1 | ######################################################################
|
---|
2 | #
|
---|
3 | # GoogleVisionAPIConverter.pm -- helper plugin that allows other plugins
|
---|
4 | # (such as ImagePlugin and PagedImagePlugin) to extend their
|
---|
5 | # processing capability through sub-classing inheritence (such as
|
---|
6 | # GoogleVisionImagePlugin and GoogleVisionPagedImagePlugin) to
|
---|
7 | # expand the image processing capabilities at ingest time to
|
---|
8 | # include the Google Vision API allowing for: metadata labelling
|
---|
9 | # of objects within a scene; and OCR text recognition.
|
---|
10 | #
|
---|
11 | # A component of the Greenstone digital library software
|
---|
12 | # from the New Zealand Digital Library Project at the
|
---|
13 | # University of Waikato, New Zealand.
|
---|
14 | #
|
---|
15 | # Copyright (C) 1999 New Zealand Digital Library Project
|
---|
16 | #
|
---|
17 | # This program is free software; you can redistribute it and/or modify
|
---|
18 | # it under the terms of the GNU General Public License as published by
|
---|
19 | # the Free Software Foundation; either version 2 of the License, or
|
---|
20 | # (at your option) any later version.
|
---|
21 | #
|
---|
22 | # This program is distributed in the hope that it will be useful,
|
---|
23 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
24 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
25 | # GNU General Public License for more details.
|
---|
26 | #
|
---|
27 | # You should have received a copy of the GNU General Public License
|
---|
28 | # along with this program; if not, write to the Free Software
|
---|
29 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
30 | #
|
---|
31 | ###########################################################################
|
---|
32 |
|
---|
33 | package GoogleVisionAPIConverter;
|
---|
34 |
|
---|
35 | use strict;
|
---|
36 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
37 | no strict 'subs';
|
---|
38 |
|
---|
39 | use gsprintf;
|
---|
40 | use FileUtils;
|
---|
41 |
|
---|
42 | ##use ImagePlugin;
|
---|
43 | use BaseMediaConverter;
|
---|
44 |
|
---|
45 | use utf8;
|
---|
46 | use JSON; # qw( from_json, encode_json );
|
---|
47 |
|
---|
48 | sub BEGIN {
|
---|
49 | @GoogleVisionAPIConverter::ISA = ('BaseMediaConverter');
|
---|
50 | }
|
---|
51 |
|
---|
52 | my $arguments = [
|
---|
53 | { 'name' => "google_application_credentials",
|
---|
54 | 'desc' => "{GoogleVisionAPIConverter.google_applicatio_credentials}",
|
---|
55 | 'type' => "string",
|
---|
56 | 'reqd' => "no",
|
---|
57 | 'deft' => "google-sa-credentials-key.json"
|
---|
58 | },
|
---|
59 | { 'name' => "enable_image_labelling",
|
---|
60 | 'desc' => "{GoogleVisionAPIConverter.enable_image_labelling}",
|
---|
61 | 'type' => "flag",
|
---|
62 | 'reqd' => "no" },
|
---|
63 | { 'name' => "enable_image_ocr",
|
---|
64 | 'desc' => "{GoogleVisionAPIConverter.enable_image_ocr}",
|
---|
65 | 'type' => "flag",
|
---|
66 | 'reqd' => "no" },
|
---|
67 | { 'name' => "enable_document_ocr",
|
---|
68 | 'desc' => "{GoogleVisionAPIConverter.enable_document_ocr}",
