source: main/trunk/greenstone2/perllib/plugouts/BasePlugout.pm@ 32540

Last change on this file since 32540 was 32540, checked in by ak19, 6 years ago

Modified BasePlugout so that plugouts now take the site_name parameter, needed by GreenstoneSQLPlugout. Similar changes were made in the previous commit for plugins, where GreenstoneSQLPlugin needed to have access to site_name. In both cases, for GS2, the site_name (and hence SQL database name) is greenstone2.
  • 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_name",
48 'desc' => "{BasPlugout.site_name}",
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_name") {
152 $v = $strArg;
153 }
154 elsif($v eq "-site_name") {
155 $self->{'site_name'} = $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 return bless $self, $class;
194
195}
196
197# implement this in subclass if you want to do some initialization after
198# loading and setting parameters, and before processing the documents.
199sub begin {
200
201 my $self= shift (@_);
202
203}
204# implement in subclasses if it needs some non-group related cleanup (post-group cleanup
205# Like begin(), end() is also called by inexport.pm
206sub end {
207 my $self= shift (@_);
208
209}
210sub print_xml_usage
211{
212 my $self = shift(@_);
213 my $header = shift(@_);
214 my $high_level_information_only = shift(@_);
215
216 # XML output is always in UTF-8
217 gsprintf::output_strings_in_UTF8;
218
219 if ($header) {
220 &PrintUsage::print_xml_header("plugout");
221 }
222 $self->print_xml($high_level_information_only);
223}
224
225
226sub print_xml
227{
228 my $self = shift(@_);
229 my $high_level_information_only = shift(@_);
230
231 my $optionlistref = $self->{'option_list'};
232 my @optionlist = @$optionlistref;
233 my $plugoutoptions = shift(@$optionlistref);
234 return if (!defined($plugoutoptions));
235
236 gsprintf(STDERR, "<PlugoutInfo>\n");
237 gsprintf(STDERR, " <Name>$plugoutoptions->{'name'}</Name>\n");
238 my $desc = gsprintf::lookup_string($plugoutoptions->{'desc'});
239 $desc =~ s/</&amp;lt;/g; # doubly escaped
240 $desc =~ s/>/&amp;gt;/g;
241 gsprintf(STDERR, " <Desc>$desc</Desc>\n");
242 gsprintf(STDERR, " <Abstract>$plugoutoptions->{'abstract'}</Abstract>\n");
243 gsprintf(STDERR, " <Inherits>$plugoutoptions->{'inherits'}</Inherits>\n");
244 unless (defined($high_level_information_only)) {
245 gsprintf(STDERR, " <Arguments>\n");
246 if (defined($plugoutoptions->{'args'})) {
247 &PrintUsage::print_options_xml($plugoutoptions->{'args'});
248 }
249 gsprintf(STDERR, " </Arguments>\n");
250
251 # Recurse up the plugout hierarchy
252 $self->print_xml();
253 }
254 gsprintf(STDERR, "</PlugoutInfo>\n");
255}
256
257
258sub print_txt_usage
259{
260 my $self = shift(@_);
261
262 # Print the usage message for a plugout (recursively)
263 my $descoffset = $self->determine_description_offset(0);
264 $self->print_plugout_usage($descoffset, 1);
265}
266
267sub determine_description_offset
268{
269 my $self = shift(@_);
270 my $maxoffset = shift(@_);
271
272 my $optionlistref = $self->{'option_list'};
273 my @optionlist = @$optionlistref;
274 my $plugoutoptions = pop(@$optionlistref);
275 return $maxoffset if (!defined($plugoutoptions));
276
277 # Find the length of the longest option string of this download
278 my $plugoutargs = $plugoutoptions->{'args'};
279 if (defined($plugoutargs)) {
280 my $longest = &PrintUsage::find_longest_option_string($plugoutargs);
281 if ($longest > $maxoffset) {
282 $maxoffset = $longest;
283 }
284 }
285
286 # Recurse up the download hierarchy
287 $maxoffset = $self->determine_description_offset($maxoffset);
288 $self->{'option_list'} = \@optionlist;
289 return $maxoffset;
290}
291
292
293sub print_plugout_usage
294{
295 my $self = shift(@_);
296 my $descoffset = shift(@_);
297 my $isleafclass = shift(@_);
298
299 my $optionlistref = $self->{'option_list'};
300 my @optionlist = @$optionlistref;
301 my $plugoutoptions = shift(@$optionlistref);
302 return if (!defined($plugoutoptions));
303
304 my $plugoutname = $plugoutoptions->{'name'};
305 my $plugoutargs = $plugoutoptions->{'args'};
306 my $plugoutdesc = $plugoutoptions->{'desc'};
307
308 # Produce the usage information using the data structure above
309 if ($isleafclass) {
310 if (defined($plugoutdesc)) {
311 gsprintf(STDERR, "$plugoutdesc\n\n");
312 }
313 gsprintf(STDERR, " {common.usage}: plugout $plugoutname [{common.options}]\n\n");
314 }
315
316 # Display the download options, if there are some
317 if (defined($plugoutargs)) {
318 # Calculate the column offset of the option descriptions
319 my $optiondescoffset = $descoffset + 2; # 2 spaces between options & descriptions
320
321 if ($isleafclass) {
322 gsprintf(STDERR, " {common.specific_options}:\n");
323 }
324 else {
325 gsprintf(STDERR, " {common.general_options}:\n", $plugoutname);
326 }
327
328 # Display the download options
329 &PrintUsage::print_options_txt($plugoutargs, $optiondescoffset);
330 }
331
332 # Recurse up the download hierarchy
333 $self->print_plugout_usage($descoffset, 0);
334 $self->{'option_list'} = \@optionlist;
335}
336
337
338sub error
339{
340 my ($strFunctionName,$strError) = @_;
341 {
342 print "Error occoured in BasePlugout.pm\n".
