source: main/trunk/greenstone2/perllib/plugouts/BasePlugout.pm@ 27646

Last change on this file since 27646 was 27646, checked in by jmt12, 11 years ago

Adding an option to allow me to suppress RSS file writing, -no_rss, as it requires file append, which isn't supported in HDFS-based collections
  • Property svn:keywords set to Author Date Id Revision
File size: 35.2 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;
40
41# suppress the annoying "subroutine redefined" warning that various
42# gets cause under perl 5.6
43$SIG{__WARN__} = sub {warn($_[0]) unless ($_[0] =~ /Subroutine\s+\S+\sredefined/)};
44
45my $arguments = [
46 { 'name' => "group_size",
47 'desc' => "{BasPlugout.group_size}",
48 'type' => "int",
49 'deft' => "1",
50 'reqd' => "no",
51 'hiddengli' => "no"},
52 { 'name' => "output_info",
53 'desc' => "{BasPlugout.output_info}",
54 'type' => "string",
55 'reqd' => "yes",
56 'hiddengli' => "yes"},
57 { 'name' => "xslt_file",
58 'desc' => "{BasPlugout.xslt_file}",
59 'type' => "string",
60 'reqd' => "no",
61 'deft' => "",
62 'hiddengli' => "no"},
63 { 'name' => "output_handle",
64 'desc' => "{BasPlugout.output_handle}",
65 'type' => "string",
66 'deft' => 'STDERR',
67 'reqd' => "no",
68 'hiddengli' => "yes"},
69 { 'name' => "verbosity",
70 'desc' => "{BasPlugout.verbosity}",
71 'type' => "int",
72 'deft' => "0",
73 'reqd' => "no",
74 'hiddengli' => "no"},
75 { 'name' => "gzip_output",
76 'desc' => "{BasPlugout.gzip_output}",
77 'type' => "flag",
78 'reqd' => "no",
79 'hiddengli' => "no"},
80 { 'name' => "debug",
81 'desc' => "{BasPlugout.debug}",
82 'type' => "flag",
83 'reqd' => "no",
84 'hiddengli' => "yes"},
85 { 'name' => "subdir_split_length",
86 'desc' => "{BasPlugout.subdir_split_length}",
87 'type' => "int",
88 'reqd' => "no",
89 'deft' => "8",
90 'hiddengli' => "no"},
91 { 'name' => "subdir_hash_prefix",
92 'desc' => "{BasPlugout.subdir_hash_prefix}",
93 'type' => "flag",
94 'reqd' => "no",
95 'deft' => "false",
96 'hiddengli' => "no"},
97 { 'name' => 'no_rss',
98 'desc' => '{BasPlugout.no_rss}',
99 'type' => 'flag',
100 'reqd' => 'no',
101 'hiddengli' => 'yes'}
102];
103
104my $options = { 'name' => "BasePlugout",
105 'desc' => "{BasPlugout.desc}",
106 'abstract' => "yes",
107 'inherits' => "no",
108 'args' => $arguments};
109
110sub new
111{
112 my $class = shift (@_);
113
114 my ($plugoutlist,$args,$hashArgOptLists) = @_;
115 push(@$plugoutlist, $class);
116
117 my $plugout_name = (defined $plugoutlist->[0]) ? $plugoutlist->[0] : $class;
118
119 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
120 push(@{$hashArgOptLists->{"OptList"}},$options);
121
122 my $self = {};
123 $self->{'plugout_type'} = $class;
124 $self->{'option_list'} = $hashArgOptLists->{"OptList"};
125 $self->{"info_only"} = 0;
126
127 # Check if gsdlinfo is in the argument list or not - if it is, don't parse
128 # the args, just return the object.
129 foreach my $strArg (@{$args})
130 {
131 if(defined $strArg && $strArg eq "-gsdlinfo")
132 {
133 $self->{"info_only"} = 1;
134 return bless $self, $class;
135 }
136 }
137
138 delete $self->{"info_only"};
139
140 if(parse2::parse($args,$hashArgOptLists->{"ArgList"},$self) == -1)
141 {
142 my $classTempClass = bless $self, $class;
143 print STDERR "<BadPlugout d=$plugout_name>\n";
144 &gsprintf(STDERR, "\n{BasPlugout.bad_general_option}\n", $plugout_name);
145 $classTempClass->print_txt_usage(""); # Use default resource bundle
146 die "\n";
147 }
148
149
150 if(defined $self->{'xslt_file'} && $self->{'xslt_file'} ne "")
151 {
152 my $full_file_path = &util::locate_config_file($self->{'xslt_file'});
153 if (!defined $full_file_path) {
154 print STDERR "Can not find $self->{'xslt_file'}, please make sure you have supplied the correct file path\n";
155 die "\n";
156 }
157 $self->{'xslt_file'} = $full_file_path;
158 }
159
160 $self->{'gs_count'} = 0;
161
162 $self->{'keep_import_structure'} = 0;
163
164 return bless $self, $class;
165
166}
167
168sub print_xml_usage
169{
