' for file output\n";
222        $archivedir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "archives");
223    }
224    }
225
226    $archivedir =~ s/[\\\/]+/\//g;
227    $archivedir =~ s/\/$//;
228    $self->{'archivedir'} = $archivedir;
229
230    if ($verbosity !~ /\d+/) {
231    if (defined $collectcfg->{'verbosity'} && $collectcfg->{'verbosity'} =~ /\d+/) {
232        $verbosity = $collectcfg->{'verbosity'};
233    } else {
234        $verbosity = 2; # the default
235    }
236    }
237    $self->{'verbosity'} = $verbosity;
238
239    if (defined $collectcfg->{'manifest'} && $self->{'manifest'} eq "") {
240    $self->{'manifest'} = $collectcfg->{'manifest'};
241    }
242
243    if (defined $collectcfg->{'gzip'} && !$self->{'gzip'}) {
244    if ($collectcfg->{'gzip'} =~ /^true$/i) {
245        $self->{'gzip'} = 1;
246    }
247    }
248
249    if ($self->{'maxdocs'} !~ /\-?\d+/) {
250    if (defined $collectcfg->{'maxdocs'} && $collectcfg->{'maxdocs'} =~ /\-?\d+/) {
251        $self->{'maxdocs'} = $collectcfg->{'maxdocs'};
252    } else {
253        $self->{'maxdocs'} = -1; # the default
254    }
255    }
256
257    if ((defined $self->{'groupsize'}) && ($self->{'groupsize'} == 1)) {
258    if (defined $collectcfg->{'groupsize'} && $collectcfg->{'groupsize'} =~ /\d+/) {
259        $self->{'groupsize'} = $collectcfg->{'groupsize'};
260    }
261    }
262
263    if (!defined $self->{'OIDtype'}
264    || ($self->{'OIDtype'} !~ /^(hash|incremental|assigned|dirname)$/ )) {
265    if (defined $collectcfg->{'OIDtype'}
266        && $collectcfg->{'OIDtype'} =~ /^(hash|incremental|assigned|dirname)$/) {
267        $self->{'OIDtype'} = $collectcfg->{'OIDtype'};
268    } else {
269        $self->{'OIDtype'} = "hash"; # the default
270    }
271    }
272
273    if ((!defined $self->{'OIDmetadata'}) || ($self->{'OIDmetadata'} eq "")) {
274    if (defined $collectcfg->{'OIDmetadata'}) {
275        $self->{'OIDmetadata'} = $collectcfg->{'OIDmetadata'};
276    } else {
277        $self->{'OIDmetadata'} = "dc.Identifier"; # the default
278    }
279    }
280
281    my $sortmeta = $self->{'sortmeta'};
282    if (defined $collectcfg->{'sortmeta'} && (!defined $sortmeta || $sortmeta eq "")) {
283    $sortmeta = $collectcfg->{'sortmeta'};
284    }
285    # sortmeta cannot be used with group size
286    $sortmeta = undef unless defined $sortmeta && $sortmeta =~ /\S/;
287    if (defined $sortmeta && $self->{'groupsize'} > 1) {
288    &gsprintf($out, "{import.cannot_sort}\n\n");
289    $sortmeta = undef;
290    }
291    $self->{'sortmeta'} = $sortmeta;
292
293    if (defined $collectcfg->{'removeprefix'} && $self->{'removeprefix'} eq "") {
294    $self->{'removeprefix'} = $collectcfg->{'removeprefix'};
295    }
296   
297    if (defined $collectcfg->{'removesuffix'} && $self->{'removesuffix'} eq "") {
298    $self->{'removesuffix'} = $collectcfg->{'removesuffix'};
299    }
300    if (defined $collectcfg->{'debug'} && $collectcfg->{'debug'} =~ /^true$/i) {
301    $self->{'debug'} = 1;
302    }
303    if (defined $collectcfg->{'gli'} && $collectcfg->{'gli'} =~ /^true$/i) {
304    $self->{'gli'} = 1;
305    }
306    $self->{'gli'} = 0 unless defined $self->{'gli'};
307       
308    # check keepold and removeold
309    my $checkdir = ($inexport_mode eq "import") ? "archives" : "export";
310
311    my ($removeold, $keepold, $incremental, $incremental_mode)
312    = &scriptutil::check_removeold_and_keepold($self->{'removeold'}, $self->{'keepold'},
313                           $self->{'incremental'}, $checkdir,
314                           $collectcfg);
315
316    $self->{'removeold'}        = $removeold;
317    $self->{'keepold'}          = $keepold;
318    $self->{'incremental'}      = $incremental;
319    $self->{'incremental_mode'} = $incremental_mode;
320}
321
322sub process_files
323{
324    my $self = shift @_;
325    my ($config_filename,$collectcfg) = @_;
326
327    my $inexport_mode = $self->{'mode'};
328
329    my $verbosity   = $self->{'verbosity'};
330    my $debug       = $self->{'debug'};
331
332    my $importdir   = $self->{'importdir'};
333    my $archivedir = $self->{'archivedir'} || $self->{'exportdir'};
334
335    my $incremental = $self->{'incremental'};