343 "In Function: ".$strFunctionName."\n".
344 "Error Message: ".$strError."\n";
345 exit(-1);
346 }
347}
348
349# OIDtype may be "hash" or "hash_on_full_filename" or "incremental" or "filename" or "dirname" or "full_filename" or "assigned"
350sub set_OIDtype {
351 my $self = shift (@_);
352 my ($type, $metadata) = @_;
353
354 if ($type =~ /^(hash|hash_on_full_filename|incremental|filename|dirname|full_filename|assigned)$/) {
355 $self->{'OIDtype'} = $type;
356 } else {
357 $self->{'OIDtype'} = "hash";
358 }
359 if ($type =~ /^assigned$/) {
360 if (defined $metadata) {
361 $self->{'OIDmetadata'} = $metadata;
362 } else {
363 $self->{'OIDmetadata'} = "dc.Identifier";
364 }
365 }
366}
367
368sub set_output_dir
369{
370 my $self = shift @_;
371 my ($output_dir) = @_;
372
373 $self->{'output_dir'} = $output_dir;
374}
375
376sub setoutputdir
377{
378 my $self = shift @_;
379 my ($output_dir) = @_;
380
381 $self->{'output_dir'} = $output_dir;
382}
383
384sub get_output_dir
385{
386 my $self = shift (@_);
387
388 return $self->{'output_dir'};
389}
390
391sub getoutputdir
392{
393 my $self = shift (@_);
394
395 return $self->{'output_dir'};
396}
397
398sub getoutputinfo
399{
400 my $self = shift (@_);
401
402 return $self->{'output_info'};
403}
404
405
406sub get_output_handler
407{
408 my $self = shift (@_);
409
410 my ($output_file_name) = @_;
411
412 my $fh;
413 &FileUtils::openFileHandle($output_file_name, '>', \$fh) or die('Can not open a file handler for: ' . $output_file_name . "\n");
414
415 return $fh;
416}
417
418sub release_output_handler
419{
420 my $self = shift (@_);
421 my ($outhandler) = @_;
422
423 close($outhandler);
424
425}
426
427sub output_xml_header {
428 my $self = shift (@_);
429 my ($handle,$docroot,$nondoctype) = @_;
430
431
432 #print $handle '<?xml version="1.0" encoding="UTF-8" standalone="no"?>' . "\n";
433
434 #For Dspace must be UTF in lower case
435 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
436
437 if (!defined $nondoctype){
438 my $doctype = (defined $docroot) ? $docroot : "Section";
439
440 # Used to be '<!DOCTYPE Archive SYSTEM ...'
441
442 print $handle "<!DOCTYPE $doctype SYSTEM \"http://greenstone.org/dtd/Archive/1.0/Archive.dtd\">\n";
443 }
444
445 print $handle "<$docroot>\n" if defined $docroot;
446}
447
448sub output_xml_footer {
449 my $self = shift (@_);
450 my ($handle,$docroot) = @_;
451 print $handle "</$docroot>\n" if defined $docroot;
452}
453
454
455sub output_general_xml_header
456{
457 my $self = shift (@_);
458 my ($handle,$docroot,$opt_attributes,$opt_dtd, $opt_doctype) = @_;
459
460 print $handle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
461
462 if (defined $opt_dtd) {
463 my $doctype = (defined $opt_doctype) ? $opt_doctype : $docroot;
464 print $handle "<!DOCTYPE $doctype SYSTEM \"$opt_dtd\">\n";
465 }
466
467 if (defined $docroot) {
468 my $full_docroot = $docroot;
469 if (defined $opt_attributes) {
470 $full_docroot .= " $opt_attributes";
471 }
472
473 print $handle "<$full_docroot>\n"
474 }
475}
476
477sub output_general_xml_footer
478{
479 output_xml_footer(@_);
480}
481
482# This is called by the plugins after read_into_doc_obj generates the doc_obj.
483sub process {
484 my $self = shift (@_);
485 my ($doc_obj) = @_;
486
487 my $output_info = $self->{'output_info'};
488 return if (!defined $output_info);
489
490 # for OAI purposes
491 $doc_obj->set_lastmodified();
492 $doc_obj->set_oailastmodified();
493
494 # find out which directory to save to
495 my $doc_dir = "";
496 if ($self->is_group()) {
497 $doc_dir = $self->get_group_doc_dir($doc_obj);
498 } else {
499 $doc_dir = $self->get_doc_dir($doc_obj);
500 }
501
502 ##############################
503 # call subclass' saveas method
504 ##############################
505 $self->saveas($doc_obj,$doc_dir);
506
507 # write out data to archiveinf-doc.db
508 if ($self->{'generate_databases'}) {
509 $self->archiveinf_db($doc_obj);
510 }
511 if ($self->is_group()) {
512 $self->{'gs_count'}++; # do we want this for all cases?