170 my $self = shift(@_);
171 my $header = shift(@_);
172 my $high_level_information_only = shift(@_);
173
174 # XML output is always in UTF-8
175 gsprintf::output_strings_in_UTF8;
176
177 if ($header) {
178 &PrintUsage::print_xml_header("plugout");
179 }
180 $self->print_xml($high_level_information_only);
181}
182
183
184sub print_xml
185{
186 my $self = shift(@_);
187 my $high_level_information_only = shift(@_);
188
189 my $optionlistref = $self->{'option_list'};
190 my @optionlist = @$optionlistref;
191 my $plugoutoptions = shift(@$optionlistref);
192 return if (!defined($plugoutoptions));
193
194 gsprintf(STDERR, "<PlugoutInfo>\n");
195 gsprintf(STDERR, " <Name>$plugoutoptions->{'name'}</Name>\n");
196 my $desc = gsprintf::lookup_string($plugoutoptions->{'desc'});
197 $desc =~ s/</&amp;lt;/g; # doubly escaped
198 $desc =~ s/>/&amp;gt;/g;
199 gsprintf(STDERR, " <Desc>$desc</Desc>\n");
200 gsprintf(STDERR, " <Abstract>$plugoutoptions->{'abstract'}</Abstract>\n");
201 gsprintf(STDERR, " <Inherits>$plugoutoptions->{'inherits'}</Inherits>\n");
202 unless (defined($high_level_information_only)) {
203 gsprintf(STDERR, " <Arguments>\n");
204 if (defined($plugoutoptions->{'args'})) {
205 &PrintUsage::print_options_xml($plugoutoptions->{'args'});
206 }
207 gsprintf(STDERR, " </Arguments>\n");
208
209 # Recurse up the plugout hierarchy
210 $self->print_xml();
211 }
212 gsprintf(STDERR, "</PlugoutInfo>\n");
213}
214
215
216sub print_txt_usage
217{
218 my $self = shift(@_);
219
220 # Print the usage message for a plugout (recursively)
221 my $descoffset = $self->determine_description_offset(0);
222 $self->print_plugout_usage($descoffset, 1);
223}
224
225sub determine_description_offset
226{
227 my $self = shift(@_);
228 my $maxoffset = shift(@_);
229
230 my $optionlistref = $self->{'option_list'};
231 my @optionlist = @$optionlistref;
232 my $plugoutoptions = pop(@$optionlistref);
233 return $maxoffset if (!defined($plugoutoptions));
234
235 # Find the length of the longest option string of this download
236 my $plugoutargs = $plugoutoptions->{'args'};
237 if (defined($plugoutargs)) {
238 my $longest = &PrintUsage::find_longest_option_string($plugoutargs);
239 if ($longest > $maxoffset) {
240 $maxoffset = $longest;
241 }
242 }
243
244 # Recurse up the download hierarchy
245 $maxoffset = $self->determine_description_offset($maxoffset);
246 $self->{'option_list'} = \@optionlist;
247 return $maxoffset;
248}
249
250
251sub print_plugout_usage
252{
253 my $self = shift(@_);
254 my $descoffset = shift(@_);
255 my $isleafclass = shift(@_);
256
257 my $optionlistref = $self->{'option_list'};
258 my @optionlist = @$optionlistref;
259 my $plugoutoptions = shift(@$optionlistref);
260 return if (!defined($plugoutoptions));
261
262 my $plugoutname = $plugoutoptions->{'name'};
263 my $plugoutargs = $plugoutoptions->{'args'};
264 my $plugoutdesc = $plugoutoptions->{'desc'};
265
266 # Produce the usage information using the data structure above
267 if ($isleafclass) {
268 if (defined($plugoutdesc)) {
269 gsprintf(STDERR, "$plugoutdesc\n\n");
270 }
271 gsprintf(STDERR, " {common.usage}: plugout $plugoutname [{common.options}]\n\n");
272 }
273
274 # Display the download options, if there are some
275 if (defined($plugoutargs)) {
276 # Calculate the column offset of the option descriptions
277 my $optiondescoffset = $descoffset + 2; # 2 spaces between options & descriptions
278
279 if ($isleafclass) {
280 gsprintf(STDERR, " {common.specific_options}:\n");
281 }
282 else {
283 gsprintf(STDERR, " {common.general_options}:\n", $plugoutname);
284 }
285
286 # Display the download options
287 &PrintUsage::print_options_txt($plugoutargs, $optiondescoffset);
288 }
289
290 # Recurse up the download hierarchy
291 $self->print_plugout_usage($descoffset, 0);
292 $self->{'option_list'} = \@optionlist;
293}
294
295
296sub error
297{
298 my ($strFunctionName,$strError) = @_;
299 {
300 print "Error occoured in BasePlugout.pm\n".
301 "In Function: ".$strFunctionName."\n".
