| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # PDFBoxConverter - helper plugin that does pdf document conversion with PDFBox
|
|---|
| 4 | #
|
|---|
| 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) 2010 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 | ###########################################################################
|
|---|
| 26 | package PDFBoxConverter;
|
|---|
| 27 |
|
|---|
| 28 | use BaseMediaConverter;
|
|---|
| 29 |
|
|---|
| 30 | use strict;
|
|---|
| 31 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 32 | no strict 'subs'; # allow barewords (eg STDERR) as function arguments
|
|---|
| 33 |
|
|---|
| 34 | #use HTML::Entities; # for encoding characters into their HTML entities when PDFBox converts to text
|
|---|
| 35 |
|
|---|
| 36 | use gsprintf 'gsprintf';
|
|---|
| 37 | use FileUtils;
|
|---|
| 38 |
|
|---|
| 39 | # these two variables mustn't be initialised here or they will get stuck
|
|---|
| 40 | # at those values.
|
|---|
| 41 | our $pdfbox_conversion_available;
|
|---|
| 42 | our $no_pdfbox_conversion_reason;
|
|---|
| 43 |
|
|---|
| 44 | BEGIN {
|
|---|
| 45 | @PDFBoxConverter::ISA = ('BaseMediaConverter');
|
|---|
| 46 |
|
|---|
| 47 | # Check that PDFBox is installed and available on the path
|
|---|
| 48 | $pdfbox_conversion_available = 1;
|
|---|
| 49 | $no_pdfbox_conversion_reason = "";
|
|---|
| 50 |
|
|---|
| 51 | if (!defined $ENV{'GEXT_PDFBOX'}) {
|
|---|
| 52 | $pdfbox_conversion_available = 0;
|
|---|
| 53 | $no_pdfbox_conversion_reason = "gextpdfboxnotinstalled";
|
|---|
| 54 | }
|
|---|
| 55 | else {
|
|---|
| 56 | my $gextpb_home = $ENV{'GEXT_PDFBOX'};
|
|---|
| 57 | my $pbajar = &FileUtils::filenameConcatenate($gextpb_home,"lib","java","pdfbox-app.jar");
|
|---|
| 58 |
|
|---|
| 59 | if (!-e $pbajar) {
|
|---|
| 60 | &gsprintf(STDERR,"**** Failed to find $pbajar\n");
|
|---|
| 61 | $pdfbox_conversion_available = 0;
|
|---|
| 62 | $no_pdfbox_conversion_reason = "gextpdfboxjarnotinstalled";
|
|---|
| 63 | }
|
|---|
| 64 | else {
|
|---|
| 65 | # test to see if java is in path
|
|---|
| 66 | # Need to run java -version instead of just java, since the %ERRORLEVEL% returned
|
|---|
| 67 | # for `java` (which is checked below for failure of the command) is 0 for JDK 1.6*
|
|---|
| 68 | # while %ERRORLEVEL% is 1 for JDK 1.7*
|
|---|
| 69 | # If `java -version` is run however, %ERRORLEVEL% returned is 0 if java is
|
|---|
| 70 | # installed, regardless of whether the JDK version is 1.6* or 1.7*.