|
---|
69 | 'type' => "flag",
|
---|
70 | 'reqd' => "no" }
|
---|
71 | ];
|
---|
72 |
|
---|
73 | my $options = { 'name' => "GoogleVisionAPIConverter",
|
---|
74 | 'desc' => "{GoogleVisionAPIConverter.desc}",
|
---|
75 | 'abstract' => "no",
|
---|
76 | 'inherits' => "yes",
|
---|
77 | 'args' => $arguments };
|
---|
78 |
|
---|
79 | sub new {
|
---|
80 | my ($class) = shift (@_);
|
---|
81 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
---|
82 | push(@$pluginlist, $class);
|
---|
83 |
|
---|
84 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
85 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
---|
86 |
|
---|
87 | my $self = new BaseMediaConverter($pluginlist, $inputargs, $hashArgOptLists, 1);
|
---|
88 |
|
---|
89 | return bless $self, $class;
|
---|
90 | }
|
---|
91 |
|
---|
92 | sub begin {
|
---|
93 | my $self = shift (@_);
|
---|
94 | my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
|
---|
95 |
|
---|
96 | if ($self->{'enable_image_ocr'} && $self->{'enable_document_ocr'}) {
|
---|
97 | print STDERR "Please use the following command syntax for vision types: (--enable_image_ocr | --enable_document_ocr) [--enable_image_labelling]\n";
|
---|
98 | print STDERR "\t\t --enable_image_ocr : optical character recognition for text within images\n";
|
---|
99 | print STDERR "\t\t --enable_document_ocr : optical character recognition for text within documents\n";
|
---|
100 | print STDERR "\t\t --enable_image_labelling : annotation labeling for objects within images\n";
|
---|
101 | exit(2);
|
---|
102 | }
|
---|
103 |
|
---|
104 | $self->SUPER::begin(@_);
|
---|
105 | }
|
---|
106 |
|
---|
107 | sub vision_monitor_line {
|
---|
108 | my ($line) = @_;
|
---|
109 |
|
---|
110 | my $had_error = 0;
|
---|
111 | my $generate_dot = 0;
|
---|
112 |
|
---|
113 | if ($line =~ m/^.*$/)
|
---|
114 | {
|
---|
115 | $generate_dot = 1;
|
---|
116 | }
|
---|
117 |
|
---|
118 | return ($had_error,$generate_dot);
|
---|
119 | }
|
---|
120 |
|
---|
121 | sub run_gv_convert {
|
---|
122 | my $self = shift (@_);
|
---|
123 | my ($filename,$file,$doc_obj,$opt_section) = @_;
|
---|
124 |
|
---|
125 | my $section = (defined $opt_section) ? $opt_section : $doc_obj->get_top_section();
|
---|
126 |
|
---|
127 | my $verbosity = $self->{'verbosity'};
|
---|
128 | my $outhandle = $self->{'outhandle'};
|
---|
129 | print $outhandle "----- GoogleVisionAPIConveter run_gv_convert -----\n";
|
---|
130 |
|
---|
131 | my @vision_types = (); # array containing target ocr / labelling type(s)
|
---|
132 |
|
---|
133 | if ($self->{'enable_image_labelling'}) { push(@vision_types, "enable_image_labelling"); }
|
---|
134 | if ($self->{'enable_image_ocr'}) { push(@vision_types, "enable_image_ocr"); }
|
---|
135 | if ($self->{'enable_document_ocr'}) { push(@vision_types, "enable_document_ocr"); }
|
---|
136 |
|
---|
137 | my $vision_types_length = scalar(@vision_types);
|
---|
138 |
|
---|
139 | if ($vision_types_length != 0) {
|
---|
140 |
|
---|
141 | $self->init_cache_for_file($filename);
|
---|
142 | my $cached_image_dir = $self->{'cached_dir'};
|
