source: main/trunk/greenstone2/perllib/plugins/PDFv1Plugin.pm@ 32783

Last change on this file since 32783 was 32783, checked in by kjdon, 2 years ago

adding missing strings and tidying up some mislabelling

File size: 14.9 KB
Line 
1###########################################################################
2#
3# PDFv1Plugin.pm -- The older pdf plugin, which uses the older pdftohtml
4# tool that can't handle newer versions of PDFs.
5# A component of the Greenstone digital library software
6# from the New Zealand Digital Library Project at the
7# University of Waikato, New Zealand.
8#
9# Copyright (C) 1999-2018 New Zealand Digital Library Project
10#
11# This program is free software; you can redistribute it and/or modify
12# it under the terms of the GNU General Public License as published by
13# the Free Software Foundation; either version 2 of the License, or
14# (at your option) any later version.
15#
16# This program is distributed in the hope that it will be useful,
17# but WITHOUT ANY WARRANTY; without even the implied warranty of
18# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19# GNU General Public License for more details.
20#
21# You should have received a copy of the GNU General Public License
22# along with this program; if not, write to the Free Software
23# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
24#
25###########################################################################
26package PDFv1Plugin;
27
28use strict;
29no strict 'refs'; # so we can use a var for filehandles (e.g. STDERR)
30no strict 'subs'; # allow filehandles to be variables and viceversa
31
32use ConvertBinaryFile;
33use ReadTextFile;
34use unicode;
35
36
37@PDFv1Plugin::ISA = ('ConvertBinaryFile', 'ReadTextFile');
38
39# PDFv1 plugin should be returned to being more like it was before AutoLoadConverters/PDFBox extension's inclusion
40# like the PDFPlugin was at http://trac.greenstone.org/browser/main/trunk/greenstone2/perllib/plugins/PDFPlugin.pm?rev=22597
41my $convert_to_list =
42 [ { 'name' => "auto",
43 'desc' => "{ConvertBinaryFile.convert_to.auto}" },
44 { 'name' => "html",
45 'desc' => "{ConvertBinaryFile.convert_to.html}" },
46 { 'name' => "text",
47 'desc' => "{ConvertBinaryFile.convert_to.text}" },
48 { 'name' => "pagedimg_jpg",
49 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_jpg}"},
50 { 'name' => "pagedimg_gif",
51 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_gif}"},
52 { 'name' => "pagedimg_png",
53 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_png}"},
54 ];
55
56
57my $arguments =
58 [
59 { 'name' => "convert_to",
60 'desc' => "{ConvertBinaryFile.convert_to}",
61 'type' => "enum",
62 'reqd' => "yes",
63 'list' => $convert_to_list,
64 'deft' => "html" },
65 { 'name' => "process_exp",
66 'desc' => "{BaseImporter.process_exp}",
67 'type' => "regexp",
68 'deft' => &get_default_process_exp(),
69 'reqd' => "no" },
70 { 'name' => "block_exp",
71 'desc' => "{CommonUtil.block_exp}",
72 'type' => "regexp",
73 'deft' => &get_default_block_exp() },
74 { 'name' => "metadata_fields",
75 'desc' => "{HTMLPlugin.metadata_fields}",
76 'type' => "string",
77 'deft' => "Title,Author,Subject,Keywords" },
78 { 'name' => "metadata_field_separator",
79 'desc' => "{HTMLPlugin.metadata_field_separator}",
80 'type' => "string",
81 'deft' => "" },
82 { 'name' => "noimages",
83 'desc' => "{PDFPlugin.noimages}",
84 'type' => "flag" },
85 { 'name' => "allowimagesonly",
