source: main/trunk/greenstone2/perllib/plugins/HathiTrustMETSPlugin.pm@ 31284

Last change on this file since 31284 was 31284, checked in by davidb, 4 years ago

Initial cut at plugin for processing HathiTrust METS files

  • Property svn:executable set to *
File size: 19.7 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' => "{BasePlugin.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' => "documenttype",
97# 'desc' => "{HathiTrustMETSPlugin.documenttype}",
98# 'type' => "enum",
99# 'list' => $type_list,
100# 'deft' => "auto",
101# 'reqd' => "no" },
102 {'name' => "processing_tmp_files",
103 'desc' => "{BasePlugin.processing_tmp_files}",
104 'type' => "flag",
105 'hiddengli' => "yes"}
106 ];
107
108my $doc_type_opt = { 'name' => "documenttype",
109 'desc' => "{HathiTrustMETSPlugin.documenttype}",
110 'type' => "enum",
111 'deft' => "auto",
112 'reqd' => "no" };
113
114my $options = { 'name' => "HathiTrustMETSPlugin",
115 'desc' => "{HathiTrustMETSPlugin.desc}",
116 'abstract' => "no",
117 'inherits' => "yes",
118 'args' => $arguments };
119
120sub new {
121 my ($class) = shift (@_);
122 my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
123 push(@$pluginlist, $class);
124
125 push(@{$hashArgOptLists->{"OptList"}},$options);
126
127 my $imc_self = new ImageConverter($pluginlist, $inputargs, $hashArgOptLists);
128
129 # we can use this plugin to check gs3 version
130 if ($imc_self->{'gs_version'} eq "3") {
131 $doc_type_opt->{'list'} = $gs3_type_list;
132 }
133 else {
134 $doc_type_opt->{'list'} = $gs2_type_list;
135 }
136 push(@$arguments,$doc_type_opt);
137 # now we add the args to the list for parsing
138 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
139
140 my $rtf_self = new ReadTextFile($pluginlist, $inputargs, $hashArgOptLists, 1);
141 my $rxf_self = new ReadXMLFile($pluginlist, $inputargs, $hashArgOptLists);
142
143 my $self = BasePlugin::merge_inheritance($imc_self,$rtf_self,$rxf_self);
144
145 # Update $self used by XML::Parser so it finds callback functions
146 # such as start_document here and not in ReadXMLFile (which is what
147 # $self was when new XML::Parser was done)
148 #
149 # If the $self returned by this constructor is the same as the one
150 # used in ReadXMLFile (e.g. in the GreenstoneXMLPlugin) then this step isn't necessary
151 #
152 # Consider embedding this type of assignment into merge_inheritance
153 # to help catch all cases?
154
155 $rxf_self->{'parser'}->{'PluginObj'} = $self;
156
157 return bless $self, $class;
158}
159
160
161sub init {
162 my $self = shift (@_);
163 my ($verbosity, $outhandle, $failhandle) = @_;
164
165 $self->SUPER::init(@_);
166 $self->ImageConverter::init();
167}
168
169sub begin {
170 my $self = shift (@_);
171 my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
172
173 $self->SUPER::begin(@_);
174 $self->ImageConverter::begin(@_);
175}
176
177sub get_default_process_exp {
178 my $self = shift (@_);
179
180 return q^\.mets.xml$^;
181}
182
183sub get_doctype {
184 my $self = shift(@_);
185
186 return "METS:mets";
187}
188
189
190# want to use BasePlugin's version of this, not ReadXMLFile's
191sub can_process_this_file {
192 my $self = shift(@_);
193 return $self->BasePlugin::can_process_this_file(@_);
194}
195
196# instead of a block exp, now we scan the file and record all text and img files mentioned there for blocking.
197sub store_block_files
198{
199 my $self = shift (@_);
200 my ($filename_full_path, $block_hash) = @_;
201
202 # do we need to do this?
203 # does BOM interfere just with XML parsing? In that case don't need it here
204 # if we do it here, we are modifying the file before we have worked out if
205 # its new or not, so it will always be reimported.