---|
143 | # my $cached_image_root = $self->{'cached_file_root'};
|
---|
144 |
|
---|
145 | # my $filename_no_path = &File::Basename::basename($filename);
|
---|
146 |
|
---|
147 | my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
|
---|
148 | my $credentials_filename = &FileUtils::filenameConcatenate($collect_dir, "etc", $self->{'google_application_credentials'});
|
---|
149 |
|
---|
150 | for my $vision_type (@vision_types) {
|
---|
151 |
|
---|
152 | my $ofile = "${vision_type}-google-vision-output.json";
|
---|
153 | my $ofilename = &FileUtils::filenameConcatenate($cached_image_dir,$ofile);
|
---|
154 |
|
---|
155 | my $vision_cmd = "vision.py --$vision_type --credentials \"$credentials_filename\" \"$filename\" \"$ofilename\"";
|
---|
156 |
|
---|
157 | $self->run_vision($file, $filename, $ofile, $ofilename, $vision_cmd, $vision_type, $doc_obj,$section);
|
---|
158 |
|
---|
159 | if ($vision_type eq "enable_document_ocr") {
|
---|
160 | my $gv_dococr_rec = { 'filename' => $ofilename, 'section' => $section };
|
---|
161 |
|
---|
162 | push(@{$self->{'gv-dococr-json-filename-recs'}}, $gv_dococr_rec);
|
---|
163 | }
|
---|
164 | }
|
---|
165 | }
|
---|
166 |
|
---|
167 | return "json";
|
---|
168 | }
|
---|
169 |
|
---|
170 | sub gv_ocr_bounding_box_rect
|
---|
171 | {
|
---|
172 | my $self = shift (@_);
|
---|
173 | my ($gv_block,) = @_;
|
---|
174 |
|
---|
175 | my $bbox_rect = undef;
|
---|
176 |
|
---|
177 | my $gv_boundingBox = $gv_block->{'boundingBox'};
|
---|
178 |
|
---|
179 | my $gv_vertices = $gv_boundingBox->{'vertices'};
|
---|
180 | my $gv_num_vertices = scalar(@$gv_vertices);
|
---|
181 |
|
---|
182 | if ($gv_num_vertices > 0) {
|
---|
183 | # print STDERR "**** gs_vertices[0] = ", JSON::encode_json($gv_vertices->[0]), "\n";
|
---|
184 |
|
---|
185 | # Discovered that sometimes the 'x' value in the 'vertices' structure is not defined
|
---|
186 | # So can't rely on picking up $gv_vertices->[0 for 'x' and 'y'
|
---|
187 | # start off with 'undef' and test for !defined in for-loop
|
---|
188 |
|
---|
189 | my $min_x = undef;
|
---|
190 | my $min_y = undef;
|
---|
191 | my $max_x = undef;
|
---|
192 | my $max_y = undef;
|
---|
193 |
|
---|
194 |
|
---|
195 | for (my $v=0; $v<$gv_num_vertices; $v++) {
|
---|
196 | my $x = $gv_vertices->[$v]->{'x'};
|
---|
197 | my $y = $gv_vertices->[$v]->{'y'};
|
---|
198 |
|
---|
199 | if (defined $x) {
|
---|
200 | $min_x = $x if (!defined $min_x || ($x < $min_x));
|
---|
201 | $max_x = $x if (!defined $max_x || ($x > $max_x));
|
---|
202 | }
|
---|
203 |
|
---|
204 | if (defined $y) {
|
---|
205 | $min_y = $y if (!defined $min_y || ($y < $min_y));
|
---|
206 | $max_y = $y if (!defined $max_y || ($y > $max_y));
|
---|
207 | }
|
---|
208 | }
|
---|
209 |
|
---|
210 | my $x_org = $min_x;
|
---|
211 | my $y_org = $min_y;
|
---|
212 | my $x_dim = $max_x - $min_x +1;
|
---|
213 | my $y_dim = $max_y - $min_y +1;
|
---|
214 |
|
---|
215 | $bbox_rect = { "x_org" => $x_org, "y_org" => $y_org, "x_dim" => $x_dim, "y_dim" => $y_dim};
|
---|
216 | }
|
---|
217 |
|
---|
218 | return $bbox_rect;
|
---|