336    my $incremental_mode = $self->{'incremental_mode'};
337
338    my $removeold   = $self->{'removeold'};
339    my $keepold     = $self->{'keepold'};
340
341    my $saveas      = $self->{'saveas'};
342    my $OIDtype     = $self->{'OIDtype'};
343    my $OIDmetadata = $self->{'OIDmetadata'};
344
345    my $out         = $self->{'out'};
346    my $faillog     = $self->{'faillog'};
347
348    my $maxdocs     = $self->{'maxdocs'};
349    my $gzip        = $self->{'gzip'};
350    my $groupsize   = $self->{'groupsize'};
351    my $sortmeta    = $self->{'sortmeta'};
352
353    my $removeprefix = $self->{'removeprefix'};
354    my $removesuffix = $self->{'removesuffix'};
355
356    my $gli          = $self->{'gli'};
357
358    my $jobs         = $self->{'jobs'};
359    my $epoch        = $self->{'epoch'};
360
361    # related to export
362    my $xsltfile         = $self->{'xsltfile'};
363    my $group_marc       = $self->{'group_marc'};
364    my $mapping_file     = $self->{'mapping_file'};
365    my $xslt_mets        = $self->{'xslt_mets'};
366    my $xslt_txt         = $self->{'xslt_txt'};
367    my $fedora_namespace = $self->{'fedora_namespace'};
368
369    if ($inexport_mode eq "import") {
370    print STDERR "<Import>\n" if $gli;
371    }
372    else {
373    print STDERR "<export>\n" if $gli;
374    }
375   
376    my $manifest_lookup = new manifest($collectcfg->{'infodbtype'},$archivedir);
377    if ($self->{'manifest'} ne "") {
378    print STDERR "parsing manifest\n";
379    my $manifest_filename = $self->{'manifest'};
380
381    if (!&util::filename_is_absolute($manifest_filename)) {
382        $manifest_filename = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, $manifest_filename);
383    }
384
385    $self->{'manifest'} =~ s/[\\\/]+/\//g;
386    $self->{'manifest'} =~ s/\/$//;
387
388    $manifest_lookup->parse($manifest_filename);
389    }
390
391    my $manifest = $self->{'manifest'};
392
393    # load all the plugins
394    my $plugins = [];
395    if (defined $collectcfg->{'plugin'}) {
396    $plugins = $collectcfg->{'plugin'};
397    }
398
399    my $plugin_incr_mode = $incremental_mode;
400    if ($manifest ne "") {
401    # if we have a manifest file, then we pretend we are fully incremental for plugins
402    $plugin_incr_mode = "all";
403    }
404    #some global options for the plugins
405    my @global_opts = ();
406
407    my $pluginfo = &plugin::load_plugins ($plugins, $verbosity, $out, $faillog, \@global_opts, $plugin_incr_mode);
408    if (scalar(@$pluginfo) == 0) {
409    &gsprintf($out, "{import.no_plugins_loaded}\n");
410    die "\n";
411    }
412
413    # remove the old contents of the archives directory (and tmp directory) if needed
414    if ($removeold) {
415    if (-e $archivedir) {
416        &gsprintf($out, "{import.removing_archives}\n");
417        &util::rm_r ($archivedir);
418    }
419    my $tmpdir = &util::filename_cat ($ENV{'GSDLCOLLECTDIR'}, "tmp");
420    $tmpdir =~ s/[\\\/]+/\//g;
421    $tmpdir =~ s/\/$//;
422    if (-e $tmpdir) {
423        &gsprintf($out, "{import.removing_tmpdir}\n");
424        &util::rm_r ($tmpdir);
425    }
426    }
427
428    # create the archives dir if needed
429    &util::mk_all_dir($archivedir);
430
431    # read the archive information file
432
433    # BACKWARDS COMPATIBILITY: Just in case there are old .ldb/.bdb files (won't do anything for other infodbtypes)
434    &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-doc"));
435    &util::rename_ldb_or_bdb_file(&util::filename_cat($archivedir, "archiveinf-src"));
436
437    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-doc", $archivedir);
438    my $arcinfo_src_filename = &dbutil::get_infodb_file_path($collectcfg->{'infodbtype'}, "archiveinf-src", $archivedir);
439                           
440    my $archive_info = new arcinfo ($collectcfg->{'infodbtype'});
441    $archive_info->load_info ($arcinfo_doc_filename);
442
443    if ($manifest eq "") {
444    # Load in list of files in import folder from last import (if present)
445    $archive_info->load_prev_import_filelist ($arcinfo_src_filename);
446    }
447
448    ####Use Plugout####
449    my $plugout;
450
451    if ($inexport_mode eq "import") {
452    if (defined $collectcfg->{'plugout'}) {
