source: main/trunk/greenstone2/perllib/plugouts/BasePlugout.pm@ 33477

Last change on this file since 33477 was 33477, checked in by kjdon, 5 years ago

need to call setup_custom_sort to allow for collection's customsorttools.pm
  • Property svn:keywords set to Author Date Id Revision
File size: 38.3 KB
Line 
1###########################################################################
2#
3# BasePlugout.pm -- base class for all the plugout modules
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) 2006 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 BasePlugout;
27
28eval {require bytes};
29
30use strict;
31no strict 'subs';
32no strict 'refs';
33
34use dbutil;
35use gsprintf 'gsprintf';
36use printusage;
37use parse2;
38use util;
39use FileUtils;
40use sorttools;
41
42# suppress the annoying "subroutine redefined" warning that various
43# gets cause under perl 5.6
44$SIG{__WARN__} = sub {warn($_[0]) unless ($_[0] =~ /Subroutine\s+\S+\sredefined/)};
45
46my $arguments = [
47 { 'name' => "site",
48 'desc' => "{BasPlugout.site}",
49 'type' => "string",
50 'reqd' => "no",
51 'hiddengli' => "yes" },
52 { 'name' => "xslt_file",
53 'desc' => "{BasPlugout.xslt_file}",
54 'type' => "string",
55 'reqd' => "no",
56 'deft' => "",
57 'hiddengli' => "no"},
58 { 'name' => "subdir_split_length",
59 'desc' => "{BasPlugout.subdir_split_length}",
60 'type' => "int",
61 'reqd' => "no",
62 'deft' => "8",
63 'hiddengli' => "no"},
64 { 'name' => "subdir_hash_prefix",
65 'desc' => "{BasPlugout.subdir_hash_prefix}",
66 'type' => "flag",
67 'reqd' => "no",
68 'deft' => "0",
69 'hiddengli' => "no"},
70 { 'name' => "gzip_output",
71 'desc' => "{BasPlugout.gzip_output}",
72 'type' => "flag",
73 'reqd' => "no",
74 'hiddengli' => "no"},
75 { 'name' => "verbosity",
76 'desc' => "{BasPlugout.verbosity}",
77 'type' => "int",
78 'deft' => "0",
79 'reqd' => "no",
80 'hiddengli' => "no"},
81 { 'name' => "output_info",
82 'desc' => "{BasPlugout.output_info}",
83 'type' => "string",
84 'reqd' => "yes",
85 'hiddengli' => "yes"},
86 { 'name' => "output_handle",
87 'desc' => "{BasPlugout.output_handle}",
88 'type' => "string",
89 'deft' => 'STDERR',
90 'reqd' => "no",
91 'hiddengli' => "yes"},
92 { 'name' => "debug",
93 'desc' => "{BasPlugout.debug}",
94 'type' => "flag",
95 'reqd' => "no",
96 'hiddengli' => "yes"},
97 { 'name' => 'no_rss',
98 'desc' => "{BasPlugout.no_rss}",
99 'type' => 'flag',
100 'reqd' => 'no',
101 'hiddengli' => 'yes'},
102 { 'name' => 'rss_title',
103 'desc' => "{BasPlugout.rss_title}",
104 'type' => 'string',
105 'deft' => 'dc.Title',
106 'reqd' => 'no',
107 'hiddengli' => 'yes'},
108 { 'name' => "no_auxiliary_databases",
109 'desc' => "{BasPlugout.no_auxiliary_databases}",
110 'type' => "flag",
111 'reqd' => "no",
112 'hiddengli' => "yes"}
113
114];
115
116my $options = { 'name' => "BasePlugout",
117 'desc' => "{BasPlugout.desc}",
118 'abstract' => "yes",
119 'inherits' => "no",
120 'args' => $arguments};
121
122sub new
123{
124 my $class = shift (@_);
125
126 my ($plugoutlist,$args,$hashArgOptLists) = @_;
127 push(@$plugoutlist, $class);
128
129 my $plugout_name = (defined $plugoutlist->[0]) ? $plugoutlist->[0] : $class;
130
131 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
132 push(@{$hashArgOptLists->{"OptList"}},$options);
133
134 my $self = {};
135 $self->{'plugout_type'} = $class;
136 $self->{'option_list'} = $hashArgOptLists->{"OptList"};
137 $self->{"info_only"} = 0;
138
139 # Check if gsdlinfo is in the argument list or not - if it is, don't parse
140 # the args, just return the object.
141 #print STDERR "#### " . join(",", @${args}) . "\n\n";
142 my $v=0;
143 foreach my $strArg (@{$args})
144 {
145 if(defined $strArg) {
146 if($strArg eq "-gsdlinfo")
147 {
148 $self->{"info_only"} = 1;
149 return bless $self, $class;
150 }
151 elsif ($strArg eq "-site") {
152 $v = $strArg;
153 }
154 elsif($v eq "-site") {
155 $self->{'site'} = $strArg;
156 }
157 }
158 }
159
160 delete $self->{"info_only"};
161
162 if(parse2::parse($args,$hashArgOptLists->{"ArgList"},$self) == -1)
163 {
164 my $classTempClass = bless $self, $class;
165 print STDERR "<BadPlugout d=$plugout_name>\n";
166 &gsprintf(STDERR, "\n{BasPlugout.bad_general_option}\n", $plugout_name);
167 $classTempClass->print_txt_usage(""); # Use default resource bundle
168 die "\n";
169 }
170
171
172 if(defined $self->{'xslt_file'} && $self->{'xslt_file'} ne "")
173 {
174 my $full_file_path = &util::locate_config_file($self->{'xslt_file'});
175 if (!defined $full_file_path) {
176 print STDERR "Can not find $self->{'xslt_file'}, please make sure you have supplied the correct file path or put the file into the collection's etc or greenstone's etc folder\n";
177 die "\n";
178 }
179 $self->{'xslt_file'} = $full_file_path;
180 }
181
182 # for group processing
183 $self->{'gs_count'} = 0;
184 $self->{'group_position'} = 1;
185
186 $self->{'keep_import_structure'} = 0;
187
188 $self->{'generate_databases'} = 1;
189 if ($self->{'no_auxiliary_databases'}) {
190 $self->{'generate_databases'} = 0;
191 }
192 undef $self->{'no_auxiliary_databases'};
193
194 &sorttools::setup_custom_sort();
195 return bless $self, $class;
196
197}
198
199# implement this in subclass if you want to do some initialization after
200# loading and setting parameters, and before processing the documents.