219 | }
|
---|
220 |
|
---|
221 | sub run_vision
|
---|
222 | {
|
---|
223 | my $self = shift (@_);
|
---|
224 | my ($file, $filename, $ofile, $ofilename, $vision_cmd, $vision_type, $doc_obj,$section) = @_;
|
---|
225 |
|
---|
226 | my $vision_regenerated;
|
---|
227 | my $vision_result;
|
---|
228 | my $vision_error;
|
---|
229 |
|
---|
230 | my $print_info = {
|
---|
231 | 'message_prefix' => "GoogleVisionAPI",
|
---|
232 | 'message' => "Sending $file to GoogleVisionAPI using vision.py"
|
---|
233 | };
|
---|
234 |
|
---|
235 | ($vision_regenerated,$vision_result,$vision_error)
|
---|
236 | = $self->run_cached_general_cmd($vision_cmd,$filename,$ofilename,$print_info);
|
---|
237 |
|
---|
238 | # Need to work a bit harder in setting up the associated JSON file
|
---|
239 | # => strip of 'enable_' in favour of 'gv_'
|
---|
240 | # => add in section number as part of the file name to avoid clashes
|
---|
241 |
|
---|
242 | my $section_file_suffix = $section;
|
---|
243 | $section_file_suffix =~ s/\./_/g;
|
---|
244 |
|
---|
245 | my $assoc_ofile = $ofile;
|
---|
246 | $assoc_ofile =~ s/^enable_/gv_/;
|
---|
247 | $assoc_ofile =~ s/-google-vision//;
|
---|
248 | $assoc_ofile =~ s/\.(.*?)$/$section_file_suffix.$1/;
|
---|
249 |
|
---|
250 | $doc_obj->associate_file($ofilename,$assoc_ofile,"application/json",$section);
|
---|
251 |
|
---|
252 | my $json_text = do { # read in json file
|
---|
253 | open(my $json_fh, "<:encoding(UTF-8)", $ofilename)
|
---|
254 | or die("Can't open \"$ofilename\": $!\n");
|
---|
255 | local $/;
|
---|
256 | <$json_fh>
|
---|
257 | };
|
---|
258 |
|
---|
259 | my $decoded_json = JSON::from_json($json_text);
|
---|
260 |
|
---|
261 | my $ocr_text;
|
---|
262 | if ($vision_type eq "enable_document_ocr" || $vision_type eq "enable_image_ocr") {
|
---|
263 | $ocr_text = $decoded_json->{'textAnnotations'}[0]{'description'}; # access full ocr content
|
---|
264 | $doc_obj->add_utf8_text($section, $ocr_text); # append text to section
|
---|
265 |
|
---|
266 | my $blocks = $decoded_json->{'fullTextAnnotation'}{'pages'}[0]{'blocks'};
|
---|
267 | my %text_and_language;
|
---|
268 |
|
---|
269 | foreach my $block (@{$blocks}) {
|
---|
270 | foreach my $paragraph (@{$block->{'paragraphs'}}) {
|
---|
271 | foreach my $word (@{$paragraph->{'words'}}) {
|
---|
272 | my $detected_language = $word->{'property'}{'detectedLanguages'}[0]{'languageCode'} || "no_lang";
|
---|
273 | my $word_text = "";
|
---|
274 | foreach my $letter (@{$word->{'symbols'}}) {
|
---|
275 | $word_text .= $letter->{'text'};
|
---|
276 | }
|
---|
277 | $text_and_language{$detected_language} .= $word_text . " ";
|
---|
278 | }
|
---|
279 | }
|
---|
280 | }
|
---|
281 |
|
---|
282 | for (keys %text_and_language) {
|
---|
283 | $doc_obj->add_utf8_metadata($section, "z_" . $_, $text_and_language{$_});
|
---|
284 | }
|
---|
285 |
|
---|
286 |
|
---|
287 | my $assoc_json_metaname = "HasGoogleVision";
|
---|
288 |
|
---|
289 | if ($vision_type eq "enable_document_ocr") {
|
---|
290 | $assoc_json_metaname .= "DocumentOCRJSON";
|
---|
291 |
|
---|