453        # If a plugout was specified in the collect.cfg file, assume it is sensible
454        # We can't check the name because it could be anything, if it is a custom plugout
455        $plugout = $collectcfg->{'plugout'};
456    }
457    else{
458        if ($saveas !~ /^(GreenstoneXML|GreenstoneMETS)$/) {
459        push @$plugout,"GreenstoneXMLPlugout";
460        }
461        else{
462        push @$plugout,$saveas."Plugout";
463        }
464    }
465    }
466    else {
467    if (defined $collectcfg->{'plugout'} && $collectcfg->{'plugout'} =~ /^(.*METS|DSpace|MARCXML)Plugout/) {
468        $plugout = $collectcfg->{'plugout'};
469    }
470    else{
471        if ($saveas !~ /^(GreenstoneMETS|FedoraMETS|DSpace|MARCXML)$/) {
472        push @$plugout,"GreenstoneMETSPlugout";
473        }
474        else{
475        push @$plugout,$saveas."Plugout";
476        }
477    }
478    }
479   
480    my $plugout_name = $plugout->[0];
481
482    push @$plugout,("-output_info",$archive_info)  if (defined $archive_info);
483    push @$plugout,("-verbosity",$verbosity)       if (defined $verbosity);
484    push @$plugout,("-debug")                      if ($debug);
485    push @$plugout,("-group_size",$groupsize)      if (defined $groupsize);
486    push @$plugout,("-gzip_output")                if ($gzip);
487    push @$plugout,("-output_handle",$out)         if (defined $out);
488
489    push @$plugout,("-xslt_file",$xsltfile)        if (defined $xsltfile && $xsltfile ne "");
490
491    if ($plugout_name =~ m/^MARCXMLPlugout$/) {
492    push @$plugout,("-group")                      if ($group_marc);
493    push @$plugout,("-mapping_file",$mapping_file) if (defined $mapping_file && $mapping_file ne "");
494    }
495    if ($plugout_name =~ m/^.*METSPlugout$/) {
496    push @$plugout,("-xslt_mets",$xslt_mets)       if (defined $xslt_mets && $xslt_mets ne "");
497    push @$plugout,("-xslt_txt",$xslt_txt)         if (defined $xslt_txt && $xslt_txt ne "");
498    }
499
500    if ($plugout_name eq "FedoraMETSPlugout") {
501    push @$plugout,("-fedora_namespace",$fedora_namespace) if (defined $fedora_namespace && $fedora_namespace ne "");
502    }
503
504
505    my $processor = &plugout::load_plugout($plugout);                       
506    $processor->setoutputdir ($archivedir);
507    $processor->set_sortmeta ($sortmeta, $removeprefix, $removesuffix) if defined $sortmeta;
508    $processor->set_OIDtype ($OIDtype, $OIDmetadata);
509   
510    &plugin::begin($pluginfo, $importdir, $processor, $maxdocs, $gli);
511   
512    if ($removeold) {
513        # occasionally, plugins may want to do something on remove old, eg pharos image indexing
514    &plugin::remove_all($pluginfo, $importdir, $processor, $maxdocs, $gli);
515    }
516
517    # process the import directory
518    my $block_hash = {};
519    $block_hash->{'new_files'} = {};
520    $block_hash->{'reindex_files'} = {};
521    my $metadata = {};
522   
523    # gobal blocking pass may set up some metadata
524    &plugin::file_block_read($pluginfo, $importdir, "", $block_hash, $metadata, $gli);
525   
526    if ($manifest ne "") {
527    #
528    # 1. Process delete files first
529    #
530
531    my @deleted_files = keys %{$manifest_lookup->{'delete'}};
532    my @full_deleted_files = ();
533
534    # ensure all filenames are absolute
535    foreach my $df (@deleted_files) {
536        #print STDERR "**delete file $df\n";
537        my $full_df =
538        (&util::filename_is_absolute($df))
539        ? $df
540        : &util::filename_cat($importdir,$df);
541
542        if (-d $full_df) {
543        &add_dir_contents_to_list($full_df, \@full_deleted_files);
544        } else {
545        push(@full_deleted_files,$full_df);
546        }
547    }
548   
549    &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_deleted_files);
550    mark_docs_for_deletion($archive_info,{},
551                   \@full_deleted_files,
552                   $archivedir, $verbosity, "delete");
553
554
555    #
556    # 2. Now files for reindexing
557    #
558
559    my @reindex_files = keys %{$manifest_lookup->{'reindex'}};
560    my @full_reindex_files = ();
561
562    # ensure all filenames are absolute