513 $self->{'group_position'}++;
514 }
515}
516
517sub store_output_info_reference {
518 my $self = shift (@_);
519 my ($doc_obj) = @_;
520
521 my $output_info = $self->{'output_info'};
522 my $metaname = $self->{'sortmeta'};
523
524 my $group_position;
525 if ($self->is_group()) {
526 $group_position = $self->{'group_position'};
527 }
528 if (!defined $metaname || $metaname !~ /\S/) {
529 my $OID = $doc_obj->get_OID();
530 $output_info->add_info($OID,$self->{'short_doc_file'}, undef, "", $group_position);
531 return;
532 }
533
534 if ($metaname eq "OID") { # sort by OID
535 my $OID = $doc_obj->get_OID();
536 $output_info->add_info($OID,$self->{'short_doc_file'}, undef, $OID, undef);
537 return;
538 }
539
540 my $metadata = "";
541 my $top_section = $doc_obj->get_top_section();
542
543 my @commameta_list = split(/,/, $metaname);
544 foreach my $cmn (@commameta_list) {
545 my $meta = $doc_obj->get_metadata_element($top_section, $cmn);
546 if ($meta) {
547 # do remove prefix/suffix - this will apply to all values
548 $meta =~ s/^$self->{'removeprefix'}// if defined $self->{'removeprefix'};
549 $meta =~ s/$self->{'removesuffix'}$// if defined $self->{'removesuffix'};
550 $meta = &sorttools::format_metadata_for_sorting($cmn, $meta, $doc_obj);
551 $metadata .= $meta if ($meta);
552 }
553 }
554
555 # store reference in the output_info
556 $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, $metadata,undef);
557
558}
559
560
561
562sub saveas {
563 my $self = shift (@_);
564 my ($doc_obj, $doc_dir) = @_;
565
566 die "Basplug::saveas function must be implemented in sub classes\n";
567}
568
569sub get_group_doc_dir {
570 my $self = shift (@_);
571 my ($doc_obj) = @_;
572
573 my $outhandle = $self->{'output_handle'};
574 my $OID = $doc_obj->get_OID();
575 $OID = "NULL" unless defined $OID;
576
577 my $groupsize = $self->{'group_size'};
578 my $gs_count = $self->{'gs_count'};
579 my $open_new_file = (($gs_count % $groupsize)==0);
580
581 my $doc_dir;
582
583 if (!$open_new_file && scalar(@{$doc_obj->get_assoc_files()})>0) {
584 # if we have some assoc files, then we will need to start a new file
585 if ($self->{'verbosity'} > 2) {
586 print $outhandle " Starting a archives folder for $OID as it has associated files\n";
587 }
588 $open_new_file = 1;
589 }
590
591 # opening a new file
592 if (($open_new_file) || !defined($self->{'gs_doc_dir'})) {
593 # first we close off the old output
594 if ($gs_count>0)
595 {
596 return if (!$self->close_group_output());
597 }
598
599 # this will create the directory
600 $doc_dir = $self->get_doc_dir ($doc_obj);
601 $self->{'new_doc_dir'} = 1;
602 $self->{'gs_doc_dir'} = $doc_dir;
603 $self->{'group_position'} = 1;
604 }
605 else {
606 $doc_dir = $self->{'gs_doc_dir'};
607 $self->{'new_doc_dir'} = 0;
608 }
609 return $doc_dir;
610
611}
612sub get_doc_dir {
613
614 my $self = shift (@_);
615 my ($doc_obj) = @_;
616
617 my $OID = $doc_obj->get_OID();
618 $OID = "NULL" unless defined $OID;
619
620 my $working_dir = $self->get_output_dir();
621 my $working_info = $self->{'output_info'};
622 return if (!defined $working_info);
623
624 my $doc_info = $working_info->get_info($OID);
625 my $doc_dir = '';
626
627 if (defined $doc_info && scalar(@$doc_info) >= 1)
628 {
629 # This OID already has an archives directory, so use it again
630 $doc_dir = $doc_info->[0];
631 $doc_dir =~ s/\/?((doc(mets)?)|(dublin_core))\.xml(\.gz)?$//;
632 }
633 elsif ($self->{'keep_import_structure'})
634 {
635 my $source_filename = $doc_obj->get_source_filename();
636 $source_filename = &File::Basename::dirname($source_filename);
637 $source_filename =~ s/[\\\/]+/\//g;
638 $source_filename =~ s/\/$//;
639
640 $doc_dir = substr($source_filename, length($ENV{'GSDLIMPORTDIR'}) + 1);
641 }
642
643 # We have to use a new archives directory for this document
644 if ($doc_dir eq "")
645 {
646 $doc_dir = $self->get_new_doc_dir ($working_info, $working_dir, $OID);
647 }
648
649 &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir));
650
651 return $doc_dir;
652}
653
654
655## @function get_new_doc_dir()
656#
657# Once a doc object is ready to write to disk (and hence has a nice OID),
658# generate a unique subdirectory to write the information to.