302 "Error Message: ".$strError."\n";
303 exit(-1);
304 }
305}
306
307# OIDtype may be "hash" or "hash_on_full_filename" or "incremental" or "filename" or "dirname" or "full_filename" or "assigned"
308sub set_OIDtype {
309 my $self = shift (@_);
310 my ($type, $metadata) = @_;
311
312 if ($type =~ /^(hash|hash_on_full_filename|incremental|filename|dirname|full_filename|assigned)$/) {
313 $self->{'OIDtype'} = $type;
314 } else {
315 $self->{'OIDtype'} = "hash";
316 }
317 if ($type =~ /^assigned$/) {
318 if (defined $metadata) {
319 $self->{'OIDmetadata'} = $metadata;
320 } else {
321 $self->{'OIDmetadata'} = "dc.Identifier";
322 }
323 }
324}
325
326sub set_output_dir
327{
328 my $self = shift @_;
329 my ($output_dir) = @_;
330
331 $self->{'output_dir'} = $output_dir;
332}
333
334sub setoutputdir
335{
336 my $self = shift @_;
337 my ($output_dir) = @_;
338
339 $self->{'output_dir'} = $output_dir;
340}
341
342sub get_output_dir
343{
344 my $self = shift (@_);
345
346 return $self->{'output_dir'};
347}
348
349sub getoutputdir
350{
351 my $self = shift (@_);
352
353 return $self->{'output_dir'};
354}
355
356sub getoutputinfo
357{
358 my $self = shift (@_);
359
360 return $self->{'output_info'};
361}
362
363
364sub get_output_handler
365{
366 my $self = shift (@_);
367
368 my ($output_file_name) = @_;
369
370 my $fh;
371 &FileUtils::openFileHandle($output_file_name, '>', \$fh) or die('Can not open a file handler for: ' . $output_file_name . "\n");
372
373 return $fh;
374}
375
376sub release_output_handler
377{
378 my $self = shift (@_);
379 my ($outhandler) = @_;
380
381 close($outhandler);
382
383}
384
385sub output_xml_header {
386 my $self = shift (@_);
387 my ($handle,$docroot,$nondoctype) = @_;
388
389
390 #print $handle '<?xml version="1.0" encoding="UTF-8" standalone="no"?>' . "\n";
391
392 #For Dspace must be UTF in lower case
393 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
394
395 if (!defined $nondoctype){
396 my $doctype = (defined $docroot) ? $docroot : "Section";
397
398 # Used to be '<!DOCTYPE Archive SYSTEM ...'
399
400 print $handle "<!DOCTYPE $doctype SYSTEM \"http://greenstone.org/dtd/Archive/1.0/Archive.dtd\">\n";
401 }
402
403 print $handle "<$docroot>\n" if defined $docroot;
404}
405
406sub output_xml_footer {
407 my $self = shift (@_);
408 my ($handle,$docroot) = @_;
409 print $handle "</$docroot>\n" if defined $docroot;
410}
411
412
413sub output_general_xml_header
414{
415 my $self = shift (@_);
416 my ($handle,$docroot,$opt_attributes,$opt_dtd, $opt_doctype) = @_;
417
418 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
419
420 if (defined $opt_dtd) {
421 my $doctype = (defined $opt_doctype) ? $opt_doctype : $docroot;
422 print $handle "<!DOCTYPE $doctype SYSTEM \"$opt_dtd\">\n";
423 }
424
425 if (defined $docroot) {
426 my $full_docroot = $docroot;
427 if (defined $opt_attributes) {
428 $full_docroot .= " $opt_attributes";
429 }
430
431 print $handle "<$full_docroot>\n"
432 }
433}
434
435sub output_general_xml_footer
436{
437 output_xml_footer(@_);
438}
439
440
441sub process {
442 my $self = shift (@_);
443 my ($doc_obj) = @_;
444
445 # for OAI purposes
446 $doc_obj->set_lastmodified();
447 $doc_obj->set_oailastmodified();
448
449 if ($self->{'group_size'} > 1) {
450 $self->group_process ($doc_obj);
451 return;
452 }
453
454 my $OID = $doc_obj->get_OID();
455 $OID = "NULL" unless defined $OID;
456
457 my $top_section = $doc_obj->get_top_section();
458
459 #get document's directory
460 my $doc_dir = $self->get_doc_dir ($OID, $doc_obj->get_source_filename());
461
462 my $output_info = $self->{'output_info'};
463 return if (!defined $output_info);
464
465 ##############################
466 # call subclass' saveas method
467 ##############################
468 $self->saveas($doc_obj,$doc_dir);
469 $self->archiveinf_db($doc_obj,$doc_dir);
470
471}
472
473sub store_output_info_reference {
474 my $self = shift (@_);
475 my ($doc_obj) = @_;
476
477 my $output_info = $self->{'output_info'};
478 my $metaname = $self->{'sortmeta'};
479 if (!defined $metaname || $metaname !~ /\S/) {
480 $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, "");
481 return;
482 }
483
484 my $metadata = "";
485 my $top_section = $doc_obj->get_top_section();
486
487 my @commameta_list = split(/,/, $metaname);
488 foreach my $cmn (@commameta_list) {
489 my $meta = $doc_obj->get_metadata_element($top_section, $cmn);
490 if ($meta) {
491 # do remove prefix/suffix - this will apply to all values
492 $meta =~ s/^$self->{'removeprefix'}// if defined $self->{'removeprefix'};
493 $meta =~ s/$self->{'removesuffix'}$// if defined $self->{'removesuffix'};
494 $meta = &sorttools::format_metadata_for_sorting($cmn, $meta, $doc_obj);
495 $metadata .= $meta if ($meta);
496 }
497 }
498
499 # store reference in the output_info
500 $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, $metadata);
501
502}
503
504sub group_process {
505
506 my $self = shift (@_);
507 my ($doc_obj) = @_;
508
509 my $OID = $doc_obj->get_OID();
510 $OID = "NULL" unless defined $OID;
511