563    foreach my $rf (@reindex_files) {       
564        my $full_rf =
565        (&util::filename_is_absolute($rf))
566        ? $rf
567        : &util::filename_cat($importdir,$rf);
568
569        if (-d $full_rf) {
570        &add_dir_contents_to_list($full_rf, \@full_reindex_files);
571        } else {
572        push(@full_reindex_files,$full_rf);
573        }
574    }
575   
576    &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@full_reindex_files);
577    mark_docs_for_deletion($archive_info,{},\@full_reindex_files, $archivedir,$verbosity, "reindex");
578
579    # And now to ensure the new version of the file processed by
580    # appropriate plugin, we need to add it to block_hash reindex list
581    foreach my $full_rf (@full_reindex_files) {
582        #print STDERR "***reindex file $full_rf\n";
583        $block_hash->{'reindex_files'}->{$full_rf} = 1;
584    }
585
586
587    #
588    # 3. Now finally any new files - add to block_hash new_files list
589    #
590
591    my @new_files = keys %{$manifest_lookup->{'index'}};
592    my @full_new_files = ();
593
594    foreach my $nf (@new_files) {
595        #print STDERR "***index file $nf\n";
596        # ensure filename is absolute
597        my $full_nf =
598        (&util::filename_is_absolute($nf))
599        ? $nf
600        : &util::filename_cat($importdir,$nf);
601
602        if (-d $full_nf) {
603        &add_dir_contents_to_list($full_nf, \@full_new_files);
604        } else {
605        push(@full_new_files,$full_nf);
606        }
607    }
608
609    foreach my $f (@full_new_files) {
610        $block_hash->{'new_files'}->{$f} = 1;
611    }
612    }
613    else {
614    # if incremental, we read through the import folder to see whats changed.
615
616    if ($incremental || $incremental_mode eq "onlyadd") {
617        prime_doc_oid_count($archivedir);
618
619        # Can now work out which files were new, already existed, and have
620        # been deleted
621       
622        new_vs_old_import_diff($archive_info,$block_hash,$importdir,
623                   $archivedir,$verbosity,$incremental_mode);
624       
625        my @new_files = sort keys %{$block_hash->{'new_files'}};
626        if (scalar(@new_files>0)) {
627        print STDERR "New files and modified metadata files since last import:\n  ";
628        print STDERR join("\n  ",@new_files), "\n";
629        }
630
631        if ($incremental) {
632               # only look for deletions if we are truely incremental
633        my @deleted_files = sort keys %{$block_hash->{'deleted_files'}};
634        # Filter out any in gsdl/tmp area
635        my @filtered_deleted_files = ();
636        my $gsdl_tmp_area = &util::filename_cat($ENV{'GSDLHOME'}, "tmp");
637        my $collect_tmp_area = &util::filename_cat($ENV{'GSDLCOLLECTDIR'}, "tmp");
638        $gsdl_tmp_area = &util::filename_to_regex($gsdl_tmp_area);
639        $collect_tmp_area = &util::filename_to_regex($collect_tmp_area);
640                 
641        foreach my $df (@deleted_files) {
642            next if ($df =~ m/^$gsdl_tmp_area/);
643            next if ($df =~ m/^$collect_tmp_area/);
644           
645            push(@filtered_deleted_files,$df);
646        }       
647       
648
649        @deleted_files = @filtered_deleted_files;
650       
651        if (scalar(@deleted_files)>0) {
652            print STDERR "Files deleted since last import:\n  ";
653            print STDERR join("\n  ",@deleted_files), "\n";
654       
655       
656            &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@deleted_files);
657           
658            mark_docs_for_deletion($archive_info,$block_hash,\@deleted_files, $archivedir,$verbosity, "delete");
659        }
660       
661        my @reindex_files = sort keys %{$block_hash->{'reindex_files'}};
662       
663        if (scalar(@reindex_files)>0) {
664            print STDERR "Files to reindex since last import:\n  ";
665            print STDERR join("\n  ",@reindex_files), "\n";
666            &plugin::remove_some($pluginfo, $collectcfg->{'infodbtype'}, $archivedir, \@reindex_files);
667            mark_docs_for_deletion($archive_info,$block_hash,\@reindex_files, $archivedir,$verbosity, "reindex");
668        }
669               