201sub begin {
202
203 my $self= shift (@_);
204
205}
206# implement in subclasses if it needs some non-group related cleanup (post-group cleanup
207# Like begin(), end() is also called by inexport.pm
208sub end {
209 my $self= shift (@_);
210
211}
212sub print_xml_usage
213{
214 my $self = shift(@_);
215 my $header = shift(@_);
216 my $high_level_information_only = shift(@_);
217
218 # XML output is always in UTF-8
219 gsprintf::output_strings_in_UTF8;
220
221 if ($header) {
222 &PrintUsage::print_xml_header("plugout");
223 }
224 $self->print_xml($high_level_information_only);
225}
226
227
228sub print_xml
229{
230 my $self = shift(@_);
231 my $high_level_information_only = shift(@_);
232
233 my $optionlistref = $self->{'option_list'};
234 my @optionlist = @$optionlistref;
235 my $plugoutoptions = shift(@$optionlistref);
236 return if (!defined($plugoutoptions));
237
238 gsprintf(STDERR, "<PlugoutInfo>\n");
239 gsprintf(STDERR, " <Name>$plugoutoptions->{'name'}</Name>\n");
240 my $desc = gsprintf::lookup_string($plugoutoptions->{'desc'});
241 $desc =~ s/</&amp;lt;/g; # doubly escaped
242 $desc =~ s/>/&amp;gt;/g;
243 gsprintf(STDERR, " <Desc>$desc</Desc>\n");
244 gsprintf(STDERR, " <Abstract>$plugoutoptions->{'abstract'}</Abstract>\n");
245 gsprintf(STDERR, " <Inherits>$plugoutoptions->{'inherits'}</Inherits>\n");
246 unless (defined($high_level_information_only)) {
247 gsprintf(STDERR, " <Arguments>\n");
248 if (defined($plugoutoptions->{'args'})) {
249 &PrintUsage::print_options_xml($plugoutoptions->{'args'});
250 }
251 gsprintf(STDERR, " </Arguments>\n");
252
253 # Recurse up the plugout hierarchy
254 $self->print_xml();
255 }
256 gsprintf(STDERR, "</PlugoutInfo>\n");
257}
258
259
260sub print_txt_usage
261{
262 my $self = shift(@_);
263
264 # Print the usage message for a plugout (recursively)
265 my $descoffset = $self->determine_description_offset(0);
266 $self->print_plugout_usage($descoffset, 1);
267}
268
269sub determine_description_offset
270{
271 my $self = shift(@_);
272 my $maxoffset = shift(@_);
273
274 my $optionlistref = $self->{'option_list'};
275 my @optionlist = @$optionlistref;
276 my $plugoutoptions = pop(@$optionlistref);
277 return $maxoffset if (!defined($plugoutoptions));
278
279 # Find the length of the longest option string of this download
280 my $plugoutargs = $plugoutoptions->{'args'};
281 if (defined($plugoutargs)) {
282 my $longest = &PrintUsage::find_longest_option_string($plugoutargs);
283 if ($longest > $maxoffset) {
284 $maxoffset = $longest;
285 }
286 }
287
288 # Recurse up the download hierarchy
289 $maxoffset = $self->determine_description_offset($maxoffset);
290 $self->{'option_list'} = \@optionlist;
291 return $maxoffset;
292}
293
294
295sub print_plugout_usage
296{
297 my $self = shift(@_);
298 my $descoffset = shift(@_);
299 my $isleafclass = shift(@_);
300
301 my $optionlistref = $self->{'option_list'};
302 my @optionlist = @$optionlistref;
303 my $plugoutoptions = shift(@$optionlistref);
304 return if (!defined($plugoutoptions));
305
306 my $plugoutname = $plugoutoptions->{'name'};
307 my $plugoutargs = $plugoutoptions->{'args'};
308 my $plugoutdesc = $plugoutoptions->{'desc'};
309
310 # Produce the usage information using the data structure above
311 if ($isleafclass) {
312 if (defined($plugoutdesc)) {
313 gsprintf(STDERR, "$plugoutdesc\n\n");
314 }
315 gsprintf(STDERR, " {common.usage}: plugout $plugoutname [{common.options}]\n\n");
316 }
317
318 # Display the download options, if there are some
319 if (defined($plugoutargs)) {
320 # Calculate the column offset of the option descriptions
321 my $optiondescoffset = $descoffset + 2; # 2 spaces between options & descriptions
322
323 if ($isleafclass) {
324 gsprintf(STDERR, " {common.specific_options}:\n");
325 }
326 else {
327 gsprintf(STDERR, " {common.general_options}:\n", $plugoutname);
328 }
329
330 # Display the download options
331 &PrintUsage::print_options_txt($plugoutargs, $optiondescoffset);
332 }
333
334 # Recurse up the download hierarchy
335 $self->print_plugout_usage($descoffset, 0);
336 $self->{'option_list'} = \@optionlist;
337}
338
339
340sub error
341{
342 my ($strFunctionName,$strError) = @_;
343 {
344 print "Error occoured in BasePlugout.pm\n".
345 "In Function: ".$strFunctionName."\n".