659#
660# - create the directory as part of this call, to try and avoid race conditions
661# found in parallel processing [jmt12]
662#
663# @todo figure out what the rule regarding $work_info->size() is meant to do
664#
665# @todo determine what $self->{'group'} is, and whether it should affect
666# directory creation
667#
668sub get_new_doc_dir
669{
670 my $self = shift (@_);
671 my($working_info,$working_dir,$OID) = @_;
672
673 my $doc_dir = "";
674 my $doc_dir_rest = $OID;
675
676 # remove any \ and / from the OID
677 $doc_dir_rest =~ s/[\\\/]//g;
678
679 # Remove ":" if we are on Windows OS, as otherwise they get confused with the drive letters
680 if ($ENV{'GSDLOS'} =~ /^windows$/i)
681 {
682 $doc_dir_rest =~ s/\://g;
683 }
684
685 # we generally create a unique directory by adding consequtive fragments of
686 # the document identifier (split by some predefined length - defaulting to
687 # 8) until we find a directory that doesn't yet exist. Note that directories
688 # that contain a document have a suffix ".dir" (whereas those that contain
689 # only subdirectories have no suffix).
690 my $doc_dir_num = 0; # how many directories deep we are
691 my $created_directory = 0; # have we successfully created a new directory
692 do
693 {
694 # (does this work on windows? - jmt12)
695 if ($doc_dir_num > 0)
696 {
697 $doc_dir .= '/';
698 }
699
700 # the default matching pattern grabs the next 'subdir_split_length'
701 # characters of the OID to act as the next subdirectory
702 my $pattern = '^(.{1,' . $self->{'subdir_split_length'} . '})';
703
704 # Do we count any "HASH" prefix against the split length limit?
705 if ($self->{'subdir_hash_prefix'} && $doc_dir_num == 0)
706 {
707 $pattern = '^((HASH)?.{1,' . $self->{'subdir_split_length'} . '})';
708 }
709
710 # Note the use of 's' to both capture the next chuck of OID and to remove
711 # it from OID at the same time
712 if ($doc_dir_rest =~ s/$pattern//i)
713 {
714 $doc_dir .= $1;
715 $doc_dir_num++;
716
717 my $full_doc_dir = &FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir');
718 if(!FileUtils::directoryExists($full_doc_dir))
719 {
720 &FileUtils::makeAllDirectories($full_doc_dir);
721 $created_directory = 1;
722 }
723
724 ###rint STDERR "[DEBUG] BasePlugout::get_new_doc_dir(<working_info>, $working_dir, $oid)\n";
725 ###rint STDERR " - create directory: $full_doc_dir => $created_directory\n";
726 ###rint STDERR " - rest: $doc_dir_rest\n";
727 ###rint STDERR " - working_info->size(): " . $working_info->size() . " [ < 1024 ?]\n";
728 ###rint STDERR " - doc_dir_num: " . $doc_dir_num . "\n";
729 }
730 }
731 while ($doc_dir_rest ne '' && ($created_directory == 0 || ($working_info->size() >= 1024 && $doc_dir_num < 2)));
732
733 # not unique yet? Add on an incremental suffix until we are unique
734 my $i = 1;
735 my $doc_dir_base = $doc_dir;
736 while ($created_directory == 0)
737 {
738 $doc_dir = $doc_dir_base . '-' . $i;
739 $created_directory = &FileUtils::makeAllDirectories(&FileUtils::filenameConcatenate($working_dir, $doc_dir . '.dir'));
740 $i++;
741 }
742
743 # in theory this should never happen
744 if (!$created_directory)
745 {
746 die("Error! Failed to create directory for document: " . $doc_dir_base . "\n");
747 }
748
749 return $doc_dir . '.dir';
750}
751## get_new_doc_dir()
752
753
754sub process_assoc_files {
755 my $self = shift (@_);
756 my ($doc_obj, $doc_dir, $handle) = @_;
757
758 my $outhandle = $self->{'output_handle'};
759
760 my $output_dir = $self->get_output_dir();
761 return if (!defined $output_dir);
762
763 &FileUtils::makeAllDirectories($output_dir) unless &FileUtils::directoryExists($output_dir);
764
765 my $working_dir = &FileUtils::filenameConcatenate($output_dir, $doc_dir);
766 &FileUtils::makeAllDirectories($working_dir) unless &FileUtils::directoryExists($working_dir);
767
768 my @assoc_files = ();
769 my $filename;;
770
771 my $source_filename = $doc_obj->get_source_filename();
772
773 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
774
775 if (defined $collect_dir) {
776 my $dirsep_regexp = &util::get_os_dirsep();
777
778 if ($collect_dir !~ /$dirsep_regexp$/) {
779 $collect_dir .= &util::get_dirsep(); # ensure there is a slash at the end
780 }
781
782 # This test is never going to fail on Windows -- is this a problem?