670        }       
671    }
672    }
673   
674    # now, whichever mode we are in, we can process the entire import folder
675    if ((defined $jobs) && ($jobs > 1))
676    {
677    # if jobs are set to >1, run in parallel using MPI helper
678    # [hs, 1 july 2010]
679    &ParallelInexport::farm_out_processes($jobs, $epoch, $importdir, $block_hash,
680                          $self->{'collection'}, $self->{'site'});
681    }
682    else
683    {
684    &plugin::read ($pluginfo, $importdir, "", $block_hash, $metadata, $processor, $maxdocs, 0, $gli);
685    }
686   
687   
688    if ($saveas eq "FedoraMETS") {
689    # create collection "doc obj" for Fedora that contains
690    # collection-level metadata
691   
692    my $doc_obj = new doc($config_filename,"nonindexed_doc","none");
693    $doc_obj->set_OID("collection");
694   
695    my $col_name = undef;
696    my $col_meta = $collectcfg->{'collectionmeta'};
697   
698    if (defined $col_meta) {       
699        store_collectionmeta($col_meta,"collectionname",$doc_obj); # in GS3 this is a collection's name
700        store_collectionmeta($col_meta,"collectionextra",$doc_obj); # in GS3 this is a collection's description     
701    }
702    $processor->process($doc_obj);
703    }
704
705    &plugin::end($pluginfo, $processor);
706
707    &plugin::deinit($pluginfo, $processor);
708
709    # Store the value of OIDCount (used in doc.pm) so it can be
710    # restored correctly to this value on an incremental build
711    store_doc_oid_count($archivedir);
712
713    # write out the archive information file
714    $processor->close_file_output() if (defined $groupsize) && ($groupsize > 1);
715    $processor->close_group_output() if $processor->is_group();
716
717    # for backwards compatability with archvies.inf file
718    if ($arcinfo_doc_filename =~ m/(contents)|(\.inf)$/) {
719    $archive_info->save_info($arcinfo_doc_filename);
720    }
721    else {
722    $archive_info->save_revinfo_db($arcinfo_src_filename);
723    }
724
725    return $pluginfo;
726}
727
728
729sub generate_statistics
730{
731    my $self = shift @_;
732    my ($pluginfo) = @_;
733
734    my $inexport_mode = $self->{'mode'};
735
736    my $statsfile   = $self->{'statsfile'};
737    my $out         = $self->{'out'};
738    my $faillogname = $self->{'faillogname'};
739    my $gli         = $self->{'gli'};
740    my $jobs        = $self->{'jobs'};
741
742    # write out import stats
743
744    if ((!defined $jobs) || ($jobs == 1))
745    {
746    # only output statistics if there are multiple jobs
747    # [hs, 1 july 2010]
748
749    my $close_stats = 0;
750    if ($statsfile !~ /^(STDERR|STDOUT)$/i) {
751        if (open (STATS, ">$statsfile")) {
752        $statsfile = 'inexport::STATS';
753        $close_stats = 1;
754        } else {
755        &gsprintf($out, "{import.cannot_open_stats_file}", $statsfile);
756        &gsprintf($out, "{import.stats_backup}\n");
757        $statsfile = 'STDERR';
758        }
759    }
760   
761    &gsprintf($out, "\n");
762    &gsprintf($out, "*********************************************\n");
763    &gsprintf($out, "{$inexport_mode.complete}\n");
764    &gsprintf($out, "*********************************************\n");
765   
766    &plugin::write_stats($pluginfo, $statsfile, $faillogname, $gli);
767    if ($close_stats) {
768        close STATS;
769    }
770    }
771
772    close OUT if $self->{'close_out'};
773    close FAILLOG;
774}
775
776
777sub store_collectionmeta
778{
779    my ($collectionmeta,$field,$doc_obj) = @_;
780   
781    my $section = $doc_obj->get_top_section();
782   
783    my $field_hash = $collectionmeta->{$field};
784   
785    foreach my $k (keys %$field_hash)
786    {
787    my $val = $field_hash->{$k};
788   
789    ### print STDERR "*** $k = $field_hash->{$k}\n";
790   
791    my $md_label = "ex.$field";
792   
793   
794    if ($k =~ m/^\[l=(.*?)\]$/)
795    {
796       
797        my $md_suffix = $1;
798        $md_label .= "^$md_suffix";
799    }
800   
801   
802    $doc_obj->add_utf8_metadata($section,$md_label, $val);
803   
804    # see collConfigxml.pm: GS2's "collectionextra" is called "description" in GS3,
805    # while "collectionname" in GS2 is called "name" in GS3.