86 'desc' => "{PDFPlugin.allowimagesonly}",
87 'type' => "flag" },
88 { 'name' => "complex",
89 'desc' => "{PDFPlugin.complex}",
90 'type' => "flag" },
91 { 'name' => "nohidden",
92 'desc' => "{PDFPlugin.nohidden}",
93 'type' => "flag" },
94 { 'name' => "zoom",
95 'desc' => "{PDFPlugin.zoom}",
96 'deft' => "2",
97 'range' => "1,3", # actually the range is 0.5-3
98 'type' => "int" },
99 { 'name' => "use_sections",
100 'desc' => "{PDFPlugin.use_sections}",
101 'type' => "flag" },
102 { 'name' => "description_tags",
103 'desc' => "{HTMLPlugin.description_tags}",
104 'type' => "flag" },
105 { 'name' => "use_realistic_book",
106 'desc' => "{PDFPlugin.use_realistic_book}",
107 'type' => "flag"}
108 ];
109
110my $options = { 'name' => "PDFv1Plugin",
111 'desc' => "{PDFv1Plugin.desc}",
112 'abstract' => "no",
113 'inherits' => "yes",
114 'srcreplaceable' => "yes", # Source docs in PDF can be replaced with GS-generated html
115 'args' => $arguments };
116
117sub new {
118 my ($class) = shift (@_);
119 my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
120 push(@$pluginlist, $class);
121
122 push(@$inputargs,"-title_sub");
123 push(@$inputargs,'^(Page\s+\d+)?(\s*1\s+)?');
124
125 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
126 push(@{$hashArgOptLists->{"OptList"}},$options);
127
128 my $self = new ConvertBinaryFile($pluginlist, $inputargs, $hashArgOptLists);
129
130 if ($self->{'info_only'}) {
131 # don't worry about any options etc
132 return bless $self, $class;
133 }
134
135 $self = bless $self, $class;
136 $self->{'file_type'} = "PDF";
137
138 # these are passed through to gsConvert.pl by ConvertBinaryFile.pm
139 my $zoom = $self->{"zoom"};
140 $self->{'convert_options'} = "-pdf_tool pdftohtml"; # PDFPluginv1 only ever uses the old pdftohtml conversion tool
141 $self->{'convert_options'} .= " -pdf_zoom $zoom";
142 $self->{'convert_options'} .= " -pdf_complex" if $self->{"complex"};
143 $self->{'convert_options'} .= " -pdf_nohidden" if $self->{"nohidden"};
144 $self->{'convert_options'} .= " -pdf_ignore_images" if $self->{"noimages"};
145 $self->{'convert_options'} .= " -pdf_allow_images_only" if $self->{"allowimagesonly"};
146
147 # check convert_to
148 if ($self->{'convert_to'} eq "text" && $ENV{'GSDLOS'} =~ /^windows$/i) {
149 &gsprintf::gsprintf(STDERR, "{PDFv1Plugin.win_old_pdftotext_unsupported}\n");
150 $self->{'convert_to'} = "html";
151 }
152 elsif ($self->{'convert_to'} eq "auto") {
153 # choose html ?? is this the best option
154 $self->{'convert_to'} = "html";
155 }
156 if ($self->{'use_realistic_book'}) {
157 if ($self->{'convert_to'} ne "html") {
158 print STDERR "PDFs will be converted to HTML for realistic book functionality\n";
159 $self->{'convert_to'} = "html";
160 }
161 }
162 # set convert_to_plugin and convert_to_ext
163 $self->set_standard_convert_settings();
164
165 my $secondary_plugin_name = $self->{'convert_to_plugin'};
166 my $secondary_plugin_options = $self->{'secondary_plugin_options'};
167
168 if (!defined $secondary_plugin_options->{$secondary_plugin_name}) {
169 $secondary_plugin_options->{$secondary_plugin_name} = [];
170 }
171 my $specific_options = $secondary_plugin_options->{$secondary_plugin_name};
172
173 # following title_sub removes "Page 1" added by pdftohtml, and a leading
174 # "1", which is often the page number at the top of the page. Bad Luck
175 # if your document title actually starts with "1 " - is there a better way?