292 | $doc_obj->add_utf8_metadata($section, "GVDocumentOCRJSON",$assoc_ofile);
|
---|
293 | }
|
---|
294 | else {
|
---|
295 | # $vision_type eq "enable_image_ocr"
|
---|
296 | $assoc_json_metaname .= "ImageOCRJSON";
|
---|
297 |
|
---|
298 | $doc_obj->add_utf8_metadata($section, "GVImageOCRJSON",$assoc_ofile);
|
---|
299 | }
|
---|
300 |
|
---|
301 | $doc_obj->add_utf8_metadata($section, $assoc_json_metaname, 1);
|
---|
302 | }
|
---|
303 | elsif ($vision_type eq "enable_image_labelling") {
|
---|
304 | $ocr_text = $decoded_json->{'labelAnnotations'};
|
---|
305 | foreach my $label (@{$ocr_text}) {
|
---|
306 | # write to metadata : 'description'='Book' 'score'='0.9' 'topicality'='0.9' 'mid'='/m/0123'
|
---|
307 | $doc_obj->add_utf8_metadata($section, "description", $label->{'description'});
|
---|
308 | $doc_obj->add_utf8_metadata($section, "score", $label->{'score'});
|
---|
309 | $doc_obj->add_utf8_metadata($section, "topicality", $label->{'topicality'});
|
---|
310 | $doc_obj->add_utf8_metadata($section, "mid", $label->{'mid'});
|
---|
311 |
|
---|
312 | # write to metadata : 'descriptions'='Book' 'Book_score'='0.9' 'Book_topicality'='0.9' 'Book_mid'='/m/0123'
|
---|
313 | $doc_obj->add_utf8_metadata($section, "descriptions", $label->{'description'});
|
---|
314 | $doc_obj->add_utf8_metadata($section, $label->{'description'} . "_score", $label->{'score'});
|
---|
315 | $doc_obj->add_utf8_metadata($section, $label->{'description'} . "_topicality", $label->{'topicality'});
|
---|
316 | $doc_obj->add_utf8_metadata($section, $label->{'description'} . "_mid", $label->{'mid'});
|
---|
317 |
|
---|
318 | }
|
---|
319 |
|
---|
320 | $doc_obj->add_utf8_metadata($section, "HasGoogleVisionImageLabellingJSON", 1);
|
---|
321 | $doc_obj->add_utf8_metadata($section, "GVImageLabellingJSON",$assoc_ofile);
|
---|
322 |
|
---|
323 | }
|
---|
324 | }
|
---|
325 |
|
---|
326 | sub start_openannotation_list
|
---|
327 | {
|
---|
328 | my $self = shift (@_);
|
---|
329 | my ($doc_obj) = @_;
|
---|
330 |
|
---|
331 | my $OID = $doc_obj->get_OID();
|
---|
332 |
|
---|
333 | my $openannotation_list = {
|
---|
334 | "\@context" => "http://www.shared-canvas.org/ns/context.json",
|
---|
335 | # "\@id" => "https://iiif.harvardartmuseums.org/manifests/object/299843/list/47174896",
|
---|
336 | "\@id" => "${OID}/openannotation-list.json", # #### **** make full URL to be unique? or greenstone3:site:collect:OID ??
|
---|
337 | "\@type" => "sc:AnnotationList",
|
---|
338 | "resources" => []
|
---|
339 | };
|
---|
340 |
|
---|
341 | $self->{'openannotation-list'} = $openannotation_list;
|
---|
342 | }
|
---|
343 |
|
---|
344 |
|
---|
345 |
|
---|
346 | sub convert_gvocr_to_openannotation_resource
|
---|
347 | {
|
---|
348 | my $self = shift (@_);
|
---|
349 | my ($gv_blocks, $doc_obj, $section) = @_;
|
---|
350 |
|
---|
351 | my $OID = $doc_obj->get_OID();
|
---|
352 | my $OID_with_section = ($section ne "") ? "${OID}_$section" : $OID;
|
---|
353 | $section = 1 if ($section eq ""); # occurs when the document is a single image
|
---|
354 |
|
---|
355 |
|
---|