|
|---|
| 71 | my $java = &util::get_java_command();
|
|---|
| 72 |
|
|---|
| 73 | my $cmd = "$java -version";
|
|---|
| 74 | if ($ENV{'GSDLOS'} =~ /^windows/i) {
|
|---|
| 75 | $cmd .= " >nul 2>&1"; # java 2>&1 >null or java >null 2>&1 both work (%ERRORLEVEL% is 0)
|
|---|
| 76 | }
|
|---|
| 77 | else {
|
|---|
| 78 | # On Ubuntu, java >/dev/null 2>&1 works,
|
|---|
| 79 | # but java 2>&1 >/dev/null doesn't work: output goes to screen anyway
|
|---|
| 80 | $cmd .= " >/dev/null 2>&1"; # " >/dev/null 2>&1 &" - don't need & at end for Linux Centos anymore (Ubuntu was already fine without it)
|
|---|
| 81 | }
|
|---|
| 82 |
|
|---|
| 83 | my $status = system($cmd);
|
|---|
| 84 |
|
|---|
| 85 | if ($status != 0) {
|
|---|
| 86 |
|
|---|
| 87 | my $error_message = "**** Testing for java\n";
|
|---|
| 88 | $error_message .= "Failed to run: $cmd\n";
|
|---|
| 89 | $error_message .= "Error variable: |$!| and status: $status\n";
|
|---|
| 90 |
|
|---|
| 91 | &gsprintf(STDERR, "PDFBoxConverter: $error_message");
|
|---|
| 92 |
|
|---|
| 93 | $pdfbox_conversion_available = 0;
|
|---|
| 94 | $no_pdfbox_conversion_reason = "couldnotrunjava";
|
|---|
| 95 | }
|
|---|
| 96 | }
|
|---|
| 97 | }
|
|---|
| 98 |
|
|---|
| 99 | }
|
|---|
| 100 |
|
|---|
| 101 | my $arguments = [ ];
|
|---|
| 102 |
|
|---|
| 103 | my $options = { 'name' => "PDFBoxConverter",
|
|---|
| 104 | 'desc' => "{PDFBoxConverter.desc}",
|
|---|
| 105 | 'abstract' => "yes",
|
|---|
| 106 | 'inherits' => "yes",
|
|---|
| 107 | 'args' => $arguments };
|
|---|
| 108 |
|
|---|
| 109 | sub new {
|
|---|
| 110 | my ($class) = shift (@_);
|
|---|
| 111 | my ($pluginlist,$inputargs,$hashArgOptLists,$auxilary) = @_;
|
|---|
| 112 | push(@$pluginlist, $class);
|
|---|
| 113 |
|
|---|
| 114 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 115 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 116 |
|
|---|
| 117 |
|
|---|
| 118 | my $self = new BaseMediaConverter($pluginlist, $inputargs,
|
|---|
| 119 | $hashArgOptLists, $auxilary);
|
|---|
| 120 |
|
|---|
| 121 | if ($self->{'info_only'}) {
|
|---|
| 122 | # don't worry about any options etc
|
|---|
| 123 | return bless $self, $class;
|
|---|
| 124 | }
|
|---|
| 125 | if ($pdfbox_conversion_available) {
|
|---|
| 126 | my $gextpb_home = $ENV{'GEXT_PDFBOX'};
|
|---|
| 127 | my $pbajar = &FileUtils::filenameConcatenate($gextpb_home,"lib","java","pdfbox-app.jar");
|
|---|
| 128 | my $java = &util::get_java_command();
|
|---|
| 129 | $self->{'pdfbox_txt_launch_cmd'} = "$java -cp \"$pbajar\" org.apache.pdfbox.tools.ExtractText";
|
|---|
| 130 | $self->{'pdfbox_html_launch_cmd'} = "$java -cp \"$pbajar\" -Dline.separator=\"<br />\" org.apache.pdfbox.tools.ExtractText";
|
|---|
| 131 | # $self->{'pdfbox_img_launch_cmd'} = "java -cp \"$pbajar\" org.apache.pdfbox.tools.PDFToImage"; # pdfbox 2.09 cmd for converting each PDF page to an image (jpg, png)
|
|---|
| 132 |
|
|---|
| 133 | # We use this next cmd to launch our new custom PDFBox class (PDFBoxToImagesAndText.java) to convert each PDF page into an image (gif, jpg, png)
|
|---|
| 134 | # AND its extracted text. Or just each page's extracted text. An item file is still generated,
|
|---|
| 135 | # but this time referring to txtfiles too, not just the images. Result: searchable paged output.
