| 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 | ###########################################################################
|
|---|
| 25 | package PDFPlugin;
|
|---|
| 26 |
|
|---|
| 27 | use ReadTextFile;
|
|---|
| 28 | use unicode;
|
|---|
| 29 | use strict;
|
|---|
| 30 | no strict 'refs'; # so we can use a var for filehandles (e.g. STDERR)
|
|---|
| 31 |
|
|---|
| 32 | sub BEGIN {
|
|---|
| 33 | $::autoloadConverterScripting="PDFBoxConverter";
|
|---|
| 34 | require AutoloadConverterScripting;
|
|---|
| 35 |
|
|---|
| 36 | @PDFPlugin::ISA = ('AutoloadConverterScripting', 'ReadTextFile');
|
|---|
| 37 | }
|
|---|
| 38 |
|
|---|
| 39 | my $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 |
|
|---|
| 55 | my $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 |
|
|---|
| 105 | my $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' => $arguments };
|
|---|
| 111 |
|
|---|
| 112 | sub new {
|
|---|
| 113 | my ($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 $self = new AutoloadConverterScripting("PDFBoxConverter",$pluginlist, $inputargs, $hashArgOptLists);
|
|---|
| 124 |
|
|---|
| 125 | if ($self->{'info_only'}) {
|
|---|
| 126 | # don't worry about any options etc
|
|---|
| 127 | return bless $self, $class;
|
|---|
| 128 | }
|
|---|
| 129 |
|
|---|
| 130 | $self->{'filename_extension'} = "pdf";
|
|---|
| 131 | $self->{'file_type'} = "PDF";
|
|---|
| 132 |
|
|---|
| 133 | # these are passed through to gsConvert.pl by ConvertBinaryFile.pm
|
|---|
| 134 | my $zoom = $self->{"zoom"};
|
|---|
| 135 | $self->{'convert_options'} = "-pdf_zoom $zoom";
|
|---|
| 136 | $self->{'convert_options'} .= " -pdf_complex" if $self->{"complex"};
|
|---|
| 137 | $self->{'convert_options'} .= " -pdf_nohidden" if $self->{"nohidden"};
|
|---|
| 138 | $self->{'convert_options'} .= " -pdf_ignore_images" if $self->{"noimages"};
|
|---|
| 139 | $self->{'convert_options'} .= " -pdf_allow_images_only" if $self->{"allowimagesonly"};
|
|---|
| 140 |
|
|---|
| 141 | # check convert_to
|
|---|
| 142 | if ($self->{'convert_to'} eq "text" && $ENV{'GSDLOS'} =~ /^windows$/i) {
|
|---|
| 143 | print STDERR "Windows does not support pdf to text. PDFs will be converted to HTML instead\n";
|
|---|
| 144 | $self->{'convert_to'} = "html";
|
|---|
| 145 | }
|
|---|
| 146 | elsif ($self->{'convert_to'} eq "auto") {
|
|---|
| 147 | # choose html ?? is this the best option
|
|---|
| 148 | $self->{'convert_to'} = "html";
|
|---|
| 149 | }
|
|---|
| 150 | # set convert_to_plugin and convert_to_ext
|
|---|
| 151 | $self->set_standard_convert_settings();
|
|---|
| 152 |
|
|---|
| 153 | my $secondary_plugin_name = $self->{'convert_to_plugin'};
|
|---|
| 154 | my $secondary_plugin_options = $self->{'secondary_plugin_options'};
|
|---|
| 155 |
|
|---|
| 156 | if (!defined $secondary_plugin_options->{$secondary_plugin_name}) {
|
|---|
| 157 | $secondary_plugin_options->{$secondary_plugin_name} = [];
|
|---|
| 158 | }
|
|---|
| 159 | my $specific_options = $secondary_plugin_options->{$secondary_plugin_name};
|
|---|
| 160 |
|
|---|
| 161 | # following title_sub removes "Page 1" added by pdftohtml, and a leading
|
|---|
| 162 | # "1", which is often the page number at the top of the page. Bad Luck
|
|---|
| 163 | # if your document title actually starts with "1 " - is there a better way?
