source: main/trunk/greenstone2/perllib/plugins/PDFPlugin.pm@ 22864

Last change on this file since 22864 was 22864, checked in by kjdon, 14 years ago

needed use ConvertBinaryFile
  • Property svn:keywords set to Author Date Id Revision
File size: 13.1 KB
Line 
1###########################################################################
2#
3# PDFPlugin.pm -- reasonably with-it pdf plugin
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) 1999-2001 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###########################################################################
25package PDFPlugin;
26
27use strict;
28no strict 'refs'; # so we can use a var for filehandles (e.g. STDERR)
29
30use ReadTextFile;
31use unicode;
32
33use AutoLoadConverters;
34use ConvertBinaryFile;
35
36@PDFPlugin::ISA = ('ConvertBinaryFile', 'AutoLoadConverters', 'ReadTextFile');
37
38
39my $convert_to_list =
40 [ { 'name' => "auto",
41 'desc' => "{ConvertBinaryFile.convert_to.auto}" },
42 { 'name' => "html",
43 'desc' => "{ConvertBinaryFile.convert_to.html}" },
44 { 'name' => "text",
45 'desc' => "{ConvertBinaryFile.convert_to.text}" },
46 { 'name' => "pagedimg_jpg",
47 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_jpg}"},
48 { 'name' => "pagedimg_gif",
49 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_gif}"},
50 { 'name' => "pagedimg_png",
51 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_png}"},
52 ];
53
54
55my $arguments =
56 [
57 { 'name' => "convert_to",
58 'desc' => "{ConvertBinaryFile.convert_to}",
59 'type' => "enum",
60 'reqd' => "yes",
61 'list' => $convert_to_list,
62 'deft' => "html" },
63 { 'name' => "process_exp",
64 'desc' => "{BasePlugin.process_exp}",
65 'type' => "regexp",
66 'deft' => &get_default_process_exp(),
67 'reqd' => "no" },
68 { 'name' => "block_exp",
69 'desc' => "{BasePlugin.block_exp}",
70 'type' => "regexp",
71 'deft' => &get_default_block_exp() },
72 { 'name' => "metadata_fields",
73 'desc' => "{HTMLPlugin.metadata_fields}",
74 'type' => "string",
75 'deft' => "" },
76 { 'name' => "metadata_field_separator",
77 'desc' => "{HTMLPlugin.metadata_field_separator}",
78 'type' => "string",
79 'deft' => "" },
80 { 'name' => "noimages",
81 'desc' => "{PDFPlugin.noimages}",
82 'type' => "flag" },
83 { 'name' => "allowimagesonly",
84 'desc' => "{PDFPlugin.allowimagesonly}",
85 'type' => "flag" },
86 { 'name' => "complex",
87 'desc' => "{PDFPlugin.complex}",
88 'type' => "flag" },
89 { 'name' => "nohidden",
90 'desc' => "{PDFPlugin.nohidden}",
91 'type' => "flag" },
92 { 'name' => "zoom",
93 'desc' => "{PDFPlugin.zoom}",
94 'deft' => "2",
95 'range' => "1,3", # actually the range is 0.5-3
96 'type' => "int" },
97 { 'name' => "use_sections",
98 'desc' => "{PDFPlugin.use_sections}",
99 'type' => "flag" },
100 { 'name' => "description_tags",
101 'desc' => "{HTMLPlugin.description_tags}",
102 'type' => "flag" }
103 ];
104
105my $options = { 'name' => "PDFPlugin",
106 'desc' => "{PDFPlugin.desc}",
107 'abstract' => "no",
108 'inherits' => "yes",
109 'srcreplaceable' => "yes", # Source docs in PDF can be replaced with GS-generated html
110 'args' => $arguments };
111
112sub new {
113 my ($class) = shift (@_);
114 my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
115 push(@$pluginlist, $class);
116
117 push(@$inputargs,"-title_sub");
118 push(@$inputargs,'^(Page\s+\d+)?(\s*1\s+)?');
119
120 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
121 push(@{$hashArgOptLists->{"OptList"}},$options);
122
123 my $auto_converter_self = new AutoLoadConverters($pluginlist,$inputargs,$hashArgOptLists,["PDFBoxConverter"],1);
124 my $cbf_self = new ConvertBinaryFile($pluginlist, $inputargs, $hashArgOptLists);
125 my $self = BasePlugin::merge_inheritance($auto_converter_self, $cbf_self);
126
127 if ($self->{'info_only'}) {
128 # don't worry about any options etc
129 return bless $self, $class;
130 }
131
132 $self = bless $self, $class;
133 $self->{'filename_extension'} = "pdf";
134 $self->{'file_type'} = "PDF";
135
136 # these are passed through to gsConvert.pl by ConvertBinaryFile.pm
137 my $zoom = $self->{"zoom"};
138 $self->{'convert_options'} = "-pdf_zoom $zoom";
139 $self->{'convert_options'} .= " -pdf_complex" if $self->{"complex"};
140 $self->{'convert_options'} .= " -pdf_nohidden" if $self->{"nohidden"};