|
|---|
| 136 | # Our new custom class PDFBoxToImagesAndText.java lives in the new build folder, so add that to the classpath for the launch cmd
|
|---|
| 137 | my $pdfbox_build = &FileUtils::filenameConcatenate($gextpb_home,"build");
|
|---|
| 138 | my $classpath = &util::pathname_cat($pbajar,$pdfbox_build);
|
|---|
| 139 | $self->{'pdfbox_imgtxt_launch_cmd'} = "java -cp \"$classpath\" org.greenstone.pdfbox.PDFBoxToImagesAndText";
|
|---|
| 140 | }
|
|---|
| 141 | else {
|
|---|
| 142 | $self->{'no_pdfbox_conversion_reason'} = $no_pdfbox_conversion_reason;
|
|---|
| 143 |
|
|---|
| 144 | my $outhandle = $self->{'outhandle'};
|
|---|
| 145 | &gsprintf($outhandle, "PDFBoxConverter: {PDFBoxConverter.noconversionavailable} ({PDFBoxConverter.$no_pdfbox_conversion_reason})\n");
|
|---|
| 146 | }
|
|---|
| 147 |
|
|---|
| 148 | $self->{'pdfbox_conversion_available'} = $pdfbox_conversion_available;
|
|---|
| 149 |
|
|---|
| 150 | return bless $self, $class;
|
|---|
| 151 |
|
|---|
| 152 | }
|
|---|
| 153 |
|
|---|
| 154 | sub init {
|
|---|
| 155 | my $self = shift(@_);
|
|---|
| 156 | my ($verbosity, $outhandle, $failhandle) = @_;
|
|---|
| 157 |
|
|---|
| 158 | $self->{'pbtmp_file_paths'} = ();
|
|---|
| 159 | }
|
|---|
| 160 |
|
|---|
| 161 | sub deinit {
|
|---|
| 162 | my $self = shift(@_);
|
|---|
| 163 |
|
|---|
| 164 | $self->clean_up_temporary_files();
|
|---|
| 165 | }
|
|---|
| 166 |
|
|---|
| 167 |
|
|---|
| 168 | sub convert {
|
|---|
| 169 | my $self = shift(@_);
|
|---|
| 170 | my ($source_file_full_path, $target_file_type) = @_;
|
|---|
| 171 |
|
|---|
| 172 | return 0 unless $pdfbox_conversion_available;
|
|---|
| 173 | # check the filename
|
|---|
| 174 | return 0 if ( !-f $source_file_full_path);
|
|---|
| 175 |
|
|---|
| 176 | # Although PDFBoxConverter inherits from AutoLoadConverters and therefore
|
|---|
| 177 | # doesn't go through gsConvert.pl, still set the -pdf_tool flag in convert_options
|
|---|
| 178 | # in case in future PDFBoxConverter no longer inherits from AutoLoadConverters
|
|---|
| 179 | # and ends up going through gsConvert.pl
|
|---|
| 180 | $self->{'convert_options'} .= " -pdf_tool pdfbox";
|
|---|
| 181 |
|
|---|
| 182 | my $img_output_mode = 0;
|
|---|
| 183 |
|
|---|
| 184 | my $convert_to = $self->{'convert_to'};
|
|---|
| 185 | my $paged_txt_output_mode = ($convert_to =~ /(pagedimgtxt|paged_text)/) ? 1 : 0;
|
|---|
| 186 |
|
|---|
| 187 | # the following line is necessary to avoid 'uninitialised variable' error
|
|---|
| 188 | # messages concerning the converted_to member variable when PDFPlugin's
|
|---|
| 189 | # use_sections option is checked.
|
|---|
| 190 | # PDFBox plugin now processes use_sections option, when working with v1.5.0
|
|---|
| 191 | # of the PDFBox jar file (which embeds each page in special <div> tags).
|
|---|
| 192 | if ($target_file_type eq "html") {
|
|---|
| 193 | $self->{'converted_to'} = "HTML";
|
|---|
| 194 | } elsif ($target_file_type eq "jpg" || $target_file_type eq "png") { # || $target_file_type eq "gif"
|
|---|
| 195 | # GIF not supported by PDFBox at present, see https://pdfbox.apache.org/1.8/commandline.html#pdftoimage
|
|---|
| 196 | $self->{'converted_to'} = $target_file_type;
|
|---|
| 197 | $img_output_mode = 1;
|
|---|
| 198 | } else {
|
|---|
| 199 | $self->{'converted_to'} = "text";
|
|---|
| 200 | }
|
|---|
| 201 |
|
|---|
| 202 | my $outhandle = $self->{'outhandle'};
|
|---|
| 203 | my $verbosity = $self->{'verbosity'};
|
|---|
| 204 |
|
|---|
| 205 | my $source_file_no_path = &File::Basename::basename($source_file_full_path);
|
|---|
| 206 | # Determine the full name and path of the output file
|
|---|
| 207 | my $target_file_path;
|
|---|
| 208 | if ($self->{'enable_cache'}) {
|
|---|
| 209 | $self->init_cache_for_file($source_file_full_path);
|
|---|
| 210 | my $cache_dir = $self->{'cached_dir'};
|
|---|
| 211 | my $file_root = $self->{'cached_file_root'};
|
|---|
| 212 | #$file_root .= "_$convert_id" if ($convert_id ne "");
|
|---|
| 213 |
|
|---|
| 214 | # append the output filetype suffix only for non-image output formats, since for
|
|---|
| 215 | # images we can be outputting multiple image files per single PDF input file
|
|---|
| 216 | my $target_file = ($img_output_mode || $paged_txt_output_mode) ? "$file_root" : "$file_root.$target_file_type";
|
|---|
| 217 |
|
|---|
| 218 | $target_file_path = &FileUtils::filenameConcatenate($cache_dir,$target_file);
|
|---|
| 219 | }
|
|---|
| 220 | else {
|
|---|
| 221 | # this is in gsdl/tmp. get a tmp filename in collection instead???