806    # Variable $nameMap variable in collConfigxml.pm maps between GS2 and GS3
807    if (($md_label eq "ex.collectionname^en") || ($md_label eq "ex.collectionname"))
808    {
809        $doc_obj->add_utf8_metadata($section,"dc.Title", $val);
810    }
811   
812    }
813}
814
815
816sub oid_count_file {
817    my ($archivedir) = @_;
818    return &util::filename_cat ($archivedir, "OIDcount");
819}
820
821
822sub prime_doc_oid_count
823{
824    my ($archivedir) = @_;
825    my $oid_count_filename = &oid_count_file($archivedir);
826
827    if (-e $oid_count_filename) {
828    if (open(OIDIN,"<$oid_count_filename")) {
829        my $OIDcount = <OIDIN>;
830        chomp $OIDcount;       
831        close(OIDIN);
832
833        $doc::OIDcount = $OIDcount;     
834    }
835    else {
836       
837        print STDERR "Warning: unable to read document OID count from $oid_count_filename\n";
838        print STDERR "Setting value to 0\n";
839    }
840    }
841   
842}
843
844sub store_doc_oid_count
845{
846    # Use the file "OIDcount" in the archives directory to record
847    # what value doc.pm got up to
848
849    my ($archivedir) = @_;
850    my $oid_count_filename = &oid_count_file($archivedir);
851
852
853    if (open(OIDOUT,">$oid_count_filename")) {
854    print OIDOUT $doc::OIDcount, "\n";
855       
856    close(OIDOUT);
857    }
858    else {
859    print STDERR "Warning: unable to store document OID count\n";
860    }
861}
862
863
864
865sub new_vs_old_import_diff
866{
867    my ($archive_info,$block_hash,$importdir,$archivedir,$verbosity,$incremental_mode) = @_;
868
869    # Get the infodbtype value for this collection from the arcinfo object
870    my $infodbtype = $archive_info->{'infodbtype'};
871
872    # in this method, we want to know if metadata files are modified or not.
873    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
874
875    my $archiveinf_timestamp = -M $arcinfo_doc_filename;
876
877    # First convert all files to absolute form
878    # This is to support the situation where the import folder is not
879    # the default
880   
881    my $prev_all_files = $archive_info->{'prev_import_filelist'};
882    my $full_prev_all_files = {};
883
884    foreach my $prev_file (keys %$prev_all_files) {
885
886    if (!&util::filename_is_absolute($prev_file)) {
887        my $full_prev_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$prev_file);
888        $full_prev_all_files->{$full_prev_file} = $prev_file;
889    }
890    else {
891        $full_prev_all_files->{$prev_file} = $prev_file;
892    }
893    }
894
895
896    # Figure out which are the new files, existing files and so
897    # by implication the files from the previous import that are not
898    # there any more => mark them for deletion
899    foreach my $curr_file (keys %{$block_hash->{'all_files'}}) {
900   
901    my $full_curr_file = $curr_file;
902
903    # entry in 'all_files' is moved to either 'existing_files',
904    # 'deleted_files', 'new_files', or 'new_or_modified_metadata_files'
905
906    if (!&util::filename_is_absolute($curr_file)) {
907        # add in import dir to make absolute
908        $full_curr_file = &util::filename_cat($importdir,$curr_file);
909    }
910
911    # figure out if new file or not
912    if (defined $full_prev_all_files->{$full_curr_file}) {
913        # delete it so that only files that need deleting are left
914        delete $full_prev_all_files->{$full_curr_file};
915       
916        # had it before. is it a metadata file?
917        if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
918       
919        # is it modified??
920        if (-M $full_curr_file < $archiveinf_timestamp) {
921            print STDERR "*** Detected a modified metadata file: $full_curr_file\n" if $verbosity > 2;
922            # its newer than last build
923            $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
924        }
925        }
926        else {
927        if ($incremental_mode eq "all") {
928           
929            # had it before
930            $block_hash->{'existing_files'}->{$full_curr_file} = 1;
931           
932        }
933        else {
934            # Warning in "onlyadd" mode, but had it before!
