source: main/trunk/greenstone2/perllib/plugins/HathiTrustMETSPlugin.pm@ 32783

Last change on this file since 32783 was 32783, checked in by kjdon, 5 years ago

adding missing strings and tidying up some mislabelling
  • Property svn:executable set to *
File size: 19.6 KB
Line 
1###########################################################################
2#
3# HathiTrustMETSPlugin.pm -- plugin for sets of HathiTrust METS OCR'd
4# text that make up a document
5#
6# A component of the Greenstone digital library software
7# from the New Zealand Digital Library Project at the
8# University of Waikato, New Zealand.
9#
10# Copyright (C) 1999 New Zealand Digital Library Project
11#
12# This program is free software; you can redistribute it and/or modify
13# it under the terms of the GNU General Public License as published by
14# the Free Software Foundation; either version 2 of the License, or
15# (at your option) any later version.
16#
17# This program is distributed in the hope that it will be useful,
18# but WITHOUT ANY WARRANTY; without even the implied warranty of
19# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20# GNU General Public License for more details.
21#
22# You should have received a copy of the GNU General Public License
23# along with this program; if not, write to the Free Software
24# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
25#
26###########################################################################
27
28# HathiTrustMETSPlugin
29# processes HathiTrust METS files that are accompanied with page-by-page
30# OCR'd txt files
31#
32# All the supplemetary text files should be in a subfolder of the same
33# name as the METS file
34#
35# As usual, running
36# 'perl -S pluginfo.pl HathiTrustMETSPlugin' will list all the options.
37
38
39package HathiTrustMETSPlugin;
40
41use Encode;
42use ReadXMLFile;
43use ReadTextFile;
44# We don't currently work with the scanned image from HathiTrust METS
45# but leave it in for future proofing
46use ImageConverter;
47use MetadataRead;
48
49use JSON;
50
51use strict;
52no strict 'refs'; # allow filehandles to be variables and viceversa
53
54sub BEGIN {
55 @HathiTrustMETSPlugin::ISA = ('MetadataRead', 'ReadXMLFile', 'ReadTextFile', 'ImageConverter'
56 );
57}
58
59# One day HathiTrust might give more than page structure
60my $gs2_type_list =
61 [
62# { 'name' => "auto",
63# 'desc' => "{PagedImagePlugin.documenttype.auto2}" },
64# { 'name' => "paged",
65# 'desc' => "{PagedImagePlugin.documenttype.paged2}" },
66 { 'name' => "hierarchy",
67 'desc' => "{PagedImagePlugin.documenttype.hierarchy}" }
68 ];
69
70my $gs3_type_list =
71 [
72# { 'name' => "auto",
73# 'desc' => "{PagedImagePlugin.documenttype.auto3}" },
74# { 'name' => "paged",
75# 'desc' => "{PagedImagePlugin.documenttype.paged3}" },
76 { 'name' => "hierarchy",
77 'desc' => "{PagedImagePlugin.documenttype.hierarchy}" }
78# { 'name' => "pagedhierarchy",
79# 'desc' => "{PagedImagePlugin.documenttype.pagedhierarchy}" }
80 ];
81
82my $arguments =
83 [ { 'name' => "process_exp",
84 'desc' => "{BaseImporter.process_exp}",
85 'type' => "string",
86 'deft' => &get_default_process_exp(),
87 'reqd' => "no" },
88 { 'name' => "title_sub",
89 'desc' => "{HTMLPlugin.title_sub}",
90 'type' => "string",
91 'deft' => "" },