|
|---|
| 222 | $target_file_path = &util::get_tmp_filename($target_file_type);
|
|---|
| 223 |
|
|---|
| 224 | # for image files, remove the suffix, since we can have many output image files
|
|---|
| 225 | # per input PDF (one img for each page of the PDF, for example)
|
|---|
| 226 | if($img_output_mode || $paged_txt_output_mode) {
|
|---|
| 227 | $target_file_path =~ s/\.[^.]*$//g;
|
|---|
| 228 | if(!&FileUtils::directoryExists($target_file_path)) {
|
|---|
| 229 | mkdir($target_file_path);
|
|---|
| 230 | }
|
|---|
| 231 |
|
|---|
| 232 | # once the item file for the imgs has been created, need to adjust target_file_path
|
|---|
| 233 |
|
|---|
| 234 | # below, we'll store the dir just created to pbtmp_file_paths, so all imgs and the
|
|---|
| 235 | # item file generated in it can be deleted in one go on clean_up
|
|---|
| 236 | }
|
|---|
| 237 |
|
|---|
| 238 | push(@{$self->{'pbtmp_file_paths'}}, $target_file_path);
|
|---|
| 239 | }
|
|---|
| 240 |
|
|---|
| 241 | # Generate and run the convert command
|
|---|
| 242 | my $convert_cmd = "";
|
|---|
| 243 |
|
|---|
| 244 | # want the filename without extension, because any images
|
|---|
| 245 | # are to be generated with the same filename as the PDF
|
|---|
| 246 | my ($tailname, $dirname, $suffix) = &File::Basename::fileparse($source_file_full_path, "\\.[^\\.]+\$");
|
|---|
| 247 |
|
|---|
| 248 | if($img_output_mode || $paged_txt_output_mode) { # converting each page to image and/or text
|
|---|
| 249 | my $output_prefix = &FileUtils::filenameConcatenate($target_file_path, $tailname);
|
|---|
| 250 |
|
|---|
| 251 | # Our custom class does renaming of the pages (simplified to just numbers) for PagedImagePlugin
|
|---|
| 252 | #$convert_cmd = $paged_txt_output_mode ? $self->{'pdfbox_imgtxt_launch_cmd'} : $self->{'pdfbox_img_launch_cmd'};
|
|---|
| 253 | $convert_cmd = $self->{'pdfbox_imgtxt_launch_cmd'};
|
|---|
| 254 | $convert_cmd .= " -imagesOnly" unless($paged_txt_output_mode); # set to images only unless there's text too
|
|---|
| 255 | if($img_output_mode) { # whether images-only or images-and-text mode
|
|---|
| 256 | $convert_cmd .= " -imageType $target_file_type";
|
|---|
| 257 | $convert_cmd .= " -dpi ". $self->{"dpi"} if defined $self->{"dpi"};
|
|---|
| 258 | } else { # img_output_mode off, so paged txt only and no images
|
|---|
| 259 | $convert_cmd .= " -textOnly";
|
|---|
| 260 | }
|
|---|
| 261 | $convert_cmd .= " -outputPrefix \"$output_prefix\"";
|
|---|
| 262 | $convert_cmd .= " \"$source_file_full_path\"";
|
|---|
| 263 |
|
|---|
| 264 | } else { # single stream of text or html
|
|---|
| 265 |
|
|---|
| 266 | if ($target_file_type eq "html") {
|
|---|
| 267 | $convert_cmd = $self->{'pdfbox_html_launch_cmd'};
|
|---|
| 268 | $convert_cmd .= " -html" if ($target_file_type eq "html");
|
|---|
| 269 | } else {
|
|---|
| 270 | $convert_cmd = $self->{'pdfbox_txt_launch_cmd'};
|
|---|
| 271 | }
|
|---|
| 272 | $convert_cmd .= " \"$source_file_full_path\" \"$target_file_path\"";
|
|---|
| 273 | }
|
|---|
| 274 |
|
|---|
| 275 | if ($verbosity>2) {
|
|---|
| 276 | &gsprintf($outhandle,"Convert command: $convert_cmd\n");
|
|---|
| 277 | }
|
|---|
| 278 |