206 #$self->tidy_item_file($filename_full_path);
207
208 my ($dir, $file) = $filename_full_path =~ /^(.*?)([^\/\\]*)$/;
209
210 # do something
211 $self->scan_xml_for_files_to_block($filename_full_path, $dir, $block_hash);
212
213}
214
215# we want to use BasePlugin's read, not ReadXMLFile's
216sub read
217{
218 my $self = shift (@_);
219
220 $self->BasePlugin::read(@_);
221}
222
223
224
225sub read_into_doc_obj {
226 my $self = shift (@_);
227 my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
228 my $outhandle = $self->{'outhandle'};
229 my $verbosity = $self->{'verbosity'};
230
231 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
232
233 print $outhandle "HathiTrustMETSPlugin processing \"$filename_full_path\"\n"
234 if $verbosity > 1;
235 print STDERR "<Processing n='$file' p='HathiTrustMETSPlugin'>\n" if ($gli);
236
237## $self->{'MaxImageWidth'} = 0;
238## $self->{'MaxImageHeight'} = 0;
239
240
241 ##$self->tidy_item_file($filename_full_path);
242
243 # careful checking needed here!! are we using local xml handlers or super ones
244 $self->ReadXMLFile::read($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli);
245 my $doc_obj = $self->{'doc_obj'};
246
247
248 my $section = $doc_obj->get_top_section();
249
250 $doc_obj->add_utf8_metadata($section, "Plugin", "$self->{'plugin_type'}");
251 $doc_obj->add_metadata($section, "FileFormat", "HathiTrustMETS");
252
253 # include any metadata passed in from previous plugins
254 # note that this metadata is associated with the top level section
255 $self->add_associated_files($doc_obj, $filename_full_path);
256 $self->extra_metadata ($doc_obj, $section, $metadata);
257 $self->auto_extract_metadata ($doc_obj);
258 $self->plugin_specific_process($base_dir, $file, $doc_obj, $gli);
259 # if we haven't found any Title so far, assign one
260 $self->title_fallback($doc_obj,$section,$filename_no_path);
261
262 $self->add_OID($doc_obj);
263 return (1,$doc_obj);
264}
265
266
267sub parse_aux_json_metadata {
268 my $self = shift(@_);
269 my ($base_dir, $file, $doc_obj, $gli) = @_;
270
271 my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
272
273 my $topsection = $doc_obj->get_top_section();
274
275 my $json_metadata_filename = $filename_full_path;
276 $json_metadata_filename =~ s/\.mets.xml$/.json/;
277
278 my $json_text = "";
279 $self->ReadTextFile::read_file($json_metadata_filename,"utf8",undef,\$json_text);
280
281 my $json_rec = decode_json $json_text;
282 my $records = $json_rec->{'records'};
283 my @keys = keys %{$records};
284
285 my $key = shift @keys; # there should only be one
286 my $record = $records->{$key};
287
288 my @md_fields = ( "recordURL", "titles", "isbns", "issns", "oclcs", "lccns", "publishDates" );
289
290 foreach my $md_field (@md_fields) {
291 my $value_array = $record->{$md_field};
292
293 my $md_name = $md_field;
294 $md_name =~ s/s$//;
295
296 foreach my $md_value (@$value_array) {
297
298 if ($md_name eq "title") {
299 $doc_obj->set_utf8_metadata_element ($topsection, "Title", $md_value);
300 $doc_obj->set_utf8_metadata_element ($topsection, "dc.Title", $md_value);
301 }
302 else {
303 $doc_obj->set_utf8_metadata_element ($topsection, $md_name, $md_value);
304 }
305 }
306 }
307
308 my $htid = $json_rec->{'items'}->[0]->{'htid'};
309 my $docName = $htid;
310 my $docNameIE = $htid;
311 $docNameIE =~ s/^.*?\.//;
312
313 $doc_obj->set_utf8_metadata_element ($topsection, "docName", $docName);
314 $doc_obj->set_utf8_metadata_element ($topsection, "docNameIE", $docNameIE);
315
316}
317
318
319# override this for an inheriting plugin to add extra metadata etc
320sub plugin_specific_process {
321 my $self = shift(@_);
322 my ($base_dir, $file, $doc_obj, $gli) = @_;
323
324 $self->parse_aux_json_metadata($base_dir,$file,$doc_obj,$gli);
325}
326
327# sub tidy_item_file {
