| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # UnknownConverterPlugin.pm -- plugin that runs the provided cmdline cmd
|
|---|
| 4 | # to launch an custom unknown external conversion application that will
|
|---|
| 5 | # convert from some custom unknown format to one of txt, html or xml.
|
|---|
| 6 | #
|
|---|
| 7 | # A component of the Greenstone digital library software
|
|---|
| 8 | # from the New Zealand Digital Library Project at the
|
|---|
| 9 | # University of Waikato, New Zealand.
|
|---|
| 10 | #
|
|---|
| 11 | # Copyright (C) 1999-2005 New Zealand Digital Library Project
|
|---|
| 12 | #
|
|---|
| 13 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 14 | # it under the terms of the GNU General Public License as published by
|
|---|
| 15 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 16 | # (at your option) any later version.
|
|---|
| 17 | #
|
|---|
| 18 | # This program is distributed in the hope that it will be useful,
|
|---|
| 19 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 20 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 21 | # GNU General Public License for more details.
|
|---|
| 22 | #
|
|---|
| 23 | # You should have received a copy of the GNU General Public License
|
|---|
| 24 | # along with this program; if not, write to the Free Software
|
|---|
| 25 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 26 | #
|
|---|
| 27 | ###########################################################################
|
|---|
| 28 |
|
|---|
| 29 | package UnknownConverterPlugin;
|
|---|
| 30 |
|
|---|
| 31 | use strict;
|
|---|
| 32 | no strict 'subs';
|
|---|
| 33 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 34 |
|
|---|
| 35 | use ConvertBinaryFile;
|
|---|
| 36 | use UnknownPlugin;
|
|---|
| 37 |
|
|---|
| 38 | # TO DO:
|
|---|
| 39 | # - error messages and other display strings need to go into strings.properties
|
|---|
| 40 | # - Have a TEMPDIR placeholder in the command, which, if present, gets replaced with the usual tempdir location
|
|---|
| 41 | # of a collection, and in which case we have to clean up intermediate files generated in there at the end?
|
|---|
| 42 | # Add a check that the generated file or files generated in the output dir match the convert_to option selected
|
|---|
| 43 | # before trying to process them
|
|---|
| 44 | # Add option that says where output comes from: stdout of the process, file that gets generated, folder.
|
|---|
| 45 | # At present, a file or folder of files is assumed.
|
|---|
| 46 | # Need to look in there for files with extension process_ext.
|
|---|
| 47 | # Do we also need a html_multi option to convert_to? If supporting html_multi as output,
|
|---|
| 48 | # see PowerPointPlugin::read(), and revision 31764 of UnknownConverterPlugin.pm
|
|---|
| 49 | # Then a folder of html files is generated per document?
|
|---|
| 50 | # OR Flag that indicates whether an html file + associated folder (such as of images) gets generated. And name of assoc folder. Such output gets generated for instance when a doc file is replaced by its html version.
|
|---|
| 51 |
|
|---|
| 52 | sub BEGIN {
|
|---|
| 53 | @UnknownConverterPlugin::ISA = ('UnknownPlugin', 'ConvertBinaryFile');
|
|---|
| 54 | }
|
|---|
| 55 |
|
|---|
| 56 | my $convert_to_list =
|
|---|
| 57 | [ { 'name' => "text",
|
|---|
| 58 | 'desc' => "{ConvertBinaryFile.convert_to.text}" },
|
|---|
| 59 | { 'name' => "html",
|
|---|
| 60 | 'desc' => "{ConvertBinaryFile.convert_to.html}" },
|
|---|
| 61 | { 'name' => "pagedimg_jpg",
|
|---|
| 62 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_jpg}" },
|
|---|
| 63 | { 'name' => "pagedimg_gif",
|
|---|
| 64 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_gif}" },
|
|---|
| 65 | { 'name' => "pagedimg_png",