356 | # Example Open Annotation resource (for single annotation):
|
---|
357 | # {
|
---|
358 | # "@context": "http://iiif.io/api/presentation/2/context.json",
|
---|
359 | # "@id": "https://iiif.harvardartmuseums.org/annotations/9641482",
|
---|
360 | # "@type": "oa:Annotation",
|
---|
361 | # "motivation": [
|
---|
362 | # "oa:commenting"
|
---|
363 | # ],
|
---|
364 | # "on": {
|
---|
365 | # "@type": "oa:SpecificResource",
|
---|
366 | # "full": "https://iiif.harvardartmuseums.org/manifests/object/299843/canvas/canvas-47174896",
|
---|
367 | # "selector": {
|
---|
368 | # "@type": "oa:FragmentSelector",
|
---|
369 | # "value": "xywh=622,591,642,940"
|
---|
370 | # },
|
---|
371 | # "within": {
|
---|
372 | # "@id": "https://iiif.harvardartmuseums.org/manifests/object/299843",
|
---|
373 | # "@type": "sc:Manifest"
|
---|
374 | # }
|
---|
375 | # },
|
---|
376 | # "resource": [
|
---|
377 | # {
|
---|
378 | # "@type": "dctypes:Text",
|
---|
379 | # "chars": "<p>age: 35-52<br/>gender: Female(66.337677%)<br/>CALM: 55.438412%<br/>CONFUSED: 3.949288%<br/>SURPRISED: 2.33092%<br/>DISGUSTED:
|
---|
380 | # 0.545727%<br/>HAPPY: 1.549943%<br/>ANGRY: 2.082294%<br/>SAD: 34.103416%<br/></p><p>Generated by AWS Rekognition</p>",
|
---|
381 | # "format": "text/html"
|
---|
382 | # }
|
---|
383 | # ]
|
---|
384 | # },
|
---|
385 |
|
---|
386 | my $self_openannotation_resources = $self->{'openannotation-list'}->{'resources'};
|
---|
387 |
|
---|
388 | my $block_i = 0;
|
---|
389 |
|
---|
390 | foreach my $block (@{$gv_blocks}) {
|
---|
391 | $block_i++;
|
---|
392 |
|
---|
393 | my $openannotation_resource = {
|
---|
394 | "\@context" => "http://iiif.io/api/presentation/2/context.json",
|
---|
395 | # "\@id" => "https://iiif.harvardartmuseums.org/annotations/9641482",
|
---|
396 | "\@id" => "${OID_with_section}/annotation/gv-block-$block_i", # #### **** make full URL to be unique? or greenstone3:site:collect:OID ??
|
---|
397 | "\@type" => "oa:Annotation",
|
---|
398 | "motivation" => [ "oa:commenting" ]
|
---|
399 | };
|
---|
400 |
|
---|
401 |
|
---|
402 | my $bbox_rect = $self->gv_ocr_bounding_box_rect($block);
|
---|
403 | my $bb_x_org = $bbox_rect->{'x_org'};
|
---|
404 | my $bb_y_org = $bbox_rect->{'y_org'};
|
---|
405 | my $bb_x_dim = $bbox_rect->{'x_dim'};
|
---|
406 | my $bb_y_dim = $bbox_rect->{'y_dim'};
|
---|
407 |
|
---|
408 | my $openannotation_on = {
|
---|
409 | "\@type" => "oa:SpecificResource",
|
---|
410 | # "full" => "https://iiif.harvardartmuseums.org/manifests/object/299843/canvas/canvas-47174896",
|
---|
411 | "full" => "${OID}/canvas/$section", # doc id + /canvas + page-i/sect # #### **** make full URL to be unique? or greenstone3:site:collect:OID ??
|
---|
412 | "selector" => {
|
---|
413 | "\@type" => "oa:FragmentSelector",
|
---|
414 | "value" => "xywh=${bb_x_org},${bb_y_org},${bb_x_dim},${bb_y_dim}"
|
---|
415 | },
|
---|
416 | "within" => {
|
---|
417 | #"\@id" => "https://iiif.harvardartmuseums.org/manifests/object/299843",
|
---|
418 | "\@id" => "${OID_with_section}/manifest", # #### **** make full URL to be unique? or greenstone3:site:collect:OID... ??