346 "Error Message: ".$strError."\n";
347 exit(-1);
348 }
349}
350
351
352# OIDtype may be "hash" or "hash_on_full_filename" or "incremental" or "filename" or "dirname" or "full_filename" or "assigned"
353sub set_OIDtype {
354 my $self = shift (@_);
355 my ($type, $metadata) = @_;
356
357 if ($type =~ /^(hash|hash_on_full_filename|incremental|filename|dirname|full_filename|assigned)$/) {
358 $self->{'OIDtype'} = $type;
359 } else {
360 $self->{'OIDtype'} = "hash";
361 }
362 if ($type =~ /^assigned$/) {
363 if (defined $metadata) {
364 $self->{'OIDmetadata'} = $metadata;
365 } else {
366 $self->{'OIDmetadata'} = "dc.Identifier";
367 }
368 }
369}
370
371sub set_output_dir
372{
373 my $self = shift @_;
374 my ($output_dir) = @_;
375
376 $self->{'output_dir'} = $output_dir;
377}
378
379sub setoutputdir
380{
381 my $self = shift @_;
382 my ($output_dir) = @_;
383
384 $self->{'output_dir'} = $output_dir;
385}
386
387sub get_output_dir
388{
389 my $self = shift (@_);
390
391 return $self->{'output_dir'};
392}
393
394sub getoutputdir
395{
396 my $self = shift (@_);
397
398 return $self->{'output_dir'};
399}
400
401sub getoutputinfo
402{
403 my $self = shift (@_);
404
405 return $self->{'output_info'};
406}
407
408
409sub get_output_handler
410{
411 my $self = shift (@_);
412
413 my ($output_file_name) = @_;
414
415 my $fh;
416 &FileUtils::openFileHandle($output_file_name, '>', \$fh) or die('Can not open a file handler for: ' . $output_file_name . "\n");
417
418 return $fh;
419}
420
421sub release_output_handler
422{
423 my $self = shift (@_);
424 my ($outhandler) = @_;
425
426 close($outhandler);
427
428}
429
430sub output_xml_header {
431 my $self = shift (@_);
432 my ($handle,$docroot,$nondoctype) = @_;
433
434
435 #print $handle '<?xml version="1.0" encoding="UTF-8" standalone="no"?>' . "\n";
436
437 #For Dspace must be UTF in lower case
438 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
439
440 if (!defined $nondoctype){
441 my $doctype = (defined $docroot) ? $docroot : "Section";
442
443 # Used to be '<!DOCTYPE Archive SYSTEM ...'
444
445 print $handle "<!DOCTYPE $doctype SYSTEM \"http://greenstone.org/dtd/Archive/1.0/Archive.dtd\">\n";
446 }
447
448 print $handle "<$docroot>\n" if defined $docroot;
449}
450
451sub output_xml_footer {
452 my $self = shift (@_);
453 my ($handle,$docroot) = @_;
454 print $handle "</$docroot>\n" if defined $docroot;
455}
456
457
458sub output_general_xml_header
459{
460 my $self = shift (@_);
461 my ($handle,$docroot,$opt_attributes,$opt_dtd, $opt_doctype) = @_;
462
463 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
464
465 if (defined $opt_dtd) {
466 my $doctype = (defined $opt_doctype) ? $opt_doctype : $docroot;
467 print $handle "<!DOCTYPE $doctype SYSTEM \"$opt_dtd\">\n";
468 }
469
470 if (defined $docroot) {
471 my $full_docroot = $docroot;
472 if (defined $opt_attributes) {
473 $full_docroot .= " $opt_attributes";
474 }
475
476 print $handle "<$full_docroot>\n"
477 }
478}
479
480sub output_general_xml_footer
481{
482 output_xml_footer(@_);
483}
484
485# This is called by the plugins after read_into_doc_obj generates the doc_obj.
486sub process {
487 my $self = shift (@_);
488 my ($doc_obj) = @_;
489
490 my $output_info = $self->{'output_info'};
491 return if (!defined $output_info);
492
493 # for OAI purposes
494 $doc_obj->set_lastmodified();
495 $doc_obj->set_oailastmodified();
496
497 # find out which directory to save to
498 my $doc_dir = "";
499 if ($self->is_group()) {
500 $doc_dir = $self->get_group_doc_dir($doc_obj);
501 } else {
502 $doc_dir = $self->get_doc_dir($doc_obj);
503 }
504
505 ##############################
506 # call subclass' saveas method
507 ##############################
508 $self->saveas($doc_obj,$doc_dir);
509
510 # write out data to archiveinf-doc.db
511 if ($self->{'generate_databases'}) {
512 $self->archiveinf_db($doc_obj);
513 }
514 if ($self->is_group()) {
515 $self->{'gs_count'}++; # do we want this for all cases?