783
784 if ($source_filename !~ /^$dirsep_regexp/) {
785 $source_filename = &FileUtils::filenameConcatenate($collect_dir, $source_filename);
786 }
787 }
788
789
790 # set the assocfile path (even if we have no assoc files - need this for lucene)
791 $doc_obj->set_utf8_metadata_element ($doc_obj->get_top_section(),
792 "assocfilepath",
793 "$doc_dir");
794 foreach my $assoc_file_rec (@{$doc_obj->get_assoc_files()}) {
795 my ($dir, $afile) = $assoc_file_rec->[1] =~ /^(.*?)([^\/\\]+)$/;
796 $dir = "" unless defined $dir;
797
798 my $utf8_real_filename = $assoc_file_rec->[0];
799
800 # for some reasons the image associate file has / before the full path
801 $utf8_real_filename =~ s/^\\(.*)/$1/i;
802
803## my $real_filename = &util::utf8_to_real_filename($utf8_real_filename);
804 my $real_filename = $utf8_real_filename;
805 $real_filename = &util::downgrade_if_dos_filename($real_filename);
806
807 if (&FileUtils::fileExists($real_filename)) {
808
809 $filename = &FileUtils::filenameConcatenate($working_dir, $afile);
810
811 &FileUtils::hardLink($real_filename, $filename, $self->{'verbosity'});
812
813 $doc_obj->add_utf8_metadata ($doc_obj->get_top_section(),
814 "gsdlassocfile",
815 "$afile:$assoc_file_rec->[2]:$dir");
816 } elsif ($self->{'verbosity'} > 1) {
817 print $outhandle "BasePlugout::process couldn't copy the associated file " .
818 "$real_filename to $afile\n";
819 }
820 }
821}
822
823
824sub process_metafiles_metadata
825{
826 my $self = shift (@_);
827 my ($doc_obj) = @_;
828
829 my $top_section = $doc_obj->get_top_section();
830 my $metafiles = $doc_obj->get_metadata($top_section,"gsdlmetafile");
831
832 foreach my $metafile_pair (@$metafiles) {
833 my ($full_metafile,$metafile) = split(/ : /,$metafile_pair);
834
835 $doc_obj->metadata_file($full_metafile,$metafile);
836 }
837
838 $doc_obj->delete_metadata($top_section,"gsdlmetafile");
839}
840
841sub archiveinf_files_to_field
842{
843 my $self = shift(@_);
844 my ($files,$field,$collect_dir,$oid_files,$reverse_lookups) = @_;
845
846 foreach my $file_rec (@$files) {
847 my $real_filename = (ref $file_rec eq "ARRAY") ? $file_rec->[0] : $file_rec;
848 my $full_file = (ref $file_rec eq "ARRAY") ? $file_rec->[1] : $file_rec;
849 # for some reasons the image associate file has / before the full path
850 $real_filename =~ s/^\\(.*)/$1/i;
851
852 my $raw_filename = &util::downgrade_if_dos_filename($real_filename);
853
854 if (&FileUtils::fileExists($raw_filename)) {
855
856# if (defined $collect_dir) {
857# my $collect_dir_re_safe = $collect_dir;
858# $collect_dir_re_safe =~ s/\\/\\\\/g; # use &util::filename_to_regex()
859# $collect_dir_re_safe =~ s/\./\\./g;##
860
861# $real_filename =~ s/^$collect_dir_re_safe//;
862# }
863
864 if (defined $reverse_lookups) {
865 $reverse_lookups->{$real_filename} = 1;
866 }
867
868 if($field =~ m@assoc-file|src-file|meta-file@) {
869 $raw_filename = &util::abspath_to_placeholders($raw_filename);
870 }
871
872### push(@{$oid_files->{$field}},$full_file);
873 push(@{$oid_files->{$field}},$raw_filename);
874 }
875 else {
876 print STDERR "Warning: archiveinf_files_to_field()\n $real_filename does not appear to be on the file system\n";
877 }
878 }
879}
880
881sub archiveinf_db
882{
883 my $self = shift (@_);
884 my ($doc_obj) = @_;
885
886 my $verbosity = $self->{'verbosity'};
887
888 my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
889 if (defined $collect_dir) {
890 my $dirsep_regexp = &util::get_os_dirsep();
891
892 if ($collect_dir !~ /$dirsep_regexp$/) {
893 # ensure there is a slash at the end
894 $collect_dir .= &util::get_dirsep();
895 }
896 }
897
898 my $oid = $doc_obj->get_OID();
899 my $source_filename = $doc_obj->get_unmodified_source_filename();
900 my $working_info = $self->{'output_info'};
901 my $doc_info = $working_info->get_info($oid);
902
903 my ($doc_file,$index_status,$sortmeta, $group_position) = @$doc_info;
904 # doc_file is the path to the archive doc.xml. Make sure it has unix
905 # slashes, then if the collection is copied to linux, it can be built without reimport
906 $doc_file =~ s/\\/\//g;
907 my $oid_files = { 'doc-file' => $doc_file,
908 'index-status' => $index_status,
909 'src-file' => $source_filename,
910 'sort-meta' => $sortmeta,
911 'assoc-file' => [],
912 'meta-file' => [] };
913 if (defined $group_position) {
914 $oid_files->{'group-position'} = $group_position;
915 }
916 my $reverse_lookups = { $source_filename => "1" };
917
918
919 $self->archiveinf_files_to_field($doc_obj->get_source_assoc_files(),"assoc-file",
920 $collect_dir,$oid_files,$reverse_lookups);
921
922
923 $self->archiveinf_files_to_field($doc_obj->get_meta_files(),"meta-file",
924 $collect_dir,$oid_files);
925
926 # Get the infodbtype value for this collection from the arcinfo object
927 my $infodbtype = $self->{'output_info'}->{'infodbtype'};
928 my $output_dir = $self->{'output_dir'};
929
930 my $doc_db = &dbutil::get_infodb_file_path($infodbtype, "archiveinf-doc", $output_dir);
931
932 ##print STDERR "*** To set in db: \n\t$doc_db\n\t$oid\n\t$doc_db_text\n";
933
934 if (!$self->{'no_rss'})
935 {
936 if (($oid_files->{'index-status'} eq "I") || ($oid_files->{'index-status'} eq "R")) {
937 my $top_section = $doc_obj->get_top_section();
938
939 # rss_title can be set in collect.cfg as follows:
940 # plugout GreenstoneXMLPlugout -rss_title "dc.Title; ex.Title"
941 # rss_title is a semi-colon or comma-separated list of the metadata field names that should
942 # be consulted in order to obtain a Title (anchor text) for the RSS document link.