92 { 'name' => "headerpage",
93 'desc' => "{HathiTrustMETSPlugin.headerpage}",
94 'type' => "flag",
95 'reqd' => "no" },
96 {'name' => "processing_tmp_files",
97 'desc' => "{BaseImporter.processing_tmp_files}",
98 'type' => "flag",
99 'hiddengli' => "yes"}
100 ];
101
102my $doc_type_opt = { 'name' => "documenttype",
103 'desc' => "{PagedImagePlugin.documenttype}",
104 'type' => "enum",
105 'deft' => "hierarchy",
106 'reqd' => "no" };
107
108my $options = { 'name' => "HathiTrustMETSPlugin",
109 'desc' => "{HathiTrustMETSPlugin.desc}",
110 'abstract' => "no",
111 'inherits' => "yes",
112 'args' => $arguments };
113
114sub new {
115 my ($class) = shift (@_);
116 my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
117 push(@$pluginlist, $class);
118
119 push(@{$hashArgOptLists->{"OptList"}},$options);
120
121 my $imc_self = new ImageConverter($pluginlist, $inputargs, $hashArgOptLists);
122
123 # we can use this plugin to check gs3 version
124 if ($imc_self->{'gs_version'} eq "3") {
125 $doc_type_opt->{'list'} = $gs3_type_list;
126 }
127 else {
128 $doc_type_opt->{'list'} = $gs2_type_list;
129 }
130 push(@$arguments,$doc_type_opt);
131 # now we add the args to the list for parsing
132 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
133
134 my $rtf_self = new ReadTextFile($pluginlist, $inputargs, $hashArgOptLists, 1);
135 my $rxf_self = new ReadXMLFile($pluginlist, $inputargs, $hashArgOptLists);
136
137 my $self = BaseImporter::merge_inheritance($imc_self,$rtf_self,$rxf_self);
138
139 # Update $self used by XML::Parser so it finds callback functions
140 # such as start_document here and not in ReadXMLFile (which is what
141 # $self was when new XML::Parser was done)
142 #
143 # If the $self returned by this constructor is the same as the one
144 # used in ReadXMLFile (e.g. in the GreenstoneXMLPlugin) then this step isn't necessary
145 #
146 # Consider embedding this type of assignment into merge_inheritance
147 # to help catch all cases?
148
149 $rxf_self->{'parser'}->{'PluginObj'} = $self;
150
151 return bless $self, $class;
152}
153
154
155sub init {
156 my $self = shift (@_);
157 my ($verbosity, $outhandle, $failhandle) = @_;
158
159 $self->SUPER::init(@_);
160 $self->ImageConverter::init();
161}
162
163sub begin {
164 my $self = shift (@_);
165 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
166
167 $self->SUPER::begin(@_);
168 $self->ImageConverter::begin(@_);
169}
170
171sub get_default_process_exp {
172 my $self = shift (@_);
173
174 return q^\.mets.xml$^;
175}
176
177sub get_doctype {
178 my $self = shift(@_);
179
180 return "METS:mets";
181}
182
183
184# want to use BaseImporter's version of this, not ReadXMLFile's
185sub can_process_this_file {
186 my $self = shift(@_);
187 return $self->BaseImporter::can_process_this_file(@_);
188}
189
190# instead of a block exp, now we scan the file and record all text and img files mentioned there for blocking.
191sub store_block_files
192{
193 my $self = shift (@_);
194 my ($filename_full_path, $block_hash) = @_;
195
196 # do we need to do this?
197 # does BOM interfere just with XML parsing? In that case don't need it here
198 # if we do it here, we are modifying the file before we have worked out if
199 # its new or not, so it will always be reimported.