516 $self->{'group_position'}++;
517 }
518}
519
520sub store_output_info_reference {
521 my $self = shift (@_);
522 my ($doc_obj) = @_;
523
524 my $output_info = $self->{'output_info'};
525 my $metaname = $self->{'sortmeta'};
526
527 my $group_position;
528 if ($self->is_group()) {
529 $group_position = $self->{'group_position'};
530 }
531 if (!defined $metaname || $metaname !~ /\S/) {
532 my $OID = $doc_obj->get_OID();
533 $output_info->add_info($OID,$self->{'short_doc_file'}, undef, "", $group_position);
534 return;
535 }
536
537 if ($metaname eq "OID") { # sort by OID
538 my $OID = $doc_obj->get_OID();
539 $output_info->add_info($OID,$self->{'short_doc_file'}, undef, $OID, undef);
540 return;
541 }
542
543 my $metadata = "";
544 my $top_section = $doc_obj->get_top_section();
545
546 my @commameta_list = split(/,/, $metaname);
547 foreach my $cmn (@commameta_list) {
548 my $meta = $doc_obj->get_metadata_element($top_section, $cmn);
549 if ($meta) {
550 # do remove prefix/suffix - this will apply to all values
551 $meta =~ s/^$self->{'removeprefix'}// if defined $self->{'removeprefix'};
552 $meta =~ s/$self->{'removesuffix'}$// if defined $self->{'removesuffix'};
553 $meta = &sorttools::format_metadata_for_sorting($cmn, $meta, $doc_obj);
554 $metadata .= $meta if ($meta);
555 }
556 }
557
558 # store reference in the output_info
559 $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, $metadata,undef);
560
561}
562
563
564
565sub saveas {
566 my $self = shift (@_);
567 my ($doc_obj, $doc_dir) = @_;
568
569 die "Basplug::saveas function must be implemented in sub classes\n";
570}
571
572sub get_group_doc_dir {
573 my $self = shift (@_);
574 my ($doc_obj) = @_;
575
576 my $outhandle = $self->{'output_handle'};
577 my $OID = $doc_obj->get_OID();
578 $OID = "NULL" unless defined $OID;
579
580 my $groupsize = $self->{'group_size'};
581 my $gs_count = $self->{'gs_count'};
582 my $open_new_file = (($gs_count % $groupsize)==0);
583
584 my $doc_dir;
585
586 if (!$open_new_file && scalar(@{$doc_obj->get_assoc_files()})>0) {
587 # if we have some assoc files, then we will need to start a new file
588 if ($self->{'verbosity'} > 2) {
589 print $outhandle " Starting a archives folder for $OID as it has associated files\n";
590 }
591 $open_new_file = 1;
592 }
593
594 # opening a new file
595 if (($open_new_file) || !defined($self->{'gs_doc_dir'})) {
596 # first we close off the old output
597 if ($gs_count>0)
598 {
599 return if (!$self->close_group_output());
600 }
601
602 # this will create the directory
603 $doc_dir = $self->get_doc_dir ($doc_obj);
604 $self->{'new_doc_dir'} = 1;
605 $self->{'gs_doc_dir'} = $doc_dir;
606 $self->{'group_position'} = 1;
607 }
608 else {
609 $doc_dir = $self->{'gs_doc_dir'};
610 $self->{'new_doc_dir'} = 0;
611 }
612 return $doc_dir;
613
614}
615sub get_doc_dir {
616
617 my $self = shift (@_);
618 my ($doc_obj) = @_;
619
620 my $OID = $doc_obj->get_OID();
621 $OID = "NULL" unless defined $OID;
622
623 my $working_dir = $self->get_output_dir();
624 my $working_info = $self->{'output_info'};
625 return if (!defined $working_info);
626
627 my $doc_info = $working_info->get_info($OID);
628 my $doc_dir = '';
629
630 if (defined $doc_info && scalar(@$doc_info) >= 1)
631 {
632 # This OID already has an archives directory, so use it again
633 $doc_dir = $doc_info->[0];
634 $doc_dir =~ s/\/?((doc(mets|sql)?)|(dublin_core))\.xml(\.gz)?$//;
635 }
636 elsif ($self->{'keep_import_structure'})
637 {
638 my $source_filename = $doc_obj->get_source_filename();
639 $source_filename = &File::Basename::dirname($source_filename);
640 $source_filename =~ s/[\\\/]+/\//g;
641 $source_filename =~ s/\/$//;
642
643 $doc_dir = substr($source_filename, length($ENV{'GSDLIMPORTDIR'}) + 1);
644 }
645
646 # We have to use a new archives directory for this document
647 if ($doc_dir eq "")
648 {
649 $doc_dir = $self->get_new_doc_dir ($working_info, $working_dir, $OID);
650 }
651
652 &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir));
653
654 return $doc_dir;
655}
656
657
658## @function get_new_doc_dir()
659#
660# Once a doc object is ready to write to disk (and hence has a nice OID),
661# generate a unique subdirectory to write the information to.
662#
663# - create the directory as part of this call, to try and avoid race conditions
664# found in parallel processing [jmt12]
665#
666# @todo figure out what the rule regarding $work_info->size() is meant to do
667#
668# @todo determine what $self->{'group'} is, and whether it should affect
669# directory creation
670#
671sub get_new_doc_dir
672{
673 my $self = shift (@_);
674 my($working_info,$working_dir,$OID) = @_;
675
676 my $doc_dir = "";
677 my $doc_dir_rest = $OID;
678
679 # remove any \ and / from the OID
680 $doc_dir_rest =~ s/[\\\/]//g;
681
682 # Remove ":" if we are on Windows OS, as otherwise they get confused with the drive letters
683 if ($ENV{'GSDLOS'} =~ /^windows$/i)
684 {
685 $doc_dir_rest =~ s/\://g;
686 }
687
688 # we generally create a unique directory by adding consequtive fragments of
689 # the document identifier (split by some predefined length - defaulting to
690 # 8) until we find a directory that doesn't yet exist. Note that directories
691 # that contain a document have a suffix ".dir" (whereas those that contain
692 # only subdirectories have no suffix).
693 my $doc_dir_num = 0; # how many directories deep we are
694 my $created_directory = 0; # have we successfully created a new directory
695 do
696 {
697 # (does this work on windows? - jmt12)
698 if ($doc_dir_num > 0)
699 {
700 $doc_dir .= '/';
701 }
702
703 # the default matching pattern grabs the next 'subdir_split_length'
704 # characters of the OID to act as the next subdirectory
705 my $pattern = '^(.{1,' . $self->{'subdir_split_length'} . '})';
706
707 # Do we count any "HASH" prefix against the split length limit?