943 # If not specified, rss_title will default to dc.Title, and fall back on Untitled
944 my $metafieldnames = $self->{'rss_title'};
945 my @metafieldarray = split(/[,;] ?/,$metafieldnames); # , or ; separator can be followed by an optional space
946 my $titles;
947 #@$titles=(); # at worst @$titles will be (), as get_metadata(dc.Titles) may return ()
948 foreach my $metafieldname (@metafieldarray) {
949 $metafieldname =~ s@^ex\.@@; # if ex.Title, need to get_metadata() on metafieldname=Title
950 $titles = $doc_obj->get_metadata($top_section,$metafieldname);
951
952 if(scalar(@$titles) != 0) { # found at least one title for one metafieldname
953 last; # break out of the loop
954 }
955 }
956
957 # if ex.Title was listed in the metafieldnames, then we'll surely have a value for title for this doc
958 # otherwise, if we have no titles at this point, add in a default of Untitled as this doc's title
959 if(scalar(@$titles) == 0) { #&& $metafieldnames !~ [email protected]@) {
960 push(@$titles, "Untitled");
961 }
962
963 # encode basic html entities like <>"& in the title(s), since the & char can break RSS links
964 for (my $i = 0; $i < scalar(@$titles); $i++) {
965 &ghtml::htmlsafe(@$titles[$i]);
966 }
967
968 my $dc_title = join("; ", @$titles);
969
970 if ($oid_files->{'index-status'} eq "R") {
971 $dc_title .= " (Updated)";
972 }
973
974 my $rss_entry = "<item>\n";
975 $rss_entry .= " <title>$dc_title</title>\n";
976 if(&util::is_gs3()) {
977 $rss_entry .= " <link>_httpdomain__httpcollection_/document/$oid</link>\n";
978 } else {
979 $rss_entry .= " <link>_httpdomainHtmlsafe__httpcollection_/document/$oid</link>\n";
980 }
981 $rss_entry .= "</item>";
982
983 if (defined(&dbutil::supportsRSS) && &dbutil::supportsRSS($infodbtype))
984 {
985 my $rss_db = &dbutil::get_infodb_file_path($infodbtype, 'rss-items', $output_dir);
986 my $rss_db_fh = &dbutil::open_infodb_write_handle($infodbtype, $rss_db, 'append');
987 &dbutil::write_infodb_rawentry($infodbtype, $rss_db_fh, $oid, $rss_entry);
988 &dbutil::close_infodb_write_handle($infodbtype, $rss_db_fh);
989 }
990 else
991 {
992 my $rss_filename = &FileUtils::filenameConcatenate($output_dir,"rss-items.rdf");
993 my $rss_fh;
994 if (&FileUtils::openFileHandle($rss_filename, '>>', \$rss_fh, "utf8"))
995 {
996 print $rss_fh $rss_entry . "\n";
997 &FileUtils::closeFileHandle($rss_filename, \$rss_fh);
998 }
999 else
1000 {
1001 print STDERR "**** Failed to open $rss_filename\n$!\n";
1002 }
1003 }
1004 }
1005 }
1006
1007 $oid_files->{'doc-file'} = [ $oid_files->{'doc-file'} ];
1008 $oid_files->{'index-status'} = [ $oid_files->{'index-status'} ];
1009 $oid_files->{'src-file'} = &util::abspath_to_placeholders($oid_files->{'src-file'});
1010 $oid_files->{'src-file'} = [ $oid_files->{'src-file'} ];
1011 $oid_files->{'sort-meta'} = [ $oid_files->{'sort-meta'} ];
1012 if (defined $oid_files->{'group-position'}) {
1013 $oid_files->{'group-position'} = [ $oid_files->{'group-position'} ];
1014 }
1015
1016 my $infodb_file_handle = &dbutil::open_infodb_write_handle($infodbtype, $doc_db, "append");
1017 &dbutil::write_infodb_entry($infodbtype, $infodb_file_handle, $oid, $oid_files);
1018 &dbutil::close_infodb_write_handle($infodbtype, $infodb_file_handle);
1019
1020 foreach my $rl (keys %$reverse_lookups) {
1021 $working_info->add_reverseinfo($rl,$oid);
1022 }
1023
1024 # meta files not set in reverse entry, but need to set the metadata flag
1025 if (defined $doc_obj->get_meta_files()) {
1026 foreach my $meta_file_rec(@{$doc_obj->get_meta_files()}) {
1027 my $full_file = (ref $meta_file_rec eq "ARRAY") ? $meta_file_rec->[0] : $meta_file_rec;
1028 $working_info->set_meta_file_flag($full_file);
1029 }
1030 }
1031}
1032
1033# This sub is called for every metadata.xml accepted for processing by by MetdataXMLPlugin
1034# and adds an entry into archiveinf-src.db for that file in the form:
1035# [@THISCOLLECTPATH@/import/metadata.xml]
1036# <meta-file>1
1037# This prevents blind reprocessing of the same old docs upon *incremental* building whenever
1038# we encounter a default empty metadata.xml that has no actual <FileSet> content defined.