200 #$self->tidy_item_file($filename_full_path);
201
202 my ($dir, $file) = $filename_full_path =~ /^(.*?)([^\/\\]*)$/;
203
204 # do something
205 $self->scan_xml_for_files_to_block($filename_full_path, $dir, $block_hash);
206
207}
208
209# we want to use BaseImporter's read, not ReadXMLFile's
210sub read
211{
212 my $self = shift (@_);
213
214 $self->BaseImporter::read(@_);
215}
216
217
218
219sub read_into_doc_obj {
220 my $self = shift (@_);
221 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
222 my $outhandle = $self->{'outhandle'};
223 my $verbosity = $self->{'verbosity'};
224
225 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
226
227 print $outhandle "HathiTrustMETSPlugin processing \"$filename_full_path\"\n"
228 if $verbosity > 1;
229 print STDERR "<Processing n='$file' p='HathiTrustMETSPlugin'>\n" if ($gli);
230
231## $self->{'MaxImageWidth'} = 0;
232## $self->{'MaxImageHeight'} = 0;
233
234
235 ##$self->tidy_item_file($filename_full_path);
236
237 # careful checking needed here!! are we using local xml handlers or super ones
238 $self->ReadXMLFile::read($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli);
239 my $doc_obj = $self->{'doc_obj'};
240
241
242 my $section = $doc_obj->get_top_section();
243
244 $doc_obj->add_utf8_metadata($section, "Plugin", "$self->{'plugin_type'}");
245 $doc_obj->add_metadata($section, "FileFormat", "HathiTrustMETS");
246
247 # include any metadata passed in from previous plugins
248 # note that this metadata is associated with the top level section
249 $self->add_associated_files($doc_obj, $filename_full_path);
250 $self->extra_metadata ($doc_obj, $section, $metadata);
251 $self->auto_extract_metadata ($doc_obj);
252 $self->plugin_specific_process($base_dir, $file, $doc_obj, $gli);
253 # if we haven't found any Title so far, assign one
254 $self->title_fallback($doc_obj,$section,$filename_no_path);
255
256 $self->add_OID($doc_obj);
257 return (1,$doc_obj);
258}
259
260
261sub parse_aux_json_metadata {
262 my $self = shift(@_);
263 my ($base_dir, $file, $doc_obj, $gli) = @_;
264
265 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
266
267 my $topsection = $doc_obj->get_top_section();
268
269 my $json_metadata_filename = $filename_full_path;
270 $json_metadata_filename =~ s/\.mets.xml$/.json/;
271
272 my $json_text = "";
273 $self->ReadTextFile::read_file($json_metadata_filename,"utf8",undef,\$json_text);
274
275 my $json_rec = decode_json $json_text;
276 my $records = $json_rec->{'records'};
277 my @keys = keys %{$records};
278
279 my $key = shift @keys; # there should only be one
280 my $record = $records->{$key};
281
282 my @md_fields = ( "recordURL", "titles", "isbns", "issns", "oclcs", "lccns", "publishDates" );
283
284 foreach my $md_field (@md_fields) {
285 my $value_array = $record->{$md_field};
286
287 my $md_name = $md_field;
288 $md_name =~ s/s$//;
289
290 foreach my $md_value (@$value_array) {
291
292 if ($md_name eq "title") {
293 $doc_obj->set_utf8_metadata_element ($topsection, "Title", $md_value);
294 $doc_obj->set_utf8_metadata_element ($topsection, "dc.Title", $md_value);
295 }
296 else {
297 $doc_obj->set_utf8_metadata_element ($topsection, $md_name, $md_value);
298 }
299 }
300 }
301
302 my $htid = $json_rec->{'items'}->[0]->{'htid'};
303 my $docName = $htid;
304 my $docNameIE = $htid;
305 $docNameIE =~ s/^.*?\.//;
306
307 $doc_obj->set_utf8_metadata_element ($topsection, "docName", $docName);
308 $doc_obj->set_utf8_metadata_element ($topsection, "docNameIE", $docNameIE);
309
310}
311
312
313# override this for an inheriting plugin to add extra metadata etc
314sub plugin_specific_process {
315 my $self = shift(@_);
316 my ($base_dir, $file, $doc_obj, $gli) = @_;
317
318 $self->parse_aux_json_metadata($base_dir,$file,$doc_obj,$gli);
319}
320
321# sub tidy_item_file {