328# ... see PagedImagePlugin
329# }
330
331# sub rotate_image {
332# ... see PagedImagePlugin
333# }
334
335# sub process_image {
336# ... see PagedImagePlugin
337# }
338
339
340
341sub xml_start_tag {
342 my $self = shift(@_);
343 my ($expat, $element) = @_;
344 $self->{'element'} = $element;
345
346 my $doc_obj = $self->{'doc_obj'};
347 if ($element eq "METS:mets") {
348 $self->{'current_section'} = $doc_obj->get_top_section();
349# } elsif ($element eq "PageGroup" || $element eq "Page") {
350## if ($element eq "PageGroup") {
351## $self->{'has_internal_structure'} = 1;
352 }
353 elsif (($element eq "METS:FLocat") && ($_{'xlink:href'} =~ m/\.txt$/)) {
354 # e.g. <METS:FLocat LOCTYPE="OTHER" OTHERLOCTYPE="SYSTEM" xlink:href="00000000.txt"/>
355
356 # create a new section as a child
357 $self->{'current_section'} = $doc_obj->insert_section($doc_obj->get_end_child($self->{'current_section'}));
358 $self->{'num_pages'}++;
359 # assign pagenum as ... what?? => use page sequence number
360 my $txtfile = $_{'xlink:href'};
361 my ($pagenum) = ($txtfile =~ m/^(\d+)/);
362
363 if (defined $pagenum) {
364 my $pagenum_int = int($pagenum);
365 $doc_obj->set_utf8_metadata_element($self->{'current_section'}, "Title", "Page $pagenum_int");
366 }
367## my ($imgfile) = $_{'imgfile'};
368## if (defined $imgfile) {
369## # *****
370## # What about support for rotate image (e.g. old ':r' notation)?
371## $self->process_image($self->{'xml_file_dir'}.$imgfile, $imgfile, $doc_obj, $self->{'current_section'});
372## }
373
374## my ($txtfile) = $_{'txtfile'};
375 if (defined($txtfile)&& $txtfile ne "") {
376 my $full_txt_filename = &FileUtils::filenameConcatenate($self->{'xml_file_dir'},$txtfile);
377 $self->process_text ($full_txt_filename, $txtfile, $doc_obj, $self->{'current_section'});
378 } else {
379 $self->add_dummy_text($doc_obj, $self->{'current_section'});
380 }
381 }
382## elsif ($element eq "Metadata") {
383## $self->{'metadata_name'} = $_{'name'};
384## }
385}
386
387sub xml_end_tag {
388 my $self = shift(@_);
389 my ($expat, $element) = @_;
390
391 my $doc_obj = $self->{'doc_obj'};
392## if ($element eq "Page" || $element eq "PageGroup") {
393 if (($element eq "METS:FLocat") && ($_{'xlink:href'} =~ m/\.txt$/)) {
394 # if Title hasn't been assigned, set PageNum as Title
395 if (!defined $doc_obj->get_metadata_element ($self->{'current_section'}, "Title") && defined $doc_obj->get_metadata_element ($self->{'current_section'}, "PageNum" )) {
396 $doc_obj->add_utf8_metadata ($self->{'current_section'}, "Title", $doc_obj->get_metadata_element ($self->{'current_section'}, "PageNum" ));
397 }
398 # move the current section back to the parent
399 $self->{'current_section'} = $doc_obj->get_parent_section($self->{'current_section'});
400 } elsif ($element eq "Metadata") {
401
402 # text read in by XML::Parser is in Perl's binary byte value
403 # form ... need to explicitly make it UTF-8
404 my $meta_name = decode("utf-8",$self->{'metadata_name'});
405 my $metadata_value = decode("utf-8",$self->{'metadata_value'});
406
407 if ($meta_name =~ /\./) {
408 $meta_name = "ex.$meta_name";
409 }
410
411 $doc_obj->add_utf8_metadata ($self->{'current_section'}, $meta_name, $metadata_value);
412 $self->{'metadata_name'} = "";
413 $self->{'metadata_value'} = "";
414
415 }
416 # otherwise we ignore the end tag
417}
418
419
420sub xml_text {
421 my $self = shift(@_);
422 my ($expat) = @_;
423
424 if ($self->{'element'} eq "Metadata" && $self->{'metadata_name'}) {
425 $self->{'metadata_value'} .= $_;
426 }
427}
428
429sub xml_doctype {
430}
431
432sub open_document {
433 my $self = shift(@_);
434
435 # create a new document
436 $self->{'doc_obj'} = new doc ($self->{'filename'}, "indexed_doc", $self->{'file_rename_method'});
437 # TODO is file filenmae_no_path??