|
|---|
| 164 | push(@$specific_options , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
|
|---|
| 165 | my $associate_tail_re = $self->{'associate_tail_re'};
|
|---|
| 166 | if ((defined $associate_tail_re) && ($associate_tail_re ne "")) {
|
|---|
| 167 | push(@$specific_options, "-associate_tail_re", $associate_tail_re);
|
|---|
| 168 | }
|
|---|
| 169 | push(@$specific_options, "-file_rename_method", "none");
|
|---|
| 170 |
|
|---|
| 171 | if ($secondary_plugin_name eq "HTMLPlugin") {
|
|---|
| 172 | # pdftohtml always produces utf8
|
|---|
| 173 | push(@$specific_options, "-input_encoding", "utf8");
|
|---|
| 174 | push(@$specific_options, "-extract_language") if $self->{'extract_language'};
|
|---|
| 175 | push(@$specific_options, "-processing_tmp_files");
|
|---|
| 176 | # Instruct HTMLPlug (when eventually accessed through read_into_doc_obj)
|
|---|
| 177 | # to extract these metadata fields from the HEAD META fields
|
|---|
| 178 | if (defined $self->{'metadata_fields'} && $self->{'metadata_fields'} =~ /\S/) {
|
|---|
| 179 | push(@$specific_options,"-metadata_fields",$self->{'metadata_fields'});
|
|---|
| 180 | } else {
|
|---|
| 181 | push(@$specific_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
|
|---|
| 182 | }
|
|---|
| 183 | if (defined $self->{'metadata_field_separator'} && $self->{'metadata_field_separator'} =~ /\S/) {
|
|---|
| 184 | push(@$specific_options,"-metadata_field_separator",$self->{'metadata_field_separator'});
|
|---|
| 185 | }
|
|---|
| 186 | if ($self->{'use_sections'} || $self->{'description_tags'}) {
|
|---|
| 187 | $self->{'description_tags'} = 1;
|
|---|
| 188 | push(@$specific_options, "-description_tags");
|
|---|
| 189 | }
|
|---|
| 190 | }
|
|---|
| 191 | elsif ($secondary_plugin_name eq "PagedImagePlugin") {
|
|---|
| 192 | push(@$specific_options, "-screenviewsize", "1000");
|
|---|
| 193 | push(@$specific_options, "-enable_cache");
|
|---|
| 194 | push(@$specific_options, "-processing_tmp_files");
|
|---|
| 195 | }
|
|---|
| 196 |
|
|---|
| 197 | $self = bless $self, $class;
|
|---|
| 198 | $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
|
|---|
| 199 | return $self;
|
|---|
| 200 | }
|
|---|
| 201 |
|
|---|
| 202 | sub get_default_process_exp {
|
|---|
| 203 | my $self = shift (@_);
|
|---|
| 204 |
|
|---|
| 205 | return q^(?i)\.pdf$^;
|
|---|
| 206 | }
|
|---|
| 207 |
|
|---|
| 208 | # so we don't inherit HTMLPlug's block exp...
|
|---|
| 209 | sub get_default_block_exp {
|
|---|
| 210 | return "";
|
|---|
| 211 | }
|
|---|
| 212 |
|
|---|
| 213 | sub convert_post_process
|
|---|
| 214 | {
|
|---|
| 215 | my $self = shift (@_);
|
|---|
| 216 | my ($conv_filename) = @_;
|
|---|
| 217 |
|
|---|
| 218 | my $outhandle=$self->{'outhandle'};
|
|---|
| 219 |
|
|---|
| 220 | #$self->{'input_encoding'} = "utf8"; # The output is always in utf8 (is it?? it is for html, but what about other types?)
|
|---|
| 221 | #my ($language, $encoding) = $self->textcat_get_language_encoding ($conv_filename);
|
|---|
| 222 |
|
|---|
| 223 | # read in file ($text will be in utf8)
|
|---|
| 224 | my $text = "";
|
|---|
| 225 | # encoding will be utf8 for html files - what about other types? will we do this step for them anyway?