322# ... see PagedImagePlugin
323# }
324
325# sub rotate_image {
326# ... see PagedImagePlugin
327# }
328
329# sub process_image {
330# ... see PagedImagePlugin
331# }
332
333
334
335sub xml_start_tag {
336 my $self = shift(@_);
337 my ($expat, $element) = @_;
338 $self->{'element'} = $element;
339
340 my $doc_obj = $self->{'doc_obj'};
341 if ($element eq "METS:mets") {
342 $self->{'current_section'} = $doc_obj->get_top_section();
343# } elsif ($element eq "PageGroup" || $element eq "Page") {
344## if ($element eq "PageGroup") {
345## $self->{'has_internal_structure'} = 1;
346 }
347 elsif (($element eq "METS:FLocat") && ($_{'xlink:href'} =~ m/\.txt$/)) {
348 # e.g. <METS:FLocat LOCTYPE="OTHER" OTHERLOCTYPE="SYSTEM" xlink:href="00000000.txt"/>
349
350 # create a new section as a child
351 $self->{'current_section'} = $doc_obj->insert_section($doc_obj->get_end_child($self->{'current_section'}));
352 $self->{'num_pages'}++;
353 # assign pagenum as ... what?? => use page sequence number
354 my $txtfile = $_{'xlink:href'};
355 my ($pagenum) = ($txtfile =~ m/^(\d+)/);
356
357 if (defined $pagenum) {
358 my $pagenum_int = int($pagenum);
359 $doc_obj->set_utf8_metadata_element($self->{'current_section'}, "Title", "Page $pagenum_int");
360 }
361## my ($imgfile) = $_{'imgfile'};
362## if (defined $imgfile) {
363## # *****
364## # What about support for rotate image (e.g. old ':r' notation)?
365## $self->process_image($self->{'xml_file_dir'}.$imgfile, $imgfile, $doc_obj, $self->{'current_section'});
366## }
367
368## my ($txtfile) = $_{'txtfile'};
369 if (defined($txtfile)&& $txtfile ne "") {
370 my $full_txt_filename = &FileUtils::filenameConcatenate($self->{'xml_file_dir'},$txtfile);
371 $self->process_text ($full_txt_filename, $txtfile, $doc_obj, $self->{'current_section'});
372 } else {
373 $self->add_dummy_text($doc_obj, $self->{'current_section'});
374 }
375 }
376## elsif ($element eq "Metadata") {
377## $self->{'metadata_name'} = $_{'name'};
378## }
379}
380
381sub xml_end_tag {
382 my $self = shift(@_);
383 my ($expat, $element) = @_;
384
385 my $doc_obj = $self->{'doc_obj'};
386## if ($element eq "Page" || $element eq "PageGroup") {
387 if (($element eq "METS:FLocat") && ($_{'xlink:href'} =~ m/\.txt$/)) {
388 # if Title hasn't been assigned, set PageNum as Title
389 if (!defined $doc_obj->get_metadata_element ($self->{'current_section'}, "Title") && defined $doc_obj->get_metadata_element ($self->{'current_section'}, "PageNum" )) {
390 $doc_obj->add_utf8_metadata ($self->{'current_section'}, "Title", $doc_obj->get_metadata_element ($self->{'current_section'}, "PageNum" ));
391 }
392 # move the current section back to the parent
393 $self->{'current_section'} = $doc_obj->get_parent_section($self->{'current_section'});
394 } elsif ($element eq "Metadata") {
395
396 # text read in by XML::Parser is in Perl's binary byte value
397 # form ... need to explicitly make it UTF-8
398 my $meta_name = decode("utf-8",$self->{'metadata_name'});
399 my $metadata_value = decode("utf-8",$self->{'metadata_value'});
400
401 if ($meta_name =~ /\./) {
402 $meta_name = "ex.$meta_name";
403 }
404
405 $doc_obj->add_utf8_metadata ($self->{'current_section'}, $meta_name, $metadata_value);
406 $self->{'metadata_name'} = "";
407 $self->{'metadata_value'} = "";
408
409 }
410 # otherwise we ignore the end tag
411}
412
413
414sub xml_text {
415 my $self = shift(@_);
416 my ($expat) = @_;
417
418 if ($self->{'element'} eq "Metadata" && $self->{'metadata_name'}) {
419 $self->{'metadata_value'} .= $_;
420 }
421}
422
423sub xml_doctype {
424}
425
426sub open_document {
427 my $self = shift(@_);
428
429 # create a new document
430 $self->{'doc_obj'} = new doc ($self->{'filename'}, "indexed_doc", $self->{'file_rename_method'});
431 # TODO is file filenmae_no_path??