708 if ($self->{'subdir_hash_prefix'} && $doc_dir_num == 0)
709 {
710 $pattern = '^((HASH)?.{1,' . $self->{'subdir_split_length'} . '})';
711 }
712
713 # Note the use of 's' to both capture the next chuck of OID and to remove
714 # it from OID at the same time
715 if ($doc_dir_rest =~ s/$pattern//i)
716 {
717 $doc_dir .= $1;
718 $doc_dir_num++;
719
720 my $full_doc_dir = &FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir');
721 if(!FileUtils::directoryExists($full_doc_dir))
722 {
723 &FileUtils::makeAllDirectories($full_doc_dir);
724 $created_directory = 1;
725 }
726
727 ###rint STDERR "[DEBUG] BasePlugout::get_new_doc_dir(<working_info>, $working_dir, $oid)\n";
728 ###rint STDERR " - create directory: $full_doc_dir => $created_directory\n";
729 ###rint STDERR " - rest: $doc_dir_rest\n";
730 ###rint STDERR " - working_info->size(): " . $working_info->size() . " [ < 1024 ?]\n";
731 ###rint STDERR " - doc_dir_num: " . $doc_dir_num . "\n";
732 }
733 }
734 while ($doc_dir_rest ne '' && ($created_directory == 0 || ($working_info->size() >= 1024 && $doc_dir_num < 2)));
735
736 # not unique yet? Add on an incremental suffix until we are unique
737 my $i = 1;
738 my $doc_dir_base = $doc_dir;
739 while ($created_directory == 0)
740 {
741 $doc_dir = $doc_dir_base . '-' . $i;
742 $created_directory = &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir'));
743 $i++;
744 }
745
746 # in theory this should never happen
747 if (!$created_directory)
748 {
749 die("Error! Failed to create directory for document: " . $doc_dir_base . "\n");
750 }
751
752 return $doc_dir . '.dir';
753}
754## get_new_doc_dir()
755
756
757sub process_assoc_files {
758 my $self = shift (@_);
759 my ($doc_obj, $doc_dir, $handle) = @_;
760
761 my $outhandle = $self->{'output_handle'};
762
763 my $output_dir = $self->get_output_dir();
764 return if (!defined $output_dir);
765
766 &FileUtils::makeAllDirectories($output_dir) unless &FileUtils::directoryExists($output_dir);
767
768 my $working_dir = &FileUtils::filenameConcatenate($output_dir, $doc_dir);
769 &FileUtils::makeAllDirectories($working_dir) unless &FileUtils::directoryExists($working_dir);
770
771 my @assoc_files = ();
772 my $filename;;
773
774 my $source_filename = $doc_obj->get_source_filename();
775
776 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
777
778 if (defined $collect_dir) {
779 my $dirsep_regexp = &util::get_os_dirsep();
780
781 if ($collect_dir !~ /$dirsep_regexp$/) {
782 $collect_dir .= &util::get_dirsep(); # ensure there is a slash at the end
783 }
784
785 # This test is never going to fail on Windows -- is this a problem?
786
787 if ($source_filename !~ /^$dirsep_regexp/) {
788 $source_filename = &FileUtils::filenameConcatenate($collect_dir, $source_filename);
789 }
790 }
791
792
793 # set the assocfile path (even if we have no assoc files - need this for lucene)
794 $doc_obj->set_utf8_metadata_element ($doc_obj->get_top_section(),
795 "assocfilepath",
796 "$doc_dir");
797 foreach my $assoc_file_rec (@{$doc_obj->get_assoc_files()}) {
798 my ($dir, $afile) = $assoc_file_rec->[1] =~ /^(.*?)([^\/\\]+)$/;
799 $dir = "" unless defined $dir;
800
801 my $utf8_real_filename = $assoc_file_rec->[0];
802
803 # for some reasons the image associate file has / before the full path
804 $utf8_real_filename =~ s/^\\(.*)/$1/i;
805
806## my $real_filename = &util::utf8_to_real_filename($utf8_real_filename);
807 my $real_filename = $utf8_real_filename;
808 $real_filename = &util::downgrade_if_dos_filename($real_filename);
809
810 if (&FileUtils::fileExists($real_filename)) {
811
812 $filename = &FileUtils::filenameConcatenate($working_dir, $afile);
813
814 &FileUtils::hardLink($real_filename, $filename, $self->{'verbosity'});
815
816 $doc_obj->add_utf8_metadata ($doc_obj->get_top_section(),
817 "gsdlassocfile",
818 "$afile:$assoc_file_rec->[2]:$dir");
819 } elsif ($self->{'verbosity'} > 1) {
820 print $outhandle "BasePlugout::process couldn't copy the associated file " .
821 "$real_filename to $afile\n";
822 }
823 }
824}
825
826
827sub process_metafiles_metadata
828{
829 my $self = shift (@_);
830 my ($doc_obj) = @_;
831
832 my $top_section = $doc_obj->get_top_section();
833 my $metafiles = $doc_obj->get_metadata($top_section,"gsdlmetafile");
834
835 foreach my $metafile_pair (@$metafiles) {
836 my ($full_metafile,$metafile) = split(/ : /,$metafile_pair);
837
838 $doc_obj->metadata_file($full_metafile,$metafile);
839 }
840
841 $doc_obj->delete_metadata($top_section,"gsdlmetafile");
842}
843
844sub archiveinf_files_to_field
845{
846 my $self = shift(@_);
847 my ($files,$field,$collect_dir,$oid_files,$reverse_lookups) = @_;
848
849 foreach my $file_rec (@$files) {
850 my $real_filename = (ref $file_rec eq "ARRAY") ? $file_rec->[0] : $file_rec;
851 my $full_file = (ref $file_rec eq "ARRAY") ? $file_rec->[1] : $file_rec;
852 # for some reasons the image associate file has / before the full path