141 $self->{'convert_options'} .= " -pdf_ignore_images" if $self->{"noimages"};
142 $self->{'convert_options'} .= " -pdf_allow_images_only" if $self->{"allowimagesonly"};
143
144 # check convert_to
145 if ($self->{'convert_to'} eq "text" && $ENV{'GSDLOS'} =~ /^windows$/i) {
146 print STDERR "Windows does not support pdf to text. PDFs will be converted to HTML instead\n";
147 $self->{'convert_to'} = "html";
148 }
149 elsif ($self->{'convert_to'} eq "auto") {
150 # choose html ?? is this the best option
151 $self->{'convert_to'} = "html";
152 }
153 # set convert_to_plugin and convert_to_ext
154 $self->set_standard_convert_settings();
155
156 my $secondary_plugin_name = $self->{'convert_to_plugin'};
157 my $secondary_plugin_options = $self->{'secondary_plugin_options'};
158
159 if (!defined $secondary_plugin_options->{$secondary_plugin_name}) {
160 $secondary_plugin_options->{$secondary_plugin_name} = [];
161 }
162 my $specific_options = $secondary_plugin_options->{$secondary_plugin_name};
163
164 # following title_sub removes "Page 1" added by pdftohtml, and a leading
165 # "1", which is often the page number at the top of the page. Bad Luck
166 # if your document title actually starts with "1 " - is there a better way?
167 push(@$specific_options , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
168 my $associate_tail_re = $self->{'associate_tail_re'};
169 if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
170 push(@$specific_options, "-associate_tail_re", $associate_tail_re);
171 }
172 push(@$specific_options, "-file_rename_method", "none");
173
174 if ($secondary_plugin_name eq "HTMLPlugin") {
175 # pdftohtml always produces utf8 - What about pdfbox???
176 push(@$specific_options, "-input_encoding", "utf8");
177 push(@$specific_options, "-extract_language") if $self->{'extract_language'};
178 push(@$specific_options, "-processing_tmp_files");
179 # Instruct HTMLPlug (when eventually accessed through read_into_doc_obj)
180 # to extract these metadata fields from the HEAD META fields
181 if (defined $self->{'metadata_fields'} && $self->{'metadata_fields'} =~ /\S/) {
182 push(@$specific_options,"-metadata_fields",$self->{'metadata_fields'});
183 } else {
184 push(@$specific_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
185 }
186 if (defined $self->{'metadata_field_separator'} && $self->{'metadata_field_separator'} =~ /\S/) {
187 push(@$specific_options,"-metadata_field_separator",$self->{'metadata_field_separator'});
188 }
189 if ($self->{'use_sections'} || $self->{'description_tags'}) {
190 $self->{'description_tags'} = 1;
191 push(@$specific_options, "-description_tags");
192 }
193 }
194 elsif ($secondary_plugin_name eq "PagedImagePlugin") {
195 push(@$specific_options, "-screenviewsize", "1000");
196 push(@$specific_options, "-enable_cache");
197 push(@$specific_options, "-processing_tmp_files");
198 }
199
200 $self = bless $self, $class;
201 $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
202 return $self;
203}
204
205sub get_default_process_exp {
206 my $self = shift (@_);
207
208 return q^(?i)\.pdf$^;
209}
210
211# so we don't inherit HTMLPlug's block exp...
212sub get_default_block_exp {
213 return "";
214}
215
216sub init {
217 my $self = shift (@_);
218
219 # ConvertBinaryFile init
220 $self->SUPER::init(@_);
221 $self->AutoLoadConverters::init();
222
223}
224
225sub begin {
226 my $self = shift (@_);
227
228 $self->AutoLoadConverters::begin();
229 $self->SUPER::begin(@_);
230
231}
232
233sub deinit {
234 my $self = shift (@_);
235
236 $self->AutoLoadConverters::deinit();
237 $self->SUPER::deinit(@_);
238
239}
240
241
242sub tmp_area_convert_file {
243
244 my $self = shift (@_);
245 return $self->AutoLoadConverters::tmp_area_convert_file(@_);
246
247}
248
249sub convert_post_process
250{
251 my $self = shift (@_);
252 my ($conv_filename) = @_;
253
254 my $outhandle=$self->{'outhandle'};
255
256 #$self->{'input_encoding'} = "utf8"; # The output is always in utf8 (is it?? it is for html, but what about other types?)
257 #my ($language, $encoding) = $self->textcat_get_language_encoding ($conv_filename);
258
259 # read in file ($text will be in utf8)
260 my $text = "";
261 # encoding will be utf8 for html files - what about other types? will we do this step for them anyway?