935            print STDERR "Warning: File $full_curr_file previously imported.\n";
936            print STDERR "         Treating as new file\n";
937           
938            $block_hash->{'new_files'}->{$full_curr_file} = 1;
939           
940        }
941        }
942    }
943    else {
944        if ($block_hash->{'metadata_files'}->{$full_curr_file}) {
945        # the new file is the special sort of file greenstone uses
946        # to attach metadata to src documents
947        # i.e metadata.xml
948        # (but note, the filename used is not constrained in
949        # Greenstone to always be this)
950
951        print STDERR "***** Detected new metadata file: $full_curr_file\n" if $verbosity > 2;
952        $block_hash->{'new_or_modified_metadata_files'}->{$full_curr_file} = 1;
953        }
954        else {
955        $block_hash->{'new_files'}->{$full_curr_file} = 1;
956        }
957    }
958
959   
960    delete $block_hash->{'all_files'}->{$curr_file};
961    }
962
963
964
965
966    # Deal with complication of new or modified metadata files by forcing
967    # everything from this point down in the file hierarchy to
968    # be freshly imported. 
969    #
970    # This may mean files that have not changed are reindexed, but does
971    # guarantee by the end of processing all new metadata is correctly
972    # associated with the relevant document(s).
973
974    foreach my $new_mdf (keys %{$block_hash->{'new_or_modified_metadata_files'}}) {
975    my ($fileroot,$situated_dir,$ext) = fileparse($new_mdf, "\\.[^\\.]+\$");
976
977    $situated_dir =~ s/[\\\/]+$//; # remove tailing slashes
978    $situated_dir =~ s/\\/\\\\/g;  # need to protect windows slash \ in regular expression
979   
980    # Go through existing_files, and mark anything that is contained
981    # within 'situated_dir' to be reindexed (in case some of the metadata
982    # attaches to one of these files)
983
984    my $reindex_files = [];
985
986    foreach my $existing_f (keys %{$block_hash->{'existing_files'}}) {
987   
988        if ($existing_f =~ m/^$situated_dir/) {
989        push(@$reindex_files,$existing_f);
990        $block_hash->{'reindex_files'}->{$existing_f} = 1;
991        delete $block_hash->{'existing_files'}->{$existing_f};
992
993        }
994    }
995   
996    # metadata file needs to be in new_files list so parsed by MetadataXMLPlug
997    # (or equivalent)
998    $block_hash->{'new_files'}->{$new_mdf} = 1;
999
1000    }
1001
1002    # go through remaining existing files and work out what has changed and needs to be reindexed.
1003    my @existing_files = sort keys %{$block_hash->{'existing_files'}};
1004
1005    my $reindex_files = [];
1006
1007    foreach my $existing_filename (@existing_files) {
1008    if (-M $existing_filename < $archiveinf_timestamp) {
1009        # file is newer than last build
1010       
1011        my $existing_file = $existing_filename;
1012        #my $collectdir = &util::filename_cat($ENV{'GSDLCOLLECTDIR'});
1013
1014        #my $collectdir_resafe = &util::filename_to_regex($collectdir);
1015        #$existing_file =~ s/^$collectdir_resafe(\\|\/)?//;
1016       
1017        print STDERR "**** Reindexing existing file: $existing_file\n";
1018
1019        push(@$reindex_files,$existing_file);
1020        $block_hash->{'reindex_files'}->{$existing_filename} = 1;
1021    }
1022
1023    }
1024
1025   
1026    # By this point full_prev_all_files contains the files
1027    # mentioned in archiveinf-src.db but are not in the 'import'
1028    # folder (or whatever was specified through -importdir ...)
1029
1030    # This list can contain files that were created in the 'tmp' or
1031    # 'cache' areas (such as screen-size and thumbnail images).