1039sub add_metaxml_file_entry_to_archiveinfsrc {
1040 my $self = shift (@_);
1041 my ($full_file) = @_;
1042
1043 print STDERR "**** Adding metaxml file entry for full_file: $full_file\n";
1044 my $working_info = $self->{'output_info'};
1045 $working_info->set_meta_file_flag($full_file);
1046}
1047
1048
1049sub set_sortmeta {
1050 my $self = shift (@_);
1051 my ($sortmeta, $removeprefix, $removesuffix) = @_;
1052
1053 $self->{'sortmeta'} = $sortmeta;
1054 if (defined ($removeprefix) && $removeprefix ) {
1055 $removeprefix =~ s/^\^//; # don't need a leading ^
1056 $self->{'removeprefix'} = $removeprefix;
1057 }
1058 if (defined ($removesuffix) && $removesuffix) {
1059 $removesuffix =~ s/\$$//; # don't need a trailing $
1060 $self->{'removesuffix'} = $removesuffix;
1061 }
1062}
1063
1064
1065
1066sub open_xslt_pipe
1067{
1068 my $self = shift @_;
1069 my ($output_file_name, $xslt_file)=@_;
1070
1071 return unless defined $xslt_file and $xslt_file ne "" and &FileUtils::fileExists($xslt_file);
1072
1073 my $java_class_path = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","ApplyXSLT.jar");
1074
1075 my $mapping_file_path = "";
1076
1077 if ($ENV{'GSDLOS'} eq "windows"){
1078 $java_class_path .=";".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
1079 # this file:/// bit didn't work for me on windows XP
1080 #$xslt_file = "\"file:///".$xslt_file."\"";
1081 #$mapping_file_path = "\"file:///";
1082 }
1083 else{
1084 $java_class_path .=":".&FileUtils::filenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
1085 }
1086
1087
1088 $java_class_path = "\"".$java_class_path."\"";
1089
1090 my $cmd = "| java -cp $java_class_path org.nzdl.gsdl.ApplyXSLT -t \"$xslt_file\" ";
1091
1092 if (defined $self->{'mapping_file'} and $self->{'mapping_file'} ne ""){
1093 my $mapping_file_path = "\"".$self->{'mapping_file'}."\"";
1094 $cmd .= "-m $mapping_file_path";
1095 }
1096
1097 open(*XMLWRITER, $cmd)
1098 or die "can't open pipe to xslt: $!";
1099
1100
1101 $self->{'xslt_writer'} = *XMLWRITER;
1102
1103 print XMLWRITER "<?DocStart?>\n";
1104 print XMLWRITER "$output_file_name\n";
1105
1106
1107 }
1108
1109
1110sub close_xslt_pipe
1111{
1112 my $self = shift @_;
1113
1114
1115 return unless defined $self->{'xslt_writer'} ;
1116
1117 my $xsltwriter = $self->{'xslt_writer'};
1118
1119 print $xsltwriter "<?DocEnd?>\n";
1120 close($xsltwriter);
1121
1122 undef $self->{'xslt_writer'};
1123
1124}
1125
1126
1127
1128#the subclass should implement this method if is_group method could return 1.
1129sub close_group_output{
1130 my $self = shift (@_);
1131}
1132
1133sub is_group {
1134 my $self = shift (@_);
1135 return 0;
1136}
1137
1138my $dc_set = { Title => 1,
1139 Creator => 1,
1140 Subject => 1,
1141 Description => 1,
1142 Publisher => 1,
1143 Contributor => 1,
1144 Date => 1,
1145 Type => 1,
1146 Format => 1,
1147 Identifier => 1,
1148 Source => 1,
1149 Language => 1,
1150 Relation => 1,
1151 Coverage => 1,
1152 Rights => 1};
1153
1154
1155# returns an XML representation of the dublin core metadata
1156# if dc meta is not found, try ex meta
1157# This method is not used by the DSpacePlugout, which has its
1158# own method to save its dc metadata
1159sub get_dc_metadata {
1160 my $self = shift(@_);
1161 my ($doc_obj, $section, $version) = @_;
1162
1163 # build up string of dublin core metadata
1164 $section="" unless defined $section;
1165
1166 my $section_ptr = $doc_obj->_lookup_section($section);
1167 return "" unless defined $section_ptr;
1168
1169
1170 my $explicit_dc = {};
1171 my $explicit_ex_dc = {};
1172 my $explicit_ex = {};
1173
1174 my $all_text="";
1175
1176 # We want high quality dc metadata to go in first, so we store all the
1177 # assigned dc.* values first. Then, for all those dc metadata names in
1178 # the official dc set that are as yet unassigned, we look to see whether
1179 # embedded ex.dc.* metadata has defined some values for them. If not,
1180 # then for the same missing dc metadata names, we look in ex metadata.