853 $real_filename =~ s/^\\(.*)/$1/i;
854
855 my $raw_filename = &util::downgrade_if_dos_filename($real_filename);
856
857 if (&FileUtils::fileExists($raw_filename)) {
858
859# if (defined $collect_dir) {
860# my $collect_dir_re_safe = $collect_dir;
861# $collect_dir_re_safe =~ s/\\/\\\\/g; # use &util::filename_to_regex()
862# $collect_dir_re_safe =~ s/\./\\./g;##
863
864# $real_filename =~ s/^$collect_dir_re_safe//;
865# }
866
867 if (defined $reverse_lookups) {
868 $reverse_lookups->{$real_filename} = 1;
869 }
870
871 if($field =~ m@assoc-file|src-file|meta-file@) {
872 $raw_filename = &util::abspath_to_placeholders($raw_filename);
873 }
874
875### push(@{$oid_files->{$field}},$full_file);
876 push(@{$oid_files->{$field}},$raw_filename);
877 }
878 else {
879 print STDERR "Warning: archiveinf_files_to_field()\n $real_filename does not appear to be on the file system\n";
880 }
881 }
882}
883
884sub archiveinf_db
885{
886 my $self = shift (@_);
887 my ($doc_obj) = @_;
888
889 my $verbosity = $self->{'verbosity'};
890
891 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
892 if (defined $collect_dir) {
893 my $dirsep_regexp = &util::get_os_dirsep();
894
895 if ($collect_dir !~ /$dirsep_regexp$/) {
896 # ensure there is a slash at the end
897 $collect_dir .= &util::get_dirsep();
898 }
899 }
900
901 my $oid = $doc_obj->get_OID();
902 my $source_filename = $doc_obj->get_unmodified_source_filename();
903 my $working_info = $self->{'output_info'};
904 my $doc_info = $working_info->get_info($oid);
905
906 my ($doc_file,$index_status,$sortmeta, $group_position) = @$doc_info;
907 # doc_file is the path to the archive doc.xml. Make sure it has unix
908 # slashes, then if the collection is copied to linux, it can be built without reimport
909 $doc_file =~ s/\\/\//g;
910 my $oid_files = { 'doc-file' => $doc_file,
911 'index-status' => $index_status,
912 'src-file' => $source_filename,
913 'sort-meta' => $sortmeta,
914 'assoc-file' => [],
915 'meta-file' => [] };
916 if (defined $group_position) {
917 $oid_files->{'group-position'} = $group_position;
918 }
919 my $reverse_lookups = { $source_filename => "1" };
920
921
922 $self->archiveinf_files_to_field($doc_obj->get_source_assoc_files(),"assoc-file",
923 $collect_dir,$oid_files,$reverse_lookups);
924
925
926 $self->archiveinf_files_to_field($doc_obj->get_meta_files(),"meta-file",
927 $collect_dir,$oid_files);
928
929 # Get the infodbtype value for this collection from the arcinfo object
930 my $infodbtype = $self->{'output_info'}->{'infodbtype'};
931 my $output_dir = $self->{'output_dir'};
932
933 my $doc_db = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $output_dir);
934
935 ##print STDERR "*** To set in db: \n\t$doc_db\n\t$oid\n\t$doc_db_text\n";
936
937 if (!$self->{'no_rss'})
938 {
939 if (($oid_files->{'index-status'} eq "I") || ($oid_files->{'index-status'} eq "R")) {
940 my $top_section = $doc_obj->get_top_section();
941
942 # rss_title can be set in collect.cfg as follows:
943 # plugout GreenstoneXMLPlugout -rss_title "dc.Title; ex.Title"
944 # rss_title is a semi-colon or comma-separated list of the metadata field names that should
945 # be consulted in order to obtain a Title (anchor text) for the RSS document link.
946 # If not specified, rss_title will default to dc.Title, and fall back on Untitled
947 my $metafieldnames = $self->{'rss_title'};
948 my @metafieldarray = split(/[,;] ?/,$metafieldnames); # , or ; separator can be followed by an optional space
949 my $titles;
950 #@$titles=(); # at worst @$titles will be (), as get_metadata(dc.Titles) may return ()
951 foreach my $metafieldname (@metafieldarray) {
952 $metafieldname =~ s@^ex\.@@; # if ex.Title, need to get_metadata() on metafieldname=Title
953 $titles = $doc_obj->get_metadata($top_section,$metafieldname);
954
955 if(scalar(@$titles) != 0) { # found at least one title for one metafieldname
956 last; # break out of the loop
957 }
958 }
959
960 # if ex.Title was listed in the metafieldnames, then we'll surely have a value for title for this doc
961 # otherwise, if we have no titles at this point, add in a default of Untitled as this doc's title
962 if(scalar(@$titles) == 0) { #&& $metafieldnames !~ [email protected]@) {
963 push(@$titles, "Untitled");
964 }
965
966 # encode basic html entities like <>"& in the title(s), since the & char can break RSS links
967 for (my $i = 0; $i < scalar(@$titles); $i++) {
968 &ghtml::htmlsafe(@$titles[$i]);
969 }
970
971 my $dc_title = join("; ", @$titles);
972
973 if ($oid_files->{'index-status'} eq "R") {
974 $dc_title .= " (Updated)";
975 }
976
977 my $rss_entry = "<item>\n";
978 $rss_entry .= " <title>$dc_title</title>\n";
979 if(&util::is_gs3()) {
980 $rss_entry .= " <link>_httpdomain__httpcollection_/document/$oid</link>\n";
981 } else {
982 $rss_entry .= " <link>_httpdomainHtmlsafe__httpcollection_/document/$oid</link>\n";
983 }
984 $rss_entry .= "</item>";
985