262 $self->read_file ($conv_filename, "utf8", "", \$text);
263
264 # Calculate number of pages based on <a ...> tags (we have a <a name=1> etc
265 # for each page). Metadata based on this calculation not set until process()
266 #
267 # Note: this is done even if we are not breaking to document into pages as it might
268 # be useful to give an indication of document length in browser through setting
269 # num_pages as metadata.
270 my @pages = ($text =~ /\<[Aa] name=\"?\w+\"?>/ig);
271 my $num_pages = scalar(@pages);
272 $self->{'num_pages'} = $num_pages;
273
274 if ($self->{'use_sections'}
275 && $self->{'converted_to'} eq "HTML") {
276
277 print $outhandle "PDFPlugin: Calculating sections...\n";
278
279 # we have "<a name=1></a>" etc for each page
280 # it may be <A name=
281 my @sections = split('<[Aa] name=', $text);
282
283 my $top_section = "";
284
285 if (scalar (@sections) == 1) { #only one section - no split!
286 print $outhandle "PDFPlugin: warning - no sections found\n";
287 } else {
288 $top_section .= shift @sections; # keep HTML header etc as top_section
289 }
290
291 # handle first section specially for title? Or all use first 100...
292
293 my $title = $sections[0];
294 $title =~ s/^\"?\w+\"?>//; # specific for pdftohtml...
295 $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
296 $title =~ s/<[^>]*>/ /g;
297 $title =~ s/(?:&nbsp;|\xc2\xa0)/ /g; # utf-8 for nbsp...
298 $title =~ s/^\s+//s;
299 $title =~ s/\s+$//;
300 $title =~ s/\s+/ /gs;
301 $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
302 $title =~ s/^\s+//s; # in case title_sub introduced any...
303 $title = substr ($title, 0, 100);
304 $title =~ s/\s\S*$/.../;
305
306
307 if (scalar (@sections) == 1) { # no sections found
308 $top_section .= $sections[0];
309 @sections=();
310 } else {
311 $top_section .= "<!--<Section>\n<Metadata name=\"Title\">$title</Metadata>\n-->\n <!--</Section>-->\n";
312 }
313
314 # add metadata per section...
315 foreach my $section (@sections) {
316 # section names are not always just digits, may be like "outline"
317 $section =~ s@^\"?(\w+)\"?></a>@@; # leftover from split expression...
318
319 $title = $1; # Greenstone does magic if sections are titled digits
320 if (! defined($title) ) {
321 print STDERR "no title: $section\n";
322 $title = " "; # get rid of the undefined warning in next line
323 }
324 my $newsection = "<!-- from PDFPlugin -->\n<!-- <Section>\n";
325 $newsection .= "<Metadata name=\"Title\">" . $title
326 . "</Metadata>\n--><p>\n";
327 $newsection .= $section;
328 $newsection .= "<!--</Section>-->\n";
329 $section = $newsection;
330 }
331
332 $text=join('', ($top_section, @sections));
333 }
334
335 # turn any high bytes that aren't valid utf-8 into utf-8.
336 unicode::ensure_utf8(\$text);
337
338 # Write it out again!
339 $self->utf8_write_file (\$text, $conv_filename);
340}
341
342
343# do plugin specific processing of doc_obj for HTML type
344sub process {
345 my $self = shift (@_);
346 my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
347
348 my $result = $self->process_type($base_dir,$file,$doc_obj);
349
350 # fix up the extracted date metadata to be in Greenstone date format,
351 # and fix the capitalisation of 'date'
352 my $cursection = $doc_obj->get_top_section();
353 foreach my $datemeta (@{$doc_obj->get_metadata($cursection, "date")}) {
354 $doc_obj->delete_metadata($cursection, "date", $datemeta);
355
356 # We're just interested in the date bit, not the time
357 # some pdf creators (eg "Acrobat 5.0 Scan Plug-in for Windows")
358 # set a /CreationDate, and set /ModDate to 000000000. pdftohtml
359 # extracts the ModDate, so it is 0...
360 $datemeta =~ /(\d+)-(\d+)-(\d+)/;
361 my ($year, $month, $day) = ($1,$2,$3);
362 if (defined($year) && defined($month) && defined($day)) {
363 if ($year == 0) {next}
364 if ($year < 100) {$year += 1900} # just to be safe
365 if ($month =~ /^\d$/) {$month="0$month"} # single digit
366 if ($day =~ /^\d$/) {$day="0$day"} # single digit
367 my $date="$year$month$day";
368 $doc_obj->add_utf8_metadata($cursection, "Date", $date);
369 }
370 }
371
372 $doc_obj->add_utf8_metadata($cursection, "NumPages", $self->{'num_pages'});
373
374 if ($self->{'use_sections'} && $self->{'converted_to'} eq "HTML") {
375 # we explicitly make it a paged document, cos greenstone won't get it
376 # right if any section has an empty title, or one with letters in it
377 $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Paged");
378 }
379
380 return $result;
381}
382
3831;
Note: See TracBrowser for help on using the repository browser.