|
|---|
| 226 | $self->read_file ($conv_filename, "utf8", "", \$text);
|
|---|
| 227 |
|
|---|
| 228 | # Calculate number of pages based on <a ...> tags (we have a <a name=1> etc
|
|---|
| 229 | # for each page). Metadata based on this calculation not set until process()
|
|---|
| 230 | #
|
|---|
| 231 | # Note: this is done even if we are not breaking to document into pages as it might
|
|---|
| 232 | # be useful to give an indication of document length in browser through setting
|
|---|
| 233 | # num_pages as metadata.
|
|---|
| 234 | my @pages = ($text =~ /\<[Aa] name=\"?\w+\"?>/ig);
|
|---|
| 235 | my $num_pages = scalar(@pages);
|
|---|
| 236 | $self->{'num_pages'} = $num_pages;
|
|---|
| 237 |
|
|---|
| 238 | if ($self->{'use_sections'}
|
|---|
| 239 | && $self->{'converted_to'} eq "HTML") {
|
|---|
| 240 |
|
|---|
| 241 | print $outhandle "PDFPlugin: Calculating sections...\n";
|
|---|
| 242 |
|
|---|
| 243 | # we have "<a name=1></a>" etc for each page
|
|---|
| 244 | # it may be <A name=
|
|---|
| 245 | my @sections = split('<[Aa] name=', $text);
|
|---|
| 246 |
|
|---|
| 247 | my $top_section = "";
|
|---|
| 248 |
|
|---|
| 249 | if (scalar (@sections) == 1) { #only one section - no split!
|
|---|
| 250 | print $outhandle "PDFPlugin: warning - no sections found\n";
|
|---|
| 251 | } else {
|
|---|
| 252 | $top_section .= shift @sections; # keep HTML header etc as top_section
|
|---|
| 253 | }
|
|---|
| 254 |
|
|---|
| 255 | # handle first section specially for title? Or all use first 100...
|
|---|
| 256 |
|
|---|
| 257 | my $title = $sections[0];
|
|---|
| 258 | $title =~ s/^\"?\w+\"?>//; # specific for pdftohtml...
|
|---|
| 259 | $title =~ s/<\/([^>]+)><\1>//g; # (eg) </b><b> - no space
|
|---|
| 260 | $title =~ s/<[^>]*>/ /g;
|
|---|
| 261 | $title =~ s/(?: |\xc2\xa0)/ /g; # utf-8 for nbsp...
|
|---|
| 262 | $title =~ s/^\s+//s;
|
|---|
| 263 | $title =~ s/\s+$//;
|
|---|
| 264 | $title =~ s/\s+/ /gs;
|
|---|
| 265 | $title =~ s/^$self->{'title_sub'}// if ($self->{'title_sub'});
|
|---|
| 266 | $title =~ s/^\s+//s; # in case title_sub introduced any...
|
|---|
| 267 | $title = substr ($title, 0, 100);
|
|---|
| 268 | $title =~ s/\s\S*$/.../;
|
|---|
| 269 |
|
|---|
| 270 |
|
|---|
| 271 | if (scalar (@sections) == 1) { # no sections found
|
|---|
| 272 | $top_section .= $sections[0];
|
|---|
| 273 | @sections=();
|
|---|
| 274 | } else {
|
|---|
| 275 | $top_section .= "<!--<Section>\n<Metadata name=\"Title\">$title</Metadata>\n-->\n <!--</Section>-->\n";
|
|---|
| 276 | }
|
|---|
| 277 |
|
|---|
| 278 | # add metadata per section...
|
|---|
| 279 | foreach my $section (@sections) {
|
|---|
| 280 | # section names are not always just digits, may be like "outline"
|
|---|
| 281 | $section =~ s@^\"?(\w+)\"?></a>@@; # leftover from split expression...