986 if (defined(&dbutil::supportsRSS) && &dbutil::supportsRSS($infodbtype))
987 {
988 my $rss_db = &dbutil::get_infodb_file_path($infodbtype, 'rss-items', $output_dir);
989 my $rss_db_fh = &dbutil::open_infodb_write_handle($infodbtype, $rss_db, 'append');
990 &dbutil::write_infodb_rawentry($infodbtype, $rss_db_fh, $oid, $rss_entry);
991 &dbutil::close_infodb_write_handle($infodbtype, $rss_db_fh);
992 }
993 else
994 {
995 my $rss_filename = &FileUtils::filenameConcatenate($output_dir,"rss-items.rdf");
996 my $rss_fh;
997 if (&FileUtils::openFileHandle($rss_filename, '>>', \$rss_fh, "utf8"))
998 {
999 print $rss_fh $rss_entry . "\n";
1000 &FileUtils::closeFileHandle($rss_filename, \$rss_fh);
1001 }
1002 else
1003 {
1004 print STDERR "**** Failed to open $rss_filename\n$!\n";
1005 }
1006 }
1007 }
1008 }
1009
1010 $oid_files->{'doc-file'} = [ $oid_files->{'doc-file'} ];
1011 $oid_files->{'index-status'} = [ $oid_files->{'index-status'} ];
1012 $oid_files->{'src-file'} = &util::abspath_to_placeholders($oid_files->{'src-file'});
1013 $oid_files->{'src-file'} = [ $oid_files->{'src-file'} ];
1014 $oid_files->{'sort-meta'} = [ $oid_files->{'sort-meta'} ];
1015 if (defined $oid_files->{'group-position'}) {
1016 $oid_files->{'group-position'} = [ $oid_files->{'group-position'} ];
1017 }
1018
1019 my $infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $doc_db, "append");
1020 &dbutil::write_infodb_entry($infodbtype, $infodb_file_handle, $oid, $oid_files);
1021 &dbutil::close_infodb_write_handle($infodbtype, $infodb_file_handle);
1022
1023 foreach my $rl (keys %$reverse_lookups) {
1024 $working_info->add_reverseinfo($rl,$oid);
1025 }
1026
1027 # meta files not set in reverse entry, but need to set the metadata flag
1028 if (defined $doc_obj->get_meta_files()) {
1029 foreach my $meta_file_rec(@{$doc_obj->get_meta_files()}) {
1030 my $full_file = (ref $meta_file_rec eq "ARRAY") ? $meta_file_rec->[0] : $meta_file_rec;
1031 $working_info->set_meta_file_flag($full_file);
1032 }
1033 }
1034}
1035
1036# This sub is called for every metadata.xml accepted for processing by by MetdataXMLPlugin
1037# and adds an entry into archiveinf-src.db for that file in the form:
1038# [@THISCOLLECTPATH@/import/metadata.xml]
1039# <meta-file>1
1040# This prevents blind reprocessing of the same old docs upon *incremental* building whenever
1041# we encounter a default empty metadata.xml that has no actual <FileSet> content defined.
1042sub add_metaxml_file_entry_to_archiveinfsrc {
1043 my $self = shift (@_);
1044 my ($full_file) = @_;
1045
1046 print STDERR "**** Adding metaxml file entry for full_file: $full_file\n";
1047 my $working_info = $self->{'output_info'};
1048 $working_info->set_meta_file_flag($full_file);
1049}
1050
1051
1052sub set_sortmeta {
1053 my $self = shift (@_);
1054 my ($sortmeta, $removeprefix, $removesuffix) = @_;
1055
1056 $self->{'sortmeta'} = $sortmeta;
1057 if (defined ($removeprefix) && $removeprefix ) {
1058 $removeprefix =~ s/^\^//; # don't need a leading ^
1059 $self->{'removeprefix'} = $removeprefix;
1060 }
1061 if (defined ($removesuffix) && $removesuffix) {
1062 $removesuffix =~ s/\$$//; # don't need a trailing $
1063 $self->{'removesuffix'} = $removesuffix;
1064 }
1065}
1066
1067
1068
1069sub open_xslt_pipe
1070{
1071 my $self = shift @_;
1072 my ($output_file_name, $xslt_file)=@_;
1073
1074 return unless defined $xslt_file and $xslt_file ne "" and &FileUtils::fileExists($xslt_file);
1075
1076 my $java_class_path = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","ApplyXSLT.jar");
1077
1078 my $mapping_file_path = "";
1079
1080 if ($ENV{'GSDLOS'} eq "windows"){
1081 $java_class_path .=";".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
1082 # this file:/// bit didn't work for me on windows XP
1083 #$xslt_file = "\"file:///".$xslt_file."\"";
1084 #$mapping_file_path = "\"file:///";
1085 }
1086 else{
1087 $java_class_path .=":".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
1088 }
1089
1090
1091 $java_class_path = "\"".$java_class_path."\"";
1092
1093 my $cmd = "| java -cp $java_class_path org.nzdl.gsdl.ApplyXSLT -t \"$xslt_file\" ";
1094
1095 if (defined $self->{'mapping_file'} and $self->{'mapping_file'} ne ""){
1096 my $mapping_file_path = "\"".$self->{'mapping_file'}."\"";
1097 $cmd .= "-m $mapping_file_path";
1098 }
1099
1100 open(*XMLWRITER, $cmd)
1101 or die "can't open pipe to xslt: $!";
1102
1103
1104 $self->{'xslt_writer'} = *XMLWRITER;
1105
1106 print XMLWRITER "<?DocStart?>\n";
1107 print XMLWRITER "$output_file_name\n";
1108
1109
1110 }
1111
1112
1113sub close_xslt_pipe
1114{
1115 my $self = shift @_;
1116
1117
1118 return unless defined $self->{'xslt_writer'} ;
1119
1120 my $xsltwriter = $self->{'xslt_writer'};
1121
1122 print $xsltwriter "<?DocEnd?>\n";
1123 close($xsltwriter);
1124
1125 undef $self->{'xslt_writer'};
1126
1127}
1128
1129
1130
1131#the subclass should implement this method if is_group method could return 1.