|
|---|
| 66 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_png}" }
|
|---|
| 67 | ];
|
|---|
| 68 |
|
|---|
| 69 | my $arguments =
|
|---|
| 70 | [ { 'name' => "exec_cmd",
|
|---|
| 71 | 'desc' => "{UnknownConverterPlugin.exec_cmd}",
|
|---|
| 72 | 'type' => "string",
|
|---|
| 73 | 'deft' => "",
|
|---|
| 74 | 'reqd' => "yes" },
|
|---|
| 75 | { 'name' => "convert_to",
|
|---|
| 76 | 'desc' => "{ConvertBinaryFile.convert_to}",
|
|---|
| 77 | 'type' => "enum",
|
|---|
| 78 | 'reqd' => "yes",
|
|---|
| 79 | 'list' => $convert_to_list,
|
|---|
| 80 | 'deft' => "text" } ];
|
|---|
| 81 |
|
|---|
| 82 | my $options = { 'name' => "UnknownConverterPlugin",
|
|---|
| 83 | 'desc' => "{UnknownConverterPlugin.desc}",
|
|---|
| 84 | 'abstract' => "no",
|
|---|
| 85 | 'inherits' => "yes",
|
|---|
| 86 | 'args' => $arguments };
|
|---|
| 87 |
|
|---|
| 88 |
|
|---|
| 89 | sub new {
|
|---|
| 90 | my ($class) = shift (@_);
|
|---|
| 91 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
|---|
| 92 | push(@$pluginlist, $class);
|
|---|
| 93 |
|
|---|
| 94 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 95 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 96 |
|
|---|
| 97 | my $unknown_converter_self = new UnknownPlugin($pluginlist, $inputargs, $hashArgOptLists, 1);
|
|---|
| 98 | my $cbf_self = new ConvertBinaryFile($pluginlist, $inputargs, $hashArgOptLists);
|
|---|
| 99 |
|
|---|
| 100 | # Need to feed the superclass plugins to merge_inheritance() below in the order that the
|
|---|
| 101 | # superclass plugins were declared in the ISA listing earlier in this file:
|
|---|
| 102 | my $self = BaseImporter::merge_inheritance($unknown_converter_self, $cbf_self);
|
|---|
| 103 |
|
|---|
| 104 | $self = bless $self, $class;
|
|---|
| 105 |
|
|---|
| 106 | if ($self->{'info_only'}) {
|
|---|
| 107 |
|
|---|
| 108 | # If running pluginfo, we don't need to go further. Copied from other plugins like PDFPlug
|
|---|
| 109 | # Returning here when running pluginfo also means that it doesn't matter that convert_to
|
|---|
| 110 | # is not yet set to its default value at this stage: no warnings at gli start (when pluginfo
|
|---|
| 111 | # is run) about convert_to being undefined
|
|---|
| 112 |
|
|---|
| 113 | # don't worry about any options etc
|
|---|
| 114 | return $self;
|
|---|
| 115 | }
|
|---|
| 116 |
|
|---|
| 117 | my $outhandle = $self->{'outhandle'};
|
|---|
| 118 |
|
|---|
| 119 | # Convert_To set up, including secondary_plugins for processing the text or html generated
|
|---|
| 120 | # set convert_to_plugin and convert_to_ext
|
|---|
| 121 | $self->set_standard_convert_settings();
|
|---|
| 122 |
|
|---|
| 123 | # one-time replacements in exec_cmd happen here
|
|---|
| 124 | # don't nest GSDL env paths in (escaped) double quotes, as they're likely to be part of larger paths
|
|---|
| 125 | # It's up to the user to do the (escaped) double-quoting around paths.
|
|---|
| 126 | my $cmd = $self->{'exec_cmd'};
|
|---|
| 127 | if($cmd) {
|
|---|
| 128 | $cmd =~ s@%%GSDL3SRCHOME@$ENV{'GSDL3SRCHOME'}@g;
|
|---|
| 129 | $cmd =~ s@%%GSDLHOME@$ENV{'GSDLHOME'}@g;
|
|---|
| 130 | $cmd =~ s@%%GSDL3HOME@$ENV{'GSDL3HOME'}@g;
|
|---|
| 131 | $self->{'exec_cmd'} = $cmd;
|
|---|
| 132 | }
|
|---|
| 133 |
|
|---|
| 134 | my $secondary_plugin_name = $self->{'convert_to_plugin'};
|
|---|
| 135 | my $secondary_plugin_options = $self->{'secondary_plugin_options'};
|
|---|
| 136 |
|
|---|
| 137 | if (!defined $secondary_plugin_options->{$secondary_plugin_name}) {
|
|---|
| 138 | $secondary_plugin_options->{$secondary_plugin_name} = [];
|
|---|
| 139 | }
|
|---|
| 140 | my $specific_options = $secondary_plugin_options->{$secondary_plugin_name};
|
|---|
| 141 |
|
|---|
| 142 | # using defaults for secondary plugins, taken from RTFPlugin