|
|---|
| 282 |
|
|---|
| 283 | $title = $1; # Greenstone does magic if sections are titled digits
|
|---|
| 284 | if (! defined($title) ) {
|
|---|
| 285 | print STDERR "no title: $section\n";
|
|---|
| 286 | $title = " "; # get rid of the undefined warning in next line
|
|---|
| 287 | }
|
|---|
| 288 | my $newsection = "<!-- from PDFPlugin -->\n<!-- <Section>\n";
|
|---|
| 289 | $newsection .= "<Metadata name=\"Title\">" . $title
|
|---|
| 290 | . "</Metadata>\n--><p>\n";
|
|---|
| 291 | $newsection .= $section;
|
|---|
| 292 | $newsection .= "<!--</Section>-->\n";
|
|---|
| 293 | $section = $newsection;
|
|---|
| 294 | }
|
|---|
| 295 |
|
|---|
| 296 | $text=join('', ($top_section, @sections));
|
|---|
| 297 | }
|
|---|
| 298 |
|
|---|
| 299 | # turn any high bytes that aren't valid utf-8 into utf-8.
|
|---|
| 300 | unicode::ensure_utf8(\$text);
|
|---|
| 301 |
|
|---|
| 302 | # Write it out again!
|
|---|
| 303 | $self->utf8_write_file (\$text, $conv_filename);
|
|---|
| 304 | }
|
|---|
| 305 |
|
|---|
| 306 |
|
|---|
| 307 | # do plugin specific processing of doc_obj for HTML type
|
|---|
| 308 | sub process {
|
|---|
| 309 | my $self = shift (@_);
|
|---|
| 310 | my ($pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
|---|
| 311 |
|
|---|
| 312 | my $result = $self->process_type($base_dir,$file,$doc_obj);
|
|---|
| 313 |
|
|---|
| 314 | # fix up the extracted date metadata to be in Greenstone date format,
|
|---|
| 315 | # and fix the capitalisation of 'date'
|
|---|
| 316 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 317 | foreach my $datemeta (@{$doc_obj->get_metadata($cursection, "date")}) {
|
|---|
| 318 | $doc_obj->delete_metadata($cursection, "date", $datemeta);
|
|---|
| 319 |
|
|---|
| 320 | # We're just interested in the date bit, not the time
|
|---|
| 321 | # some pdf creators (eg "Acrobat 5.0 Scan Plug-in for Windows")
|
|---|
| 322 | # set a /CreationDate, and set /ModDate to 000000000. pdftohtml
|
|---|
| 323 | # extracts the ModDate, so it is 0...
|
|---|
| 324 | $datemeta =~ /(\d+)-(\d+)-(\d+)/;
|
|---|
| 325 | my ($year, $month, $day) = ($1,$2,$3);
|
|---|
| 326 | if (defined($year) && defined($month) && defined($day)) {
|
|---|
| 327 | if ($year == 0) {next}
|
|---|
| 328 | if ($year < 100) {$year += 1900} # just to be safe
|
|---|
| 329 | if ($month =~ /^\d$/) {$month="0$month"} # single digit
|
|---|
| 330 | if ($day =~ /^\d$/) {$day="0$day"} # single digit
|
|---|
| 331 | my $date="$year$month$day";
|
|---|
| 332 | $doc_obj->add_utf8_metadata($cursection, "Date", $date);
|
|---|
| 333 | }
|
|---|
| 334 | }
|
|---|
| 335 |
|
|---|
| 336 | $doc_obj->add_utf8_metadata($cursection, "NumPages", $self->{'num_pages'});
|
|---|
| 337 |
|
|---|
| 338 | if ($self->{'use_sections'} && $self->{'converted_to'} eq "HTML") {
|
|---|
| 339 | # we explicitly make it a paged document, cos greenstone won't get it
|
|---|
| 340 | # right if any section has an empty title, or one with letters in it
|
|---|
| 341 | $doc_obj->set_utf8_metadata_element ($cursection, "gsdlthistype", "Paged");
|
|---|
| 342 | }
|
|---|
| 343 |
|
|---|
| 344 | return $result;
|
|---|
| 345 | }
|
|---|
| 346 |
|
|---|
| 347 | 1;
|
|---|