176 push(@$specific_options , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
177 my $associate_tail_re = $self->{'associate_tail_re'};
178 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
179 push(@$specific_options, "-associate_tail_re", $associate_tail_re);
180 }
181 push(@$specific_options, "-file_rename_method", "none");
182
183 if ($secondary_plugin_name eq "HTMLPlugin") {
184 # pdftohtml always produces utf8
185 push(@$specific_options, "-input_encoding", "utf8");
186 push(@$specific_options, "-extract_language") if $self->{'extract_language'};
187 push(@$specific_options, "-processing_tmp_files");
188 # Instruct HTMLPlug (when eventually accessed through read_into_doc_obj)
189 # to extract these metadata fields from the HEAD META fields
190 if (defined $self->{'metadata_fields'} && $self->{'metadata_fields'} =~ /\S/) {
191 push(@$specific_options,"-metadata_fields",$self->{'metadata_fields'});
192 } else {
193 push(@$specific_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
194 }
195 if (defined $self->{'metadata_field_separator'} && $self->{'metadata_field_separator'} =~ /\S/) {
196 push(@$specific_options,"-metadata_field_separator",$self->{'metadata_field_separator'});
197 }
198 if ($self->{'use_sections'} || $self->{'description_tags'}) {
199 $self->{'description_tags'} = 1;
200 push(@$specific_options, "-description_tags");
201 }
202 if ($self->{'use_realistic_book'}) {
203 push(@$specific_options, "-use_realistic_book");
204 }
205 }
206 elsif ($secondary_plugin_name eq "PagedImagePlugin") {
207 push(@$specific_options, "-screenviewsize", "1000");
208 push(@$specific_options, "-enable_cache");
209 push(@$specific_options, "-processing_tmp_files");
210 }
211
212 $self = bless $self, $class; # Q TODO: why does it do this a 2nd time in this function?
213 $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
214 return $self;
215}
216
217sub get_default_process_exp {
218 my $self = shift (@_);
219
220 return q^(?i)\.pdf$^;
221}
222
223# so we don't inherit HTMLPlug's block exp...
224sub get_default_block_exp {
225 return "";
226}
227
228
229# By setting hashing to be on ga xml this ensures that two
230# PDF files that are identical except for the metadata
231# to hash to different values. Without this, when each PDF
232# file is converted to HTML there is a chance that they
233# will both be *identical* if the conversion utility does
234# not embed the metadata in the generated HTML. This is
235# certainly the case when PDFBOX is being used.
236
237# This change makes this convert to based plugin more
238# consistent with the original vision that the same document
239# with different metadata should
240# be seen as different.
241
242sub get_oid_hash_type {
243 my $self = shift (@_);
244 return "hash_on_ga_xml";
245}
246
247
248#sub tmp_area_convert_file {
249#
250# my $self = shift (@_);
251# return $self->AutoLoadConverters::tmp_area_convert_file(@_);
252#
253#}
254
255sub convert_post_process
256{
257 my $self = shift (@_);
258 my ($conv_filename) = @_;
259
260 my $outhandle=$self->{'outhandle'};
261
262 #$self->{'input_encoding'} = "utf8"; # The output is always in utf8 (is it?? it is for html, but what about other types?)
263 #my ($language, $encoding) = $self->textcat_get_language_encoding ($conv_filename);
264
265 # read in file ($text will be in utf8)
266 my $text = "";
267 # encoding will be utf8 for html files - what about other types? will we do this step for them anyway?
268 $self->read_file ($conv_filename, "utf8", "", \$text);
269
270 # Clean html from low and high surrogates D800–DFFF
271 $text =~ s@[\N{U+D800}-\N{U+DFFF}]@\ @g;
272
273 # Calculate number of pages based on <a ...> tags (we have a <a name=1> etc
274 # for each page). Metadata based on this calculation not set until process()
275 #
276 # Note: this is done even if we are not breaking the document into pages as it might
277 # be useful to give an indication of document length in browser through setting
278 # num_pages as metadata.
279 my @pages = ($text =~ m/\<[Aa] name=\"?\w+\"?>/ig); #<div style=\"?page-break-before:always; page-break-after:always\"?>
280 my $num_pages = scalar(@pages);
281 $self->{'num_pages'} = $num_pages;
282
283 if ($self->{'use_sections'}
284 && $self->{'converted_to'} eq "HTML") {
285
286 print $outhandle "PDFPlugin: Calculating sections...\n";
287
288 # we have "<a name=1></a>" etc for each page
289 # it may be <A name=
290 my @sections = split('<[Aa] name=', $text);
291
292 my $top_section = "";
293
294 if (scalar (@sections) == 1) { #only one section - no split!
295 print $outhandle "PDFPlugin: warning - no sections found\n";
296 } else {
297 $top_section .= shift @sections; # keep HTML header etc as top_section
298 }
299
300 # handle first section specially for title? Or all use first 100...
301
302 my $title = $sections[0];
303 $title =~ s/^\"?\w+\"?>//; # specific for pdftohtml...