1132sub close_group_output{
1133 my $self = shift (@_);
1134}
1135
1136sub is_group {
1137 my $self = shift (@_);
1138 return 0;
1139}
1140
1141my $dc_set = { Title => 1,
1142 Creator => 1,
1143 Subject => 1,
1144 Description => 1,
1145 Publisher => 1,
1146 Contributor => 1,
1147 Date => 1,
1148 Type => 1,
1149 Format => 1,
1150 Identifier => 1,
1151 Source => 1,
1152 Language => 1,
1153 Relation => 1,
1154 Coverage => 1,
1155 Rights => 1};
1156
1157
1158# returns an XML representation of the dublin core metadata
1159# if dc meta is not found, try ex meta
1160# This method is not used by the DSpacePlugout, which has its
1161# own method to save its dc metadata
1162sub get_dc_metadata {
1163 my $self = shift(@_);
1164 my ($doc_obj, $section, $version) = @_;
1165
1166 # build up string of dublin core metadata
1167 $section="" unless defined $section;
1168
1169 my $section_ptr = $doc_obj->_lookup_section($section);
1170 return "" unless defined $section_ptr;
1171
1172
1173 my $explicit_dc = {};
1174 my $explicit_ex_dc = {};
1175 my $explicit_ex = {};
1176
1177 my $all_text="";
1178
1179 # We want high quality dc metadata to go in first, so we store all the
1180 # assigned dc.* values first. Then, for all those dc metadata names in
1181 # the official dc set that are as yet unassigned, we look to see whether
1182 # embedded ex.dc.* metadata has defined some values for them. If not,
1183 # then for the same missing dc metadata names, we look in ex metadata.
1184
1185 foreach my $data (@{$section_ptr->{'metadata'}}){
1186 my $escaped_value = &docprint::escape_text($data->[1]);
1187 if ($data->[0]=~ m/^dc\./) {
1188 $data->[0] =~ tr/[A-Z]/[a-z]/;
1189
1190 $data->[0] =~ m/^dc\.(.*)/;
1191 my $dc_element = $1;
1192
1193 if (!defined $explicit_dc->{$dc_element}) {
1194 $explicit_dc->{$dc_element} = [];
1195 }
1196 push(@{$explicit_dc->{$dc_element}},$escaped_value);
1197
1198 if (defined $version && ($version eq "oai_dc")) {
1199 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1200 }
1201 else {
1202 # qualifier???
1203 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1204 }
1205
1206 } elsif ($data->[0]=~ m/^ex\.dc\./) { # now look through ex.dc.* to fill in as yet unassigned fields in dc metaset
1207 $data->[0] =~ m/^ex\.dc\.(.*)/;
1208 my $ex_dc_element = $1;
1209 my $lc_ex_dc_element = lc($ex_dc_element);
1210
1211 # only store the ex.dc value for this dc metaname if no dc.* was assigned for it
1212 if (defined $dc_set->{$ex_dc_element}) {
1213 if (!defined $explicit_ex_dc->{$lc_ex_dc_element}) {
1214 $explicit_ex_dc->{$lc_ex_dc_element} = [];
1215 }
1216 push(@{$explicit_ex_dc->{$lc_ex_dc_element}},$escaped_value);
1217 }
1218 }
1219 elsif (($data->[0] =~ m/^ex\./) || ($data->[0] !~ m/\./)) { # look through ex. meta (incl. meta without prefix)
1220 $data->[0] =~ m/^(ex\.)?(.*)/;
1221 my $ex_element = $2;
1222 my $lc_ex_element = lc($ex_element);
1223
1224 if (defined $dc_set->{$ex_element}) {
1225 if (!defined $explicit_ex->{$lc_ex_element}) {
1226 $explicit_ex->{$lc_ex_element} = [];
1227 }
1228 push(@{$explicit_ex->{$lc_ex_element}},$escaped_value);
1229 }
1230 }
1231 }
1232
1233 # go through dc_set and for any element *not* defined in explicit_dc
1234 # that does exist in explicit_ex, add it in as metadata
1235 foreach my $k ( keys %$dc_set ) {
1236 my $lc_k = lc($k);
1237
1238 if (!defined $explicit_dc->{$lc_k}) {
1239 # try to find if ex.dc.* defines this dc.* meta,
1240 # if not, then look for whether there's an ex.* equivalent
1241
1242 if (defined $explicit_ex_dc->{$lc_k}) {
1243 foreach my $v (@{$explicit_ex_dc->{$lc_k}}) {
1244 my $dc_element = $lc_k;
1245 my $escaped_value = $v;
1246
1247 if (defined $version && ($version eq "oai_dc")) {
1248 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1249 }
1250 else {
1251 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1252 }
1253 }
1254 } elsif (defined $explicit_ex->{$lc_k}) {
1255 foreach my $v (@{$explicit_ex->{$lc_k}}) {
1256 my $dc_element = $lc_k;
1257 my $escaped_value = $v;
1258
1259 if (defined $version && ($version eq "oai_dc")) {
1260 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1261 }
1262 else {
1263 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1264 }
1265 }
1266 }
1267 }
1268 }
1269
1270 if ($all_text eq "") {
1271 $all_text .= " There is no Dublin Core metatdata in this document\n";
1272 }
1273 $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;
1274
1275 return $all_text;
1276}
1277
1278# Build up dublin_core metadata. Priority given to dc.* over ex.*
1279# This method was apparently added by Jeffrey and committed by Shaoqun.
1280# But we don't know why it was added, so not using it anymore.
1281sub new_get_dc_metadata {
1282
1283 my $self = shift(@_);
1284 my ($doc_obj, $section, $version) = @_;
1285
1286 # build up string of dublin core metadata
1287 $section="" unless defined $section;
1288
1289 my $section_ptr=$doc_obj->_lookup_section($section);
1290 return "" unless defined $section_ptr;
1291
1292 my $all_text = "";
1293 foreach my $data (@{$section_ptr->{'metadata'}}){
1294 my $escaped_value = &docprint::escape_text($data->[1]);
1295 my $dc_element = $data->[0];
1296
1297 my @array = split('\.',$dc_element);
1298 my ($type,$name);
1299
1300 if(defined $array[1])
1301 {
1302 $type = $array[0];
1303 $name = $array[1];
1304 }
1305 else
1306 {
1307 $type = "ex";
1308 $name = $array[0];
1309 }
1310
1311 $all_text .= ' <Metadata Type="'. $type.'" Name="'.$name.'">'. $escaped_value. "</Metadata>\n";
1312 }
1313 return $all_text;
1314}
1315
1316
13171;
Note: See TracBrowser for help on using the repository browser.