438 $self->set_initial_doc_fields($self->{'doc_obj'}, $self->{'filename'}, $self->{'processor'}, $self->{'metadata'});
439
440## my ($dir, $file) = $self->{'filename'} =~ /^(.*?)([^\/\\]*)$/;
441 my ($dir, $file_ext) = $self->{'filename'} =~ /^(.*?)(\.mets\.xml)$/;
442
443 $self->{'xml_file_dir'} = $dir;
444 $self->{'num_pages'} = 0;
445## $self->{'has_internal_structure'} = 0;
446
447}
448
449sub close_document {
450 my $self = shift(@_);
451 my $doc_obj = $self->{'doc_obj'};
452
453 my $topsection = $doc_obj->get_top_section();
454
455 # add numpages metadata
456 $doc_obj->set_utf8_metadata_element ($topsection, 'NumPages', $self->{'num_pages'}); # ##### !!!!
457
458 # set the document type
459 my $final_doc_type = "";
460## if ($self->{'documenttype'} eq "auto") {
461### if ($self->{'has_internal_structure'}) {
462### if ($self->{'gs_version'} eq "3") {
463### $final_doc_type = "pagedhierarchy";
464### }
465### else {
466### $final_doc_type = "hierarchy";
467### }
468### } else {
469### $final_doc_type = "paged";
470### }
471### } else {
472## # set to what doc type option was set to
473## $final_doc_type = $self->{'documenttype'};
474## }
475# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", $final_doc_type); # #### !!!!!
476 ### capiatalisation????
477# if ($self->{'documenttype'} eq 'paged') {
478 # set the gsdlthistype metadata to Paged - this ensures this document will
479 # be treated as a Paged doc, even if Titles are not numeric
480# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", "Paged");
481# } else {
482# $doc_obj->set_utf8_metadata_element ($topsection, "gsdlthistype", "Hierarchy");
483# }
484
485## $doc_obj->set_utf8_metadata_element($topsection,"MaxImageWidth",$self->{'MaxImageWidth'});
486## $doc_obj->set_utf8_metadata_element($topsection,"MaxImageHeight",$self->{'MaxImageHeight'});
487## $self->{'MaxImageWidth'} = undef;
488## $self->{'MaxImageHeight'} = undef;
489
490}
491
492
493sub set_initial_doc_fields {
494 my $self = shift(@_);
495 my ($doc_obj, $filename_full_path, $processor, $metadata) = @_;
496
497 my $topsection = $doc_obj->get_top_section();
498
499 my $plugin_filename_encoding = $self->{'filename_encoding'};
500 my $filename_encoding = $self->deduce_filename_encoding($filename_full_path,$metadata,$plugin_filename_encoding);
501 $self->set_Source_metadata($doc_obj, $filename_full_path, $filename_encoding);
502
503 # if we want a header page, we need to add some text into the top section, otherwise this section will become invisible
504 if ($self->{'headerpage'}) {
505 $self->add_dummy_text($doc_obj, $topsection);
506 }
507}
508
509sub scan_xml_for_files_to_block
510{
511 my $self = shift (@_);
512 my ($filename_full_path, $dir, $block_hash) = @_;
513
514 my ($file_root) = ($filename_full_path =~ m/^(.*)\.mets\.xml$/);
515
516 &util::block_filename($block_hash,"$file_root.zip");
517 &util::block_filename($block_hash,"$file_root.json");
518
519 my $page_dir = $file_root;
520
521 open (METSFILE, $filename_full_path) || die "couldn't open $filename_full_path to work out which files to block\n";