432 $self->set_initial_doc_fields($self->{'doc_obj'}, $self->{'filename'}, $self->{'processor'}, $self->{'metadata'});
433
434## my ($dir, $file) = $self->{'filename'} =~ /^(.*?)([^\/\\]*)$/;
435 my ($dir, $file_ext) = $self->{'filename'} =~ /^(.*?)(\.mets\.xml)$/;
436
437 $self->{'xml_file_dir'} = $dir;
438 $self->{'num_pages'} = 0;
439## $self->{'has_internal_structure'} = 0;
440
441}
442
443sub close_document {
444 my $self = shift(@_);
445 my $doc_obj = $self->{'doc_obj'};
446
447 my $topsection = $doc_obj->get_top_section();
448
449 # add numpages metadata
450 $doc_obj->set_utf8_metadata_element ($topsection, 'NumPages', $self->{'num_pages'}); # ##### !!!!
451
452 # set the document type
453 my $final_doc_type = "";
454## if ($self->{'documenttype'} eq "auto") {
455### if ($self->{'has_internal_structure'}) {
456### if ($self->{'gs_version'} eq "3") {
457### $final_doc_type = "pagedhierarchy";
458### }
459### else {
460### $final_doc_type = "hierarchy";
461### }
462### } else {
463### $final_doc_type = "paged";
464### }
465### } else {
466## # set to what doc type option was set to
467## $final_doc_type = $self->{'documenttype'};
468## }
469# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", $final_doc_type); # #### !!!!!
470 ### capiatalisation????
471# if ($self->{'documenttype'} eq 'paged') {
472 # set the gsdlthistype metadata to Paged - this ensures this document will
473 # be treated as a Paged doc, even if Titles are not numeric
474# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", "Paged");
475# } else {
476# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", "Hierarchy");
477# }
478
479## $doc_obj->set_utf8_metadata_element($topsection,"MaxImageWidth",$self->{'MaxImageWidth'});
480## $doc_obj->set_utf8_metadata_element($topsection,"MaxImageHeight",$self->{'MaxImageHeight'});
481## $self->{'MaxImageWidth'} = undef;
482## $self->{'MaxImageHeight'} = undef;
483
484}
485
486
487sub set_initial_doc_fields {
488 my $self = shift(@_);
489 my ($doc_obj, $filename_full_path, $processor, $metadata) = @_;
490
491 my $topsection = $doc_obj->get_top_section();
492
493 my $plugin_filename_encoding = $self->{'filename_encoding'};
494 my $filename_encoding = $self->deduce_filename_encoding($filename_full_path,$metadata,$plugin_filename_encoding);
495 $self->set_Source_metadata($doc_obj, $filename_full_path, $filename_encoding);
496
497 # if we want a header page, we need to add some text into the top section, otherwise this section will become invisible
498 if ($self->{'headerpage'}) {
499 $self->add_dummy_text($doc_obj, $topsection);
500 }
501}
502
503sub scan_xml_for_files_to_block
504{
505 my $self = shift (@_);
506 my ($filename_full_path, $dir, $block_hash) = @_;
507
508 my ($file_root) = ($filename_full_path =~ m/^(.*)\.mets\.xml$/);
509
510 $self->block_raw_filename($block_hash,"$file_root.zip");
511 $self->block_raw_filename($block_hash,"$file_root.json");
512
513 my $page_dir = $file_root;
514
515 open (METSFILE, $filename_full_path) || die "couldn't open $filename_full_path to work out which files to block\n";