304 $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
305 $title =~ s/<[^>]*>/ /g;
306 $title =~ s/(?:&nbsp;|\xc2\xa0)/ /g; # utf-8 for nbsp...
307 $title =~ s/^\s+//s;
308 $title =~ s/\s+$//;
309 $title =~ s/\s+/ /gs;
310 $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
311 $title =~ s/^\s+//s; # in case title_sub introduced any...
312 $title = substr ($title, 0, 100);
313 $title =~ s/\s\S*$/.../;
314
315
316 if (scalar (@sections) == 1) { # no sections found
317 $top_section .= $sections[0];
318 @sections=();
319 } else {
320 $top_section .= "<!--<Section>\n<Metadata name=\"Title\">$title</Metadata>\n-->\n <!--</Section>-->\n";
321 }
322
323 # add metadata per section...
324 foreach my $section (@sections) {
325 # section names are not always just digits, may be like "outline"
326 $section =~ s@^\"?(\w+)\"?></a>@@; # leftover from split expression...
327
328 $title = $1; # Greenstone does magic if sections are titled digits
329 if (! defined($title) ) {
330 print STDERR "no title: $section\n";
331 $title = " "; # get rid of the undefined warning in next line
332 }
333 my $newsection = "<!-- from PDFPlugin -->\n<!-- <Section>\n";
334 $newsection .= "<Metadata name=\"Title\">" . $title
335 . "</Metadata>\n--><br />\n"; #TODO: . "</Metadata>\n--><p>\n";
336 $newsection .= $section;
337 $newsection .= "<!--</Section>-->\n";
338 $section = $newsection;
339 }
340
341 $text=join('', ($top_section, @sections));
342 }
343
344 if ($self->{'use_sections'}
345 && $self->{'converted_to'} eq "text") {
346 print STDERR "**** When converting PDF to text, cannot apply use_sections\n";
347 }
348
349
350 # The following should no longer be needed, now that strings
351 # read in are Unicode aware (in the Perl sense) rather than
352 # raw binary strings that just happen to be UTF-8 compliant
353
354 # turn any high bytes that aren't valid utf-8 into utf-8.
355## unicode::ensure_utf8(\$text);
356
357 # Write it out again!
358 $self->utf8_write_file (\$text, $conv_filename);
359}
360
361
362# do plugin specific processing of doc_obj for HTML type
363sub process {
364 my $self = shift (@_);
365 my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
366
367 my $result = $self->process_type($base_dir,$file,$doc_obj);
368
369 # fix up the extracted date metadata to be in Greenstone date format,
370 # and fix the capitalisation of 'date'
371 my $cursection = $doc_obj->get_top_section();
372 foreach my $datemeta (@{$doc_obj->get_metadata($cursection, "date")}) {
373 $doc_obj->delete_metadata($cursection, "date", $datemeta);
374
375 # We're just interested in the date bit, not the time
376 # some pdf creators (eg "Acrobat 5.0 Scan Plug-in for Windows")
377 # set a /CreationDate, and set /ModDate to 000000000. pdftohtml
378 # extracts the ModDate, so it is 0...
379 $datemeta =~ /(\d+)-(\d+)-(\d+)/;
380 my ($year, $month, $day) = ($1,$2,$3);
381 if (defined($year) && defined($month) && defined($day)) {
382 if ($year == 0) {next}
383 if ($year < 100) {$year += 1900} # just to be safe
384 if ($month =~ /^\d$/) {$month="0$month"} # single digit
385 if ($day =~ /^\d$/) {$day="0$day"} # single digit
386 my $date="$year$month$day";
387 $doc_obj->add_utf8_metadata($cursection, "Date", $date);
388 }
389 }
390
391 $doc_obj->add_utf8_metadata($cursection, "NumPages", $self->{'num_pages'}) if defined $self->{'num_pages'};
392
393 if ($self->{'use_sections'} && $self->{'converted_to'} eq "HTML") {
394 # For gs2 we explicitly make it a paged document, cos greenstone won't get it
395 # right if any section has an empty title, or one with letters in it
396 if (&util::is_gs3()) {
397 # but for gs3, paged docs currently use image slider which is ugly if there are no images
398 $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Hierarchy");
399 } else {
400 $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Paged");
401 }
402 }
403
404 return $result;
405}
406
4071;
Note: See TracBrowser for help on using the repository browser.