512 my $groupsize = $self->{'group_size'};
513 my $gs_count = $self->{'gs_count'};
514 my $open_new_file = (($gs_count % $groupsize)==0);
515 my $outhandle = $self->{'output_handle'};
516
517 # opening a new file, or document has assoicated files => directory needed
518 if (($open_new_file) || (scalar(@{$doc_obj->get_assoc_files()})>0)) {
519
520 # The directory the archive file (doc.xml) and all associated files
521 # should end up in
522 my $doc_dir;
523 # If we've determined its time for a new file, open it now
524 if ($open_new_file || !defined($self->{'gs_doc_dir'}))
525 {
526 $doc_dir = $self->get_doc_dir ($OID, $doc_obj->get_source_filename());
527 # only if opening new file
528 my $output_dir = $self->get_output_dir();
529 &FileUtils::makeAllDirectories($output_dir) unless &FileUtils::directoryExists($output_dir);
530 my $doc_file = &FileUtils::filenameConcatenate($output_dir, $doc_dir, "doc.xml");
531 my $short_doc_file = &FileUtils::filenameConcatenate($doc_dir, "doc.xml");
532
533 if ($gs_count>0)
534 {
535 return if (!$self->close_file_output());
536 }
537
538 open (GROUPPROCESS, ">$doc_file") or (print $outhandle "BasePlugout::group_process could not write to file $doc_file\n" and return);
539
540
541 $self->{'gs_filename'} = $doc_file;
542 $self->{'short_doc_file'} = $short_doc_file;
543 $self->{'gs_OID'} = $OID;
544 $self->{'gs_doc_dir'} = $doc_dir;
545
546 $self->output_xml_header('BasePlugout::GROUPPROCESS','Archive');
547 }
548 # Otherwise load the same archive document directory used last time
549 else
550 {
551 $doc_dir = $self->{'gs_doc_dir'};
552 }
553
554 # copy all the associated files, add this information as metadata
555 # to the document
556 print $outhandle "Writing associated files to $doc_dir\n";
557 $self->process_assoc_files ($doc_obj, $doc_dir);
558
559 # look up 'gsdlmetafile' metadata and store that information
560 # explicitly in $doc_obj
561 $self->process_metafiles_metadata ($doc_obj);
562 }
563
564 # save this document
565 my $section_text = &docprint::get_section_xml($doc_obj,$doc_obj->get_top_section());
566 print GROUPPROCESS $section_text;
567
568 $self->{'gs_count'}++;
569}
570
571
572sub saveas {
573 my $self = shift (@_);
574
575 die "Basplug::saveas function must be implemented in sub classes\n";
576}
577
578sub get_doc_dir {
579 my $self = shift (@_);
580 my ($OID, $source_filename) = @_;
581
582 my $working_dir = $self->get_output_dir();
583 my $working_info = $self->{'output_info'};
584 return if (!defined $working_info);
585
586 my $doc_info = $working_info->get_info($OID);
587 my $doc_dir = '';
588
589 if (defined $doc_info && scalar(@$doc_info) >= 1)
590 {
591 # This OID already has an archives directory, so use it again
592 $doc_dir = $doc_info->[0];
593 $doc_dir =~ s/\/?((doc(mets)?)|(dublin_core))\.xml(\.gz)?$//;
594 }
595 elsif ($self->{'keep_import_structure'})
596 {
597 $source_filename = &File::Basename::dirname($source_filename);
598 $source_filename =~ s/[\\\/]+/\//g;
599 $source_filename =~ s/\/$//;
600
601 $doc_dir = substr($source_filename, length($ENV{'GSDLIMPORTDIR'}) + 1);
602 }
603
604 # We have to use a new archives directory for this document
605 if ($doc_dir eq "")
606 {
607 $doc_dir = $self->get_new_doc_dir ($working_info, $working_dir, $OID);
608 }
609
610 if (!defined $self->{'group'} || !$self->{'group'}){
611 &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir));
612 }
613
614 return $doc_dir;
615}
616
617## @function get_new_doc_dir()
618#
619# Once a doc object is ready to write to disk (and hence has a nice OID),
620# generate a unique subdirectory to write the information to.
621# - create the directory as part of this call, to try and avoid race conditions
622# found in parallel processing [jmt12]
623#
624sub get_new_doc_dir
625{
626 my $self = shift (@_);
627 my($working_info,$working_dir,$OID) = @_;
628
629 my $doc_dir = "";
630 my $doc_dir_rest = $OID;
631
632 # remove any \ and / from the OID
633 $doc_dir_rest =~ s/[\\\/]//g;
634
635 # Remove ":" if we are on Windows OS, as otherwise they get confused with the drive letters
636 if ($ENV{'GSDLOS'} =~ /^windows$/i)
637 {
638 $doc_dir_rest =~ s/\://g;
639 }
640
641 # we generally create a unique directory by adding consequtive fragments of
642 # the document identifier (split by some predefined length - defaulting to
643 # 8) until we find a directory that doesn't yet exist. Note that directories
644 # that contain a document have a suffix ".dir" (whereas those that contain
645 # only subdirectories have no suffix).
646 my $doc_dir_num = 0; # how many directories deep we are
647 my $created_directory = 0; # have we successfully created a new directory
648 do
649 {
650 # (does this work on windows? - jmt12)
651 if ($doc_dir_num > 0)
652 {
653 $doc_dir .= '/';
654 }
655 # the default matching pattern grabs the next 'subdir_split_length'
656 # characters of the OID to act as the next subdirectory
657 my $pattern = '^(.{1,' . $self->{'subdir_split_length'} . '})';
658 # Do we count any "HASH" prefix against the split length limit?