516 my $line = "";
517 while (defined ($line = <METSFILE>)) {
518 next unless $line =~ /\w/;
519
520 # Exaple of what we are looking for
521 # <METS:FLocat LOCTYPE="OTHER" OTHERLOCTYPE="SYSTEM" xlink:href="00000000.txt"/>
522
523 if ($line =~ /xlink:href=\"([^\"]+)\"/) {
524 my $txt_filename = &FileUtils::filenameConcatenate($page_dir,$1);
525 my $topics_filename = $txt_filename . ".topics";
526 $self->block_raw_filename($block_hash,$txt_filename);
527 $self->block_raw_filename($block_hash,$topics_filename);
528 }
529 }
530 close METSFILE;
531
532}
533
534
535sub process_text {
536 my $self = shift (@_);
537 my ($filename_full_path, $file, $doc_obj, $cursection) = @_;
538
539 # check that the text file exists!!
540 if (!-f $filename_full_path) {
541 print "HathiTrustMETSPlugin: ERROR: File $filename_full_path does not exist, skipping\n";
542 return 0;
543 }
544
545 # remember that this text file was one of our source files, but only
546 # if we are not processing a tmp file
547 if (!$self->{'processing_tmp_files'} ) {
548 $doc_obj->associate_source_file($filename_full_path);
549 }
550 # Do encoding stuff
551 my ($language, $encoding) = $self->textcat_get_language_encoding ($filename_full_path);
552
553 my $text="";
554 if ( -s $filename_full_path > 0 ) {
555 &ReadTextFile::read_file($self, $filename_full_path, $encoding, $language, \$text); # already decoded as utf8
556 }
557
558# HathiTrust often has empty files
559## if (!length ($text)) {
560## # It's a bit unusual but not out of the question to have no text, so just give a warning
561## print "HathiTrustMETSPlugin: WARNING: $filename_full_path contains no text\n";
562## }
563
564 # we need to escape the escape character, or else mg will convert into
565 # eg literal newlines, instead of leaving the text as '\n'
566 $text =~ s/\\/\\\\/g; # macro language
567 $text =~ s/_/\\_/g; # macro language
568
569
570 if ($text =~ m/<html.*?>\s*<head.*?>.*<\/head>\s*<body.*?>(.*)<\/body>\s*<\/html>\s*$/is) {
571 # looks like HTML input
572 # no need to escape < and > or put in <pre> tags
573
574 $text = $1;
575
576 # add text to document object
577 $doc_obj->add_utf8_text($cursection, "$text");
578 }
579 else {
580 $text =~ s/</&lt;/g;
581 $text =~ s/>/&gt;/g;
582
583 # insert preformat tags and add text to document object
584 $doc_obj->add_utf8_text($cursection, "<pre>\n$text\n</pre>");
585 }
586
587 my $topics_filename = $filename_full_path . ".topics";
588 if ( -s $topics_filename > 0 ) {
589
590 my $topics_text = "";
591 $self->ReadTextFile::read_file($topics_filename,"utf8",undef,\$topics_text);
592
593 my @topics_array = split(/\|/,$topics_text);
594 foreach my $topic (@topics_array) {
595 if ($topic ne "") {
596 $doc_obj->set_utf8_metadata_element ($cursection, "concept", $topic);
597 }
598 }
599 }
600
601 return 1;
602}
603
604
605sub clean_up_after_doc_obj_processing {
606 my $self = shift(@_);
607
608 $self->ImageConverter::clean_up_temporary_files();
609}
610
6111;
Note: See TracBrowser for help on using the repository browser.