|
---|
419 | "\@type" => "sc:Manifest"
|
---|
420 | }
|
---|
421 | };
|
---|
422 | $openannotation_resource->{'on'} = $openannotation_on;
|
---|
423 |
|
---|
424 |
|
---|
425 | my $block_text_html = "";
|
---|
426 |
|
---|
427 | foreach my $paragraph (@{$block->{'paragraphs'}}) {
|
---|
428 | my $para_text = "";
|
---|
429 |
|
---|
430 | foreach my $word (@{$paragraph->{'words'}}) {
|
---|
431 | my $word_text = "";
|
---|
432 |
|
---|
433 | foreach my $letter (@{$word->{'symbols'}}) {
|
---|
434 | $word_text .= $letter->{'text'};
|
---|
435 | }
|
---|
436 |
|
---|
437 | $para_text .= " " if $para_text ne "";
|
---|
438 | $para_text .= $word_text;
|
---|
439 | }
|
---|
440 |
|
---|
441 | $block_text_html .= "<p>\n$para_text\n</p>\n\n";
|
---|
442 | }
|
---|
443 |
|
---|
444 | my $openannotation_inner_resource = [{
|
---|
445 | "\@type" => "dctypes:Text",
|
---|
446 | "chars" => "$block_text_html",
|
---|
447 | "format" => "text/html"
|
---|
448 | }];
|
---|
449 |
|
---|
450 | $openannotation_resource->{'resource'} = $openannotation_inner_resource;
|
---|
451 |
|
---|
452 | push(@$self_openannotation_resources,$openannotation_resource);
|
---|
453 | }
|
---|
454 | }
|
---|
455 |
|
---|
456 |
|
---|
457 | sub convert_and_append_openannotation_resources
|
---|
458 | {
|
---|
459 | my $self = shift (@_);
|
---|
460 | my ($gv_dococr_json_filename, $doc_obj, $section) = @_;
|
---|
461 |
|
---|
462 |
|
---|
463 | # Read in JSON file
|
---|
464 | my $json_text = do {
|
---|
465 | open(my $json_fh, "<:encoding(UTF-8)", $gv_dococr_json_filename)
|
---|
466 | or die("Can't open \"$gv_dococr_json_filename\": $!\n");
|
---|
467 | local $/;
|
---|
468 | <$json_fh>
|
---|
469 | };
|
---|
470 |
|
---|
471 | my $decoded_json = JSON::from_json($json_text);
|
---|
472 |
|
---|
473 | my $gv_blocks = $decoded_json->{'fullTextAnnotation'}->{'pages'}->[0]->{'blocks'};
|
---|
474 |
|
---|
475 | $self->convert_gvocr_to_openannotation_resource($gv_blocks, $doc_obj, $section);
|
---|
476 | }
|
---|
477 |
|
---|
478 |
|
---|
479 |
|
---|
480 | sub end_openannotation_list
|
---|
481 | {
|
---|
482 | my $self = shift (@_);
|
---|
483 | my ($doc_obj,$json_ofilename) = @_;
|
---|
484 |
|
---|
485 | my $ret_status = 1;
|
---|
486 |
|
---|
487 | if (!open(JOUT, "> $json_ofilename")) {
|
---|
488 | print STDERR "Error: Failed save Open Annotation List JSON to \"$json_ofilename\":\n $!\n";
|
---|
489 | $ret_status = 0;
|
---|
490 | }
|
---|
491 | else {
|
---|
492 | binmode(JOUT, ":utf8");
|
---|
493 |
|
---|
494 | my $openannotation_list = $self->{'openannotation-list'};
|
---|
495 | my $openannotation_list_json_text = JSON::encode_json($openannotation_list);
|
---|
496 |
|
---|
497 | print JOUT $openannotation_list_json_text;
|
---|
498 | close JOUT;
|
---|
499 |
|
---|
500 | }
|
---|
501 |
|
---|
502 | $self->{'openannotation-list'} = undef;
|
---|
503 |
|
---|
504 | return $ret_status;
|
---|
505 | }
|
---|
506 |
|
---|
507 | sub openannotation_list_associate_json
|
---|
508 | {
|
---|
509 | my $self = shift (@_);
|
---|
510 | my ($doc_obj, $gv_dococr_json_filename_recs) = @_;
|
---|
511 |
|
---|
512 | my $outhandle = $self->{'outhandle'};
|
---|
513 |
|
---|
514 | # Guaranteed to have at least one value in gv_dococr_json_filename_recs
|
---|
515 | #
|
---|
516 | # Legacy code used to have a '\d+' just before the '.json' reflecting page/section number
|
---|
517 | # Keep this in the regular expression, just in case,
|
---|
518 | #
|
---|
519 | my ($gv_dococr_filename_root) = ($gv_dococr_json_filename_recs->[0]->{'filename'} =~ m/^(.+)(?:\d+)?\.json$/);
|
---|
520 |
|
---|
521 | # slight of hand so new directory spot in cache_dir picked out is where we want it!