659 if ($self->{'subdir_hash_prefix'} && $doc_dir_num == 0)
660 {
661 $pattern = '^((HASH)?.{1,' . $self->{'subdir_split_length'} . '})';
662 }
663 # Note the use of 's' to both capture the next chuck of OID and to remove
664 # it from OID at the same time
665 if ($doc_dir_rest =~ s/$pattern//i)
666 {
667 $doc_dir .= $1;
668 $doc_dir_num++;
669 $created_directory = &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir'));
670 }
671 my $new_dir = &FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir');
672 ###rint STDERR "[DEBUG] BasePlugout::get_new_doc_dir(<working_info>, $working_dir, $oid)\n";
673 ###rint STDERR " - create directory: $new_dir => $created_directory\n";
674 ###rint STDERR " - rest: $doc_dir_rest\n";
675 ###rint STDERR " - working_info->size(): " . $working_info->size() . " [ < 1024 ?]\n";
676 ###rint STDERR " - doc_dir_num: " . $doc_dir_num . "\n";
677 }
678 while ($doc_dir_rest ne '' && ($created_directory == 0 || ($working_info->size() >= 1024 && $doc_dir_num < 2)));
679
680 # not unique yet? Add on an incremental suffix until we are unique
681 my $i = 1;
682 my $doc_dir_base = $doc_dir;
683 while ($created_directory == 0)
684 {
685 $doc_dir = $doc_dir_base . '-' . $i;
686 $created_directory = &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir'));
687 $i++;
688 }
689
690 # in theory this should never happen
691 if (!$created_directory)
692 {
693 die("Error! Failed to create directory for document: " . $doc_dir_base . "\n");
694 }
695
696 return $doc_dir . '.dir';
697}
698## get_new_doc_dir()
699
700
701sub process_assoc_files {
702 my $self = shift (@_);
703 my ($doc_obj, $doc_dir, $handle) = @_;
704
705 my $outhandle = $self->{'output_handle'};
706
707 my $output_dir = $self->get_output_dir();
708 return if (!defined $output_dir);
709
710 &FileUtils::makeAllDirectories($output_dir) unless &FileUtils::directoryExists($output_dir);
711
712 my $working_dir = &FileUtils::filenameConcatenate($output_dir, $doc_dir);
713 &FileUtils::makeAllDirectories($working_dir) unless &FileUtils::directoryExists($working_dir);
714
715 my @assoc_files = ();
716 my $filename;;
717
718 my $source_filename = $doc_obj->get_source_filename();
719
720 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
721
722 if (defined $collect_dir) {
723 my $dirsep_regexp = &util::get_os_dirsep();
724
725 if ($collect_dir !~ /$dirsep_regexp$/) {
726 $collect_dir .= &util::get_dirsep(); # ensure there is a slash at the end
727 }
728
729 # This test is never going to fail on Windows -- is this a problem?
730
731 if ($source_filename !~ /^$dirsep_regexp/) {
732 $source_filename = &FileUtils::filenameConcatenate($collect_dir, $source_filename);
733 }
734 }
735
736
737 # set the assocfile path (even if we have no assoc files - need this for lucene)
738 $doc_obj->set_utf8_metadata_element ($doc_obj->get_top_section(),
739 "assocfilepath",
740 "$doc_dir");
741 foreach my $assoc_file_rec (@{$doc_obj->get_assoc_files()}) {
742 my ($dir, $afile) = $assoc_file_rec->[1] =~ /^(.*?)([^\/\\]+)$/;
743 $dir = "" unless defined $dir;
744
745 my $utf8_real_filename = $assoc_file_rec->[0];
746
747 # for some reasons the image associate file has / before the full path
748 $utf8_real_filename =~ s/^\\(.*)/$1/i;
749
750## my $real_filename = &util::utf8_to_real_filename($utf8_real_filename);
751 my $real_filename = $utf8_real_filename;
752 $real_filename = &util::downgrade_if_dos_filename($real_filename);
753
754 if (&FileUtils::fileExists($real_filename)) {
755
756 $filename = &FileUtils::filenameConcatenate($working_dir, $afile);
757
758 &FileUtils::hardLink($real_filename, $filename, $self->{'verbosity'});
759
760 $doc_obj->add_utf8_metadata ($doc_obj->get_top_section(),
761 "gsdlassocfile",
762 "$afile:$assoc_file_rec->[2]:$dir");
763 } elsif ($self->{'verbosity'} > 1) {
764 print $outhandle "BasePlugout::process couldn't copy the associated file " .