1032    #
1033    # In building the final list of files to delete, we test to see if
1034    # it exists on the filesystem and if it does (unusual for a "normal"
1035    # file in import, but possible in the case of 'tmp' files),
1036    # supress it from going into the final list
1037
1038    my $collectdir = $ENV{'GSDLCOLLECTDIR'};
1039
1040    my @deleted_files = values %$full_prev_all_files;
1041    map { my $curr_file = $_;
1042      my $full_curr_file = $curr_file;
1043
1044      if (!&util::filename_is_absolute($curr_file)) {
1045          # add in import dir to make absolute
1046
1047          $full_curr_file = &util::filename_cat($collectdir,$curr_file);
1048      }
1049
1050
1051      if (!-e $full_curr_file) {
1052          $block_hash->{'deleted_files'}->{$curr_file} = 1;
1053      }
1054      } @deleted_files;
1055
1056
1057
1058}
1059
1060
1061# this is used to delete "deleted" docs, and to remove old versions of "changed" docs
1062# $mode is 'delete' or 'reindex'
1063sub mark_docs_for_deletion
1064{
1065    my ($archive_info,$block_hash,$deleted_files,$archivedir,$verbosity,$mode) = @_;
1066
1067    my $mode_text = "deleted from index";
1068    if ($mode eq "reindex") {
1069    $mode_text = "reindexed";
1070    }
1071
1072    # Get the infodbtype value for this collection from the arcinfo object
1073    my $infodbtype = $archive_info->{'infodbtype'};
1074
1075    my $arcinfo_doc_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $archivedir);
1076    my $arcinfo_src_filename = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $archivedir);
1077
1078
1079    # record files marked for deletion in arcinfo
1080    foreach my $file (@$deleted_files) {
1081    # use 'archiveinf-src' info database file to look up all the OIDs
1082    # that this file is used in (note in most cases, it's just one OID)
1083   
1084    my $src_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_src_filename, $file);
1085    my $src_rec = &dbutil::convert_infodb_string_to_hash($src_rec_string);
1086    my $oids = $src_rec->{'oid'};
1087    my $file_record_deleted = 0;
1088
1089    # delete the src record
1090    my $src_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_src_filename, "append");
1091    &dbutil::delete_infodb_entry($infodbtype, $src_infodb_file_handle, $file);
1092    &dbutil::close_infodb_write_handle($infodbtype, $src_infodb_file_handle);
1093
1094
1095    foreach my $oid (@$oids) {
1096
1097        # find the source doc (the primary file that becomes this oid)
1098        my $doc_rec_string = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1099        my $doc_rec = &dbutil::convert_infodb_string_to_hash($doc_rec_string);
1100        my $doc_source_file = $doc_rec->{'src-file'}->[0];
1101        if (!&util::filename_is_absolute($doc_source_file)) {
1102        $doc_source_file = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},$doc_source_file);
1103        }
1104
1105        if ($doc_source_file ne $file) {
1106        # its an associated or metadata file
1107       
1108        # mark source doc for reimport as one of its assoc files has changed or deleted
1109        $block_hash->{'reindex_files'}->{$doc_source_file} = 1;
1110       
1111        }
1112        my $curr_status = $archive_info->get_status_info($oid);
1113        if (defined($curr_status) && (($curr_status ne "D"))) {
1114        if ($verbosity>1) {
1115            print STDERR "$oid ($doc_source_file) marked to be $mode_text on next buildcol.pl\n";
1116        }
1117        # mark oid for deletion (it will be deleted or reimported)
1118        $archive_info->set_status_info($oid,"D");
1119        my $val = &dbutil::read_infodb_entry($infodbtype, $arcinfo_doc_filename, $oid);
1120        $val =~ s/^<index-status>(.*)$/<index-status>D/m;
1121
1122        my $val_rec = &dbutil::convert_infodb_string_to_hash($val);
1123        my $doc_infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $arcinfo_doc_filename, "append");
1124
1125        &dbutil::write_infodb_entry($infodbtype, $doc_infodb_file_handle, $oid, $val_rec);
1126        &dbutil::close_infodb_write_handle($infodbtype, $doc_infodb_file_handle);
1127        }
1128    }
1129   
1130    }
1131    # now go through and check that we haven't marked any primary files for reindex (because their associated files have changed/deleted) when they have been deleted themselves. only in delete mode.
1132    if ($mode eq "delete") {
1133    foreach my $file (@$deleted_files) {
1134        if (defined $block_hash->{'reindex_files'}->{$file}) {
1135        delete $block_hash->{'reindex_files'}->{$file};
1136        }
1137    }
1138    }
1139
1140
1141}
1142
1143sub add_dir_contents_to_list {
1144
1145    my ($dirname, $list) = @_;
1146 
1147    # Recur over directory contents.
1148    my (@dir, $subfile);
1149   
1150    # find all the files in the directory
1151    if (!opendir (DIR, $dirname)) {
1152    print STDERR "inexport: WARNING - couldn't read directory $dirname\n";
1153    return -1; # error in processing
1154    }
1155    @dir = readdir (DIR);
1156    closedir (DIR);
1157   
1158    for (my $i = 0; $i < scalar(@dir); $i++) {
1159    my $subfile = $dir[$i];
1160    next if ($subfile =~ m/^\.\.?$/);
1161    my $full_file = &util::filename_cat($dirname, $subfile);
1162    if (-d $full_file) {
1163        &add_dir_contents_to_list($full_file, $list);
1164    } else {
1165        push (@$list, $full_file);
1166    }
1167    }
1168   
1169}
1170
1171
11721;
Note: See TracBrowser for help on using the browser.