1181
1182 foreach my $data (@{$section_ptr->{'metadata'}}){
1183 my $escaped_value = &docprint::escape_text($data->[1]);
1184 if ($data->[0]=~ m/^dc\./) {
1185 $data->[0] =~ tr/[A-Z]/[a-z]/;
1186
1187 $data->[0] =~ m/^dc\.(.*)/;
1188 my $dc_element = $1;
1189
1190 if (!defined $explicit_dc->{$dc_element}) {
1191 $explicit_dc->{$dc_element} = [];
1192 }
1193 push(@{$explicit_dc->{$dc_element}},$escaped_value);
1194
1195 if (defined $version && ($version eq "oai_dc")) {
1196 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1197 }
1198 else {
1199 # qualifier???
1200 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1201 }
1202
1203 } elsif ($data->[0]=~ m/^ex\.dc\./) { # now look through ex.dc.* to fill in as yet unassigned fields in dc metaset
1204 $data->[0] =~ m/^ex\.dc\.(.*)/;
1205 my $ex_dc_element = $1;
1206 my $lc_ex_dc_element = lc($ex_dc_element);
1207
1208 # only store the ex.dc value for this dc metaname if no dc.* was assigned for it
1209 if (defined $dc_set->{$ex_dc_element}) {
1210 if (!defined $explicit_ex_dc->{$lc_ex_dc_element}) {
1211 $explicit_ex_dc->{$lc_ex_dc_element} = [];
1212 }
1213 push(@{$explicit_ex_dc->{$lc_ex_dc_element}},$escaped_value);
1214 }
1215 }
1216 elsif (($data->[0] =~ m/^ex\./) || ($data->[0] !~ m/\./)) { # look through ex. meta (incl. meta without prefix)
1217 $data->[0] =~ m/^(ex\.)?(.*)/;
1218 my $ex_element = $2;
1219 my $lc_ex_element = lc($ex_element);
1220
1221 if (defined $dc_set->{$ex_element}) {
1222 if (!defined $explicit_ex->{$lc_ex_element}) {
1223 $explicit_ex->{$lc_ex_element} = [];
1224 }
1225 push(@{$explicit_ex->{$lc_ex_element}},$escaped_value);
1226 }
1227 }
1228 }
1229
1230 # go through dc_set and for any element *not* defined in explicit_dc
1231 # that does exist in explicit_ex, add it in as metadata
1232 foreach my $k ( keys %$dc_set ) {
1233 my $lc_k = lc($k);
1234
1235 if (!defined $explicit_dc->{$lc_k}) {
1236 # try to find if ex.dc.* defines this dc.* meta,
1237 # if not, then look for whether there's an ex.* equivalent
1238
1239 if (defined $explicit_ex_dc->{$lc_k}) {
1240 foreach my $v (@{$explicit_ex_dc->{$lc_k}}) {
1241 my $dc_element = $lc_k;
1242 my $escaped_value = $v;
1243
1244 if (defined $version && ($version eq "oai_dc")) {
1245 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1246 }
1247 else {
1248 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1249 }
1250 }
1251 } elsif (defined $explicit_ex->{$lc_k}) {
1252 foreach my $v (@{$explicit_ex->{$lc_k}}) {
1253 my $dc_element = $lc_k;
1254 my $escaped_value = $v;
1255
1256 if (defined $version && ($version eq "oai_dc")) {
1257 $all_text .= " <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
1258 }
1259 else {
1260 $all_text .= ' <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
1261 }
1262 }
1263 }
1264 }
1265 }
1266
1267 if ($all_text eq "") {
1268 $all_text .= " There is no Dublin Core metatdata in this document\n";
1269 }
1270 $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;
1271
1272 return $all_text;
1273}
1274
1275# Build up dublin_core metadata. Priority given to dc.* over ex.*
1276# This method was apparently added by Jeffrey and committed by Shaoqun.
1277# But we don't know why it was added, so not using it anymore.
1278sub new_get_dc_metadata {
1279
1280 my $self = shift(@_);
1281 my ($doc_obj, $section, $version) = @_;
1282
1283 # build up string of dublin core metadata
1284 $section="" unless defined $section;
1285
1286 my $section_ptr=$doc_obj->_lookup_section($section);
1287 return "" unless defined $section_ptr;
1288
1289 my $all_text = "";
1290 foreach my $data (@{$section_ptr->{'metadata'}}){
1291 my $escaped_value = &docprint::escape_text($data->[1]);
1292 my $dc_element = $data->[0];
1293
1294 my @array = split('\.',$dc_element);
1295 my ($type,$name);
1296
1297 if(defined $array[1])
1298 {
1299 $type = $array[0];
1300 $name = $array[1];
1301 }
1302 else
1303 {
1304 $type = "ex";
1305 $name = $array[0];
1306 }
1307
1308 $all_text .= ' <Metadata Type="'. $type.'" Name="'.$name.'">'. $escaped_value. "</Metadata>\n";
1309 }
1310 return $all_text;
1311}
1312
1313
13141;
Note: See TracBrowser for help on using the repository browser.