765 "$real_filename to $afile\n";
766 }
767 }
768}
769
770
771sub process_metafiles_metadata
772{
773 my $self = shift (@_);
774 my ($doc_obj) = @_;
775
776 my $top_section = $doc_obj->get_top_section();
777 my $metafiles = $doc_obj->get_metadata($top_section,"gsdlmetafile");
778
779 foreach my $metafile_pair (@$metafiles) {
780 my ($full_metafile,$metafile) = split(/ : /,$metafile_pair);
781
782 $doc_obj->metadata_file($full_metafile,$metafile);
783 }
784
785 $doc_obj->delete_metadata($top_section,"gsdlmetafile");
786}
787
788sub archiveinf_files_to_field
789{
790 my $self = shift(@_);
791 my ($files,$field,$collect_dir,$oid_files,$reverse_lookups) = @_;
792
793 foreach my $file_rec (@$files) {
794 my $real_filename = (ref $file_rec eq "ARRAY") ? $file_rec->[0] : $file_rec;
795 my $full_file = (ref $file_rec eq "ARRAY") ? $file_rec->[1] : $file_rec;
796 # for some reasons the image associate file has / before the full path
797 $real_filename =~ s/^\\(.*)/$1/i;
798
799 my $raw_filename = &util::downgrade_if_dos_filename($real_filename);
800
801 if (&FileUtils::fileExists($raw_filename)) {
802
803# if (defined $collect_dir) {
804# my $collect_dir_re_safe = $collect_dir;
805# $collect_dir_re_safe =~ s/\\/\\\\/g; # use &util::filename_to_regex()
806# $collect_dir_re_safe =~ s/\./\\./g;##
807
808# $real_filename =~ s/^$collect_dir_re_safe//;
809# }
810
811 if (defined $reverse_lookups) {
812 $reverse_lookups->{$real_filename} = 1;
813 }
814### push(@{$oid_files->{$field}},$full_file);
815 push(@{$oid_files->{$field}},$raw_filename);
816 }
817 else {
818 print STDERR "Warning: archiveinf_files_to_field()\n $real_filename does not appear to be on the file system\n";
819 }
820 }
821}
822
823sub archiveinf_db
824{
825 my $self = shift (@_);
826 my ($doc_obj) = @_;
827
828 my $verbosity = $self->{'verbosity'};
829
830 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
831 if (defined $collect_dir) {
832 my $dirsep_regexp = &util::get_os_dirsep();
833
834 if ($collect_dir !~ /$dirsep_regexp$/) {
835 # ensure there is a slash at the end
836 $collect_dir .= &util::get_dirsep();
837 }
838 }
839
840 my $oid = $doc_obj->get_OID();
841 my $source_filename = $doc_obj->get_unmodified_source_filename();
842 my $working_info = $self->{'output_info'};
843 my $doc_info = $working_info->get_info($oid);
844
845 my ($doc_file,$index_status,$sortmeta) = @$doc_info;
846 # doc_file is the path to the archive doc.xml. Make sure it has unix
847 # slashes, then if the collection is copied to linux, it can be built without reimport
848 $doc_file =~ s/\\/\//g;
849 my $oid_files = { 'doc-file' => $doc_file,
850 'index-status' => $index_status,
851 'src-file' => $source_filename,
852 'sort-meta' => $sortmeta,
853 'assoc-file' => [],
854 'meta-file' => [] };
855
856 my $reverse_lookups = { $source_filename => "1" };
857
858
859 $self->archiveinf_files_to_field($doc_obj->get_source_assoc_files(),"assoc-file",
860 $collect_dir,$oid_files,$reverse_lookups);
861
862
863 $self->archiveinf_files_to_field($doc_obj->get_meta_files(),"meta-file",
864 $collect_dir,$oid_files);
865
866 # Get the infodbtype value for this collection from the arcinfo object
867 my $infodbtype = $self->{'output_info'}->{'infodbtype'};
868 my $output_dir = $self->{'output_dir'};
869
870 my $doc_db = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $output_dir);
871 my $src_db = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-src", $output_dir);
872
873 ##print STDERR "*** To set in db: \n\t$doc_db\n\t$oid\n\t$doc_db_text\n";
874
875 if (!$self->{'no_rss'})
876 {
877 if (($oid_files->{'index-status'} eq "I") || ($oid_files->{'index-status'} eq "R")) {
878 my $top_section = $doc_obj->get_top_section();
879
880 my $dc_titles = $doc_obj->get_metadata($top_section,"dc.Title");
881 my $dc_title = join("; ", @$dc_titles);
882
883 if ($oid_files->{'index-status'} eq "R") {
884 $dc_title .= " (Updated)";
885 }
886
887 my $rss_filename = &FileUtils::filenameConcatenate($output_dir,"rss-items.rdf");
888 my $rss_fh;
889 if (&FileUtils::openFileHandle($rss_filename, '>>', \$rss_fh)) {
890 print $rss_fh "<item>\n";
891 print $rss_fh " <title>$dc_title</title>\n";
892 print $rss_fh " <link>_httpdomain__httpcollection_/document/$oid</link>\n";
893 print $rss_fh "</item>\n";
894 &FileUtils::closeFileHandle($rss_filename, \$rss_fh);
895 }
896 else {
897 print STDERR "**** Failed to open $rss_filename\n$!\n";
898 }
899 }
900 }
901
902 $oid_files->{'doc-file'} = [ $oid_files->{'doc-file'} ];
903 $oid_files->{'index-status'} = [ $oid_files->{'index-status'} ];
904 $oid_files->{'src-file'} = [ $oid_files->{'src-file'} ];
905 $oid_files->{'sort-meta'} = [ $oid_files->{'sort-meta'} ];
906
907 my $infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $doc_db, "append");
908 &dbutil::write_infodb_entry($infodbtype, $infodb_file_handle, $oid, $oid_files);
909 &dbutil::close_infodb_write_handle($infodbtype, $infodb_file_handle);
910
911 foreach my $rl (keys %$reverse_lookups) {