|
|---|
| 279 | my $print_info = { 'message_prefix' => "PDFBox Conversion",
|
|---|
| 280 | 'message' => "Converting $source_file_no_path to: $target_file_type" };
|
|---|
| 281 | # $print_info->{'cache_mode'} = $cache_mode if ($cache_mode ne "");
|
|---|
| 282 |
|
|---|
| 283 | my ($regenerated,$result,$had_error)
|
|---|
| 284 | = $self->autorun_general_cmd($convert_cmd,$source_file_full_path, $target_file_path,$print_info);
|
|---|
| 285 |
|
|---|
| 286 | if($img_output_mode || $paged_txt_output_mode) {
|
|---|
| 287 | # now the images have been generated, generate the "$target_file_path/tailname.item"
|
|---|
| 288 | # item file for them, which is also the target_file_path that needs to be returned
|
|---|
| 289 | $target_file_path = &util::create_itemfile($target_file_path, $tailname, $target_file_type);
|
|---|
| 290 | #print STDERR "**** item file: $target_file_path\n";
|
|---|
| 291 | }
|
|---|
| 292 | elsif ($self->{'converted_to'} eq "text") {
|
|---|
| 293 | # ensure html entities are doubly escaped for pdfbox to text conversion: & -> &amp;
|
|---|
| 294 | # conversion to html does it automatically, but conversion to text doesn't
|
|---|
| 295 | # and this results in illegal characters in doc.xml
|
|---|
| 296 |
|
|---|
| 297 | my $fulltext = &FileUtils::readUTF8File($target_file_path);
|
|---|
| 298 | if(defined $fulltext) {
|
|---|
| 299 | #$fulltext = &HTML::Entities::encode($fulltext); # doesn't seem to help
|
|---|
| 300 | $fulltext =~ s@&@&@sg; # Kathy's fix to ensure doc contents don't break XML
|
|---|
| 301 | &FileUtils::writeUTF8File($target_file_path, \$fulltext);
|
|---|
| 302 | } else {
|
|---|
| 303 | print STDERR "PDFBoxConverter::convert(): Unable to read from converted file\n";
|
|---|
| 304 | $had_error = 1;
|
|---|
| 305 | }
|
|---|
| 306 | }
|
|---|
| 307 |
|
|---|
| 308 | if ($had_error) {
|
|---|
| 309 | return (0, $result,$target_file_path);
|
|---|
| 310 | }
|
|---|
| 311 | return (1, $result,$target_file_path);
|
|---|
| 312 | }
|
|---|
| 313 |
|
|---|
| 314 | sub convert_without_result {
|
|---|
| 315 | my $self = shift(@_);
|
|---|
| 316 |
|
|---|
| 317 | my $source_file_path = shift(@_);
|
|---|
| 318 | my $target_file_type = shift(@_);
|
|---|
| 319 | my $convert_options = shift(@_) || "";
|
|---|
| 320 | my $convert_id = shift(@_) || "";
|
|---|
| 321 |
|
|---|
| 322 | return $self->convert($source_file_path,$target_file_type,
|
|---|
| 323 | $convert_options,$convert_id,"without_result");
|
|---|
| 324 | }
|
|---|
| 325 |
|
|---|
| 326 | sub clean_up_temporary_files {
|
|---|
| 327 | my $self = shift(@_);
|
|---|
| 328 |
|
|---|
| 329 | foreach my $pbtmp_file_path (@{$self->{'pbtmp_file_paths'}}) {
|
|---|
| 330 | if (-d $pbtmp_file_path) {
|
|---|
| 331 | #print STDERR "@@@@@@ cleanup called on $pbtmp_file_path\n";
|
|---|
| 332 | &FileUtils::removeFilesRecursive($pbtmp_file_path);
|
|---|
| 333 | }
|
|---|
| 334 | elsif (-e $pbtmp_file_path) {
|
|---|
| 335 | &FileUtils::removeFiles($pbtmp_file_path);
|
|---|
| 336 | }
|
|---|
| 337 | }
|
|---|
| 338 |
|
|---|
| 339 | $self->{'pbtmp_file_paths'} = ();
|
|---|
| 340 | }
|
|---|
| 341 |
|
|---|
| 342 |
|
|---|
| 343 | 1;
|
|---|