522 my $line = "";
523 while (defined ($line = <METSFILE>)) {
524 next unless $line =~ /\w/;
525
526 # Exaple of what we are looking for
527 # <METS:FLocat LOCTYPE="OTHER" OTHERLOCTYPE="SYSTEM" xlink:href="00000000.txt"/>
528
529 if ($line =~ /xlink:href=\"([^\"]+)\"/) {
530 my $txt_filename = &FileUtils::filenameConcatenate($page_dir,$1);
531 my $topics_filename = $txt_filename . ".topics";
532 &util::block_filename($block_hash,$txt_filename);
533 &util::block_filename($block_hash,$topics_filename);
534 }
535 }
536 close METSFILE;
537
538}
539
540
541sub process_text {
542 my $self = shift (@_);
543 my ($filename_full_path, $file, $doc_obj, $cursection) = @_;
544
545 # check that the text file exists!!
546 if (!-f $filename_full_path) {
547 print "HathiTrustMETSPlugin: ERROR: File $filename_full_path does not exist, skipping\n";
548 return 0;
549 }
550
551 # remember that this text file was one of our source files, but only
552 # if we are not processing a tmp file
553 if (!$self->{'processing_tmp_files'} ) {
554 $doc_obj->associate_source_file($filename_full_path);
555 }
556 # Do encoding stuff
557 my ($language, $encoding) = $self->textcat_get_language_encoding ($filename_full_path);
558
559 my $text="";
560 if ( -s $filename_full_path > 0 ) {
561 &ReadTextFile::read_file($self, $filename_full_path, $encoding, $language, \$text); # already decoded as utf8
562 }
563
564# HathiTrust often has empty files
565## if (!length ($text)) {
566## # It's a bit unusual but not out of the question to have no text, so just give a warning
567## print "HathiTrustMETSPlugin: WARNING: $filename_full_path contains no text\n";
568## }
569
570 # we need to escape the escape character, or else mg will convert into
571 # eg literal newlines, instead of leaving the text as '\n'
572 $text =~ s/\\/\\\\/g; # macro language
573 $text =~ s/_/\\_/g; # macro language
574
575
576 if ($text =~ m/<html.*?>\s*<head.*?>.*<\/head>\s*<body.*?>(.*)<\/body>\s*<\/html>\s*$/is) {
577 # looks like HTML input
578 # no need to escape < and > or put in <pre> tags
579
580 $text = $1;
581
582 # add text to document object
583 $doc_obj->add_utf8_text($cursection, "$text");
584 }
585 else {
586 $text =~ s/</&lt;/g;
587 $text =~ s/>/&gt;/g;
588
589 # insert preformat tags and add text to document object
590 $doc_obj->add_utf8_text($cursection, "<pre>\n$text\n</pre>");
591 }
592
593 my $topics_filename = $filename_full_path . ".topics";
594 if ( -s $topics_filename > 0 ) {
595
596 my $topics_text = "";
597 $self->ReadTextFile::read_file($topics_filename,"utf8",undef,\$topics_text);
598
599 my @topics_array = split(/\|/,$topics_text);
600 foreach my $topic (@topics_array) {
601 if ($topic ne "") {
602 $doc_obj->set_utf8_metadata_element ($cursection, "concept", $topic);
603 }
604 }
605 }
606
607 return 1;
608}
609
610
611sub clean_up_after_doc_obj_processing {
612 my $self = shift(@_);
613
614 $self->ImageConverter::clean_up_temporary_files();
615}
616
6171;
Note: See TracBrowser for help on using the repository browser.