912 $working_info->add_reverseinfo($rl,$oid);
913 }
914
915 # meta files not set in reverese entry, but need to set the metadata flag
916 if (defined $doc_obj->get_meta_files()) {
917 foreach my $meta_file_rec(@{$doc_obj->get_meta_files()}) {
918 my $full_file = (ref $meta_file_rec eq "ARRAY") ? $meta_file_rec->[0] : $meta_file_rec;
919 $working_info->set_meta_file_flag($full_file);
920 }
921 }
922}
923
924
925sub set_sortmeta {
926 my $self = shift (@_);
927 my ($sortmeta, $removeprefix, $removesuffix) = @_;
928
929 $self->{'sortmeta'} = $sortmeta;
930 if (defined ($removeprefix) && $removeprefix ) {
931 $removeprefix =~ s/^\^//; # don't need a leading ^
932 $self->{'removeprefix'} = $removeprefix;
933 }
934 if (defined ($removesuffix) && $removesuffix) {
935 $removesuffix =~ s/\$$//; # don't need a trailing $
936 $self->{'removesuffix'} = $removesuffix;
937 }
938}
939
940sub open_xslt_pipe
941{
942 my $self = shift @_;
943 my ($output_file_name, $xslt_file)=@_;
944
945 return unless defined $xslt_file and $xslt_file ne "" and &FileUtils::fileExists($xslt_file);
946
947 my $java_class_path = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","ApplyXSLT.jar");
948
949 my $mapping_file_path = "";
950
951 if ($ENV{'GSDLOS'} eq "windows"){
952 $java_class_path .=";".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
953 # this file:/// bit didn't work for me on windows XP
954 #$xslt_file = "\"file:///".$xslt_file."\"";
955 #$mapping_file_path = "\"file:///";
956 }
957 else{
958 $java_class_path .=":".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
959 }
960
961
962 $java_class_path = "\"".$java_class_path."\"";
963
964 my $cmd = "| java -cp $java_class_path org.nzdl.gsdl.ApplyXSLT -t \"$xslt_file\" ";
965
966 if (defined $self->{'mapping_file'} and $self->{'mapping_file'} ne ""){
967 my $mapping_file_path = "\"".$self->{'mapping_file'}."\"";
968 $cmd .= "-m $mapping_file_path";
969 }
970
971 open(*XMLWRITER, $cmd)
972 or die "can't open pipe to xslt: $!";
973
974
975 $self->{'xslt_writer'} = *XMLWRITER;
976
977 print XMLWRITER "<?DocStart?>\n";
978 print XMLWRITER "$output_file_name\n";
979
980
981 }
982
983
984sub close_xslt_pipe
985{
986 my $self = shift @_;
987
988
989 return unless defined $self->{'xslt_writer'} ;
990
991 my $xsltwriter = $self->{'xslt_writer'};
992
993 print $xsltwriter "<?DocEnd?>\n";
994 close($xsltwriter);
995
996 undef $self->{'xslt_writer'};
997
998}
999
1000sub close_file_output
1001{
1002 my ($self) = @_;
1003
1004 # make sure that the handle has been opened - it won't be if we failed
1005 # to import any documents...
1006 if (defined(fileno(GROUPPROCESS))) {
1007 $self->output_xml_footer('GROUPPROCESS','Archive');
1008 close GROUPPROCESS;
1009 }
1010
1011 my $OID = $self->{'gs_OID'};
1012 my $short_doc_file = $self->{'short_doc_file'};
1013
1014 if ($self->{'gzip'}) {
1015 my $doc_file = $self->{'gs_filename'};
1016 `gzip $doc_file`;
1017 $doc_file .= ".gz";
1018 $short_doc_file .= ".gz";
1019 if (!&FileUtils::fileExists($doc_file)) {
1020 my $outhandle = $self->{'output_handle'};
1021 print $outhandle "error while gzipping: $doc_file doesn't exist\n";
1022 return 0;
1023 }
1024 }
1025
1026 # store reference in output_info
1027 my $output_info = $self->{'output_info'};
1028 return 0 if (!defined $output_info);
1029 $output_info->add_info($OID, $short_doc_file, undef, undef);
1030 return 1;
1031}
1032
1033
1034#the subclass should implement this method if is_group method could return 1.
1035sub close_group_output{
1036 my $self = shift (@_);
1037}
1038
1039sub is_group {
1040 my $self = shift (@_);
1041 return 0;
1042}
1043
1044my $dc_set = { Title => 1,
1045 Creator => 1,
1046 Subject => 1,
1047 Description => 1,
1048 Publisher => 1,
1049 Contributor => 1,
1050 Date => 1,
1051 Type => 1,
1052 Format => 1,
1053 Identifier => 1,
1054 Source => 1,
1055 Language => 1,
1056 Relation => 1,
1057 Coverage => 1,
1058 Rights => 1};
1059
1060
1061# returns an XML representation of the dublin core metadata
1062# if dc meta is not found, try ex meta
1063# This method is not used by the DSpacePlugout, which has its
1064# own method to save its dc metadata
1065sub get_dc_metadata {
1066 my $self = shift(@_);
1067 my ($doc_obj, $section, $version) = @_;
1068
1069 # build up string of dublin core metadata
1070 $section="" unless defined $section;
1071
1072 my $section_ptr = $doc_obj->_lookup_section($section);
1073 return "" unless defined $section_ptr;
1074
1075
1076 my $explicit_dc = {};
1077 my $explicit_ex_dc = {};
1078 my $explicit_ex = {};
1079
1080 my $all_text="";
1081
1082 # We want high quality dc metadata to go in first, so we store all the
1083 # assigned dc.* values first. Then, for all those dc metadata names in
1084 # the official dc set that are as yet unassigned, we look to see whether
1085 # embedded ex.dc.* metadata has defined some values for them. If not,
1086 # then for the same missing dc metadata names, we look in ex metadata.