|
---|
522 | $gv_dococr_filename_root .= "/";
|
---|
523 |
|
---|
524 | my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
|
---|
525 | my $toplevel_cached_dir = &FileUtils::filenameConcatenate($collect_dir,"cached");
|
---|
526 |
|
---|
527 | $self->init_cache_for_file($gv_dococr_filename_root);
|
---|
528 | my $cached_dir = $self->{'cached_dir'};
|
---|
529 |
|
---|
530 | my $assoc_openannotation_json_ofile = "openannotation-list.json";
|
---|
531 | my $cached_openannotation_json_ofilename = &FileUtils::filenameConcatenate($cached_dir,$assoc_openannotation_json_ofile);
|
---|
532 |
|
---|
533 | my $needs_json_regen = 0;
|
---|
534 |
|
---|
535 | if (!-f $cached_openannotation_json_ofilename) {
|
---|
536 | $needs_json_regen = 1;
|
---|
537 | }
|
---|
538 | else {
|
---|
539 | for my $gv_json_filename_rec (@$gv_dococr_json_filename_recs) {
|
---|
540 | my $gv_json_filename = $gv_json_filename_rec->{'filename'};
|
---|
541 | if (-M $gv_json_filename > -M $cached_openannotation_json_ofilename) {
|
---|
542 | $needs_json_regen = 1;
|
---|
543 | last;
|
---|
544 | }
|
---|
545 | }
|
---|
546 | }
|
---|
547 |
|
---|
548 | my $saved_ok = 1;
|
---|
549 |
|
---|
550 | if ($needs_json_regen) {
|
---|
551 |
|
---|
552 | print $outhandle " OpenAnnotation-List: Generating $cached_openannotation_json_ofilename\n";
|
---|
553 |
|
---|
554 | $self->start_openannotation_list($doc_obj);
|
---|
555 |
|
---|
556 | for my $gv_json_filename_rec (@$gv_dococr_json_filename_recs) {
|
---|
557 | my $gv_json_filename = $gv_json_filename_rec->{'filename'};
|
---|
558 | my $section = $gv_json_filename_rec->{'section'};
|
---|
559 | $self->convert_and_append_openannotation_resources($gv_json_filename, $doc_obj,$section);
|
---|
560 | }
|
---|
561 |
|
---|
562 | $saved_ok = $self->end_openannotation_list($doc_obj,$cached_openannotation_json_ofilename);
|
---|
563 | }
|
---|
564 |
|
---|
565 | if ($saved_ok) {
|
---|
566 | print $outhandle " OpenAnnotation-List: Cached file $cached_openannotation_json_ofilename already exists\n";
|
---|
567 |
|
---|
568 | my $top_section = $doc_obj->get_top_section();
|
---|
569 | $doc_obj->associate_file($cached_openannotation_json_ofilename,$assoc_openannotation_json_ofile,"application/json",$top_section);
|
---|
570 | }
|
---|
571 |
|
---|
572 | return $saved_ok;
|
---|
573 | }
|
---|
574 |
|
---|
575 | sub opt_run_gen_openannotation
|
---|
576 | {
|
---|
577 | my $self = shift (@_);
|
---|
578 | my ($doc_obj) = @_;
|
---|
579 |
|
---|
580 | my $gv_dococr_json_filename_recs = $self->{'gv-dococr-json-filename-recs'};
|
---|
581 | my $num_gv_dococr_json_filename_recs = scalar(@$gv_dococr_json_filename_recs);
|
---|
582 |
|
---|
583 | my $ret_val_ok = 1;
|
---|
584 |
|
---|
585 | if ($num_gv_dococr_json_filename_recs > 0) {
|
---|
586 | $ret_val_ok = $self->openannotation_list_associate_json($doc_obj,$gv_dococr_json_filename_recs);
|
---|
587 | }
|
---|
588 |
|
---|
589 | return $ret_val_ok;
|
---|
590 | }
|
---|
591 |
|
---|
592 | 1;
|
---|
593 |
|
---|