1087
1088 foreach my $data (@{$section_ptr->{'metadata'}}){
1089 my $escaped_value = &docprint::escape_text($data->[1]);
1090 if ($data->[0]=~ m/^dc\./) {
1091 $data->[0] =~ tr/[A-Z]/[a-z]/;
1092
1093 $data->[0] =~ m/^dc\.(.*)/;
1094 my $dc_element = $1;
1095
1096 if (!defined $explicit_dc->{$dc_element}) {
1097 $explicit_dc->{$dc_element} = [];
1098 }
1099 push(@{$explicit_dc->{$dc_element}},$escaped_value);
1100
1101 if (defined $version && ($version eq "oai_dc")) {
1102 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1103 }
1104 else {
1105 # qualifier???
1106 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1107 }
1108
1109 } elsif ($data->[0]=~ m/^ex\.dc\./) { # now look through ex.dc.* to fill in as yet unassigned fields in dc metaset
1110 $data->[0] =~ m/^ex\.dc\.(.*)/;
1111 my $ex_dc_element = $1;
1112 my $lc_ex_dc_element = lc($ex_dc_element);
1113
1114 # only store the ex.dc value for this dc metaname if no dc.* was assigned for it
1115 if (defined $dc_set->{$ex_dc_element}) {
1116 if (!defined $explicit_ex_dc->{$lc_ex_dc_element}) {
1117 $explicit_ex_dc->{$lc_ex_dc_element} = [];
1118 }
1119 push(@{$explicit_ex_dc->{$lc_ex_dc_element}},$escaped_value);
1120 }
1121 }
1122 elsif (($data->[0] =~ m/^ex\./) || ($data->[0] !~ m/\./)) { # look through ex. meta (incl. meta without prefix)
1123 $data->[0] =~ m/^(ex\.)?(.*)/;
1124 my $ex_element = $2;
1125 my $lc_ex_element = lc($ex_element);
1126
1127 if (defined $dc_set->{$ex_element}) {
1128 if (!defined $explicit_ex->{$lc_ex_element}) {
1129 $explicit_ex->{$lc_ex_element} = [];
1130 }
1131 push(@{$explicit_ex->{$lc_ex_element}},$escaped_value);
1132 }
1133 }
1134 }
1135
1136 # go through dc_set and for any element *not* defined in explicit_dc
1137 # that does exist in explicit_ex, add it in as metadata
1138 foreach my $k ( keys %$dc_set ) {
1139 my $lc_k = lc($k);
1140
1141 if (!defined $explicit_dc->{$lc_k}) {
1142 # try to find if ex.dc.* defines this dc.* meta,
1143 # if not, then look for whether there's an ex.* equivalent
1144
1145 if (defined $explicit_ex_dc->{$lc_k}) {
1146 foreach my $v (@{$explicit_ex_dc->{$lc_k}}) {
1147 my $dc_element = $lc_k;
1148 my $escaped_value = $v;
1149
1150 if (defined $version && ($version eq "oai_dc")) {
1151 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1152 }
1153 else {
1154 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1155 }
1156 }
1157 } elsif (defined $explicit_ex->{$lc_k}) {
1158 foreach my $v (@{$explicit_ex->{$lc_k}}) {
1159 my $dc_element = $lc_k;
1160 my $escaped_value = $v;
1161
1162 if (defined $version && ($version eq "oai_dc")) {
1163 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1164 }
1165 else {
1166 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1167 }
1168 }
1169 }
1170 }
1171 }
1172
1173 if ($all_text eq "") {
1174 $all_text .= " There is no Dublin Core metatdata in this document\n";
1175 }
1176 $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;
1177
1178 return $all_text;
1179}
1180
1181# Build up dublin_core metadata. Priority given to dc.* over ex.*
1182# This method was apparently added by Jeffrey and committed by Shaoqun.
1183# But we don't know why it was added, so not using it anymore.
1184sub new_get_dc_metadata {
1185
1186 my $self = shift(@_);
1187 my ($doc_obj, $section, $version) = @_;
1188
1189 # build up string of dublin core metadata
1190 $section="" unless defined $section;
1191
1192 my $section_ptr=$doc_obj->_lookup_section($section);
1193 return "" unless defined $section_ptr;
1194
1195 my $all_text = "";
1196 foreach my $data (@{$section_ptr->{'metadata'}}){
1197 my $escaped_value = &docprint::escape_text($data->[1]);
1198 my $dc_element = $data->[0];
1199
1200 my @array = split('\.',$dc_element);
1201 my ($type,$name);
1202
1203 if(defined $array[1])
1204 {
1205 $type = $array[0];
1206 $name = $array[1];
1207 }
1208 else
1209 {
1210 $type = "ex";
1211 $name = $array[0];
1212 }
1213
1214 $all_text .= ' <Metadata Type="'. $type.'" Name="'.$name.'">'. $escaped_value. "</Metadata>\n";
1215 }
1216 return $all_text;
1217}
1218
1219
12201;
Note: See TracBrowser for help on using the repository browser.