|
|---|
| 143 | push(@$specific_options, "-file_rename_method", "none");
|
|---|
| 144 | push(@$specific_options, "-extract_language") if $self->{'extract_language'};
|
|---|
| 145 | if ($secondary_plugin_name eq "TextPlugin") {
|
|---|
| 146 | push(@$specific_options, "-input_encoding", "utf8");
|
|---|
| 147 | }
|
|---|
| 148 | elsif ($secondary_plugin_name eq "HTMLPlugin") {
|
|---|
| 149 | push(@$specific_options, "-description_tags") if $self->{'description_tags'};
|
|---|
| 150 | push(@$specific_options, "-processing_tmp_files");
|
|---|
| 151 | }
|
|---|
| 152 | elsif ($secondary_plugin_name eq "PagedImagePlugin") {
|
|---|
| 153 | push(@$specific_options, "-screenviewsize", "1000");
|
|---|
| 154 | push(@$specific_options, "-enable_cache");
|
|---|
| 155 | push(@$specific_options, "-processing_tmp_files");
|
|---|
| 156 | }
|
|---|
| 157 |
|
|---|
| 158 | # bless again, copied from PDFPlugin, PowerPointPlugin
|
|---|
| 159 | $self = bless $self, $class;
|
|---|
| 160 | $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
|
|---|
| 161 | return $self;
|
|---|
| 162 | }
|
|---|
| 163 |
|
|---|
| 164 | # Called by UnknownPlugin::process()
|
|---|
| 165 | # Overriding here to ensure that the NoText flag (metadata) and dummy text are not set,
|
|---|
| 166 | # since, unlike UnknownPlugin, this plugin has a chance of extracting text from the unknown file format
|
|---|
| 167 | sub add_dummy_text {
|
|---|
| 168 | my $self = shift(@_);
|
|---|
| 169 | }
|
|---|
| 170 |
|
|---|
| 171 | # Are init, begin and deinit necessary (will they not get called automatically)?
|
|---|
| 172 | # Dr Bainbridge says it doesn't hurt for these to be explicitly defined here.
|
|---|
| 173 | # Copied here from PDFPlugin, PowerPointPlugin
|
|---|
| 174 | # https://stackoverflow.com/questions/42885207/why-doesnt-class-supernew-call-the-constructors-of-all-parent-classes-when
|
|---|
| 175 | # "$class->SUPER::new always calls A::new because A comes before B in @ISA. See method resolution order in perlobj: ..."
|
|---|
| 176 | # https://stackoverflow.com/questions/15414696/when-using-multiple-inheritance-in-perl-is-there-a-way-to-indicate-which-super-f
|
|---|
| 177 | sub init {
|
|---|
| 178 | my $self = shift (@_);
|
|---|
| 179 |
|
|---|
| 180 | # ConvertBinaryFile init
|
|---|
| 181 | $self->ConvertBinaryFile::init(@_);
|
|---|
| 182 | }
|
|---|
| 183 |
|
|---|
| 184 | sub begin {
|
|---|
| 185 | my $self = shift (@_);
|
|---|
| 186 |
|
|---|
| 187 | $self->ConvertBinaryFile::begin(@_);
|
|---|
| 188 |
|
|---|
| 189 | }
|
|---|
| 190 |
|
|---|
| 191 | sub deinit {
|
|---|
| 192 | my $self = shift (@_);
|
|---|
| 193 |
|
|---|
| 194 | $self->ConvertBinaryFile::deinit(@_);
|
|---|
| 195 |
|
|---|
| 196 | }
|
|---|
| 197 |
|
|---|
| 198 | # Called by ConvertBinaryFile::tmp_area_convert_file() to do the actual conversion
|
|---|
| 199 | # In order to call the custom conversion command, UnknownConverterPlugin needs to know the actual
|
|---|
| 200 | # input filename (which is the tmp_filename parameter) and the output file name, which this subroutine
|
|---|
| 201 | # will work out. Then it will run the conversion command.
|
|---|
| 202 | sub run_conversion_command {
|
|---|
| 203 | my $self = shift (@_);
|
|---|
| 204 | my ($tmp_dirname, $tmp_filename, $utf8_tailname, $lc_suffix, $tailname, $suffix) = @_;
|
|---|
| 205 |
|
|---|
| 206 | my $outhandle = $self->{'outhandle'};
|
|---|
| 207 | my $convert_to = $self->{'convert_to'};
|
|---|
| 208 | my $failhandle = $self->{'failhandle'};
|
|---|
| 209 | my $verbosity = $self->{'verbosity'};
|
|---|
| 210 |
|
|---|
| 211 | my $convert_to_ext = $self->{'convert_to_ext'};
|
|---|
| 212 | if ($verbosity > 0) {
|
|---|
| 213 | print $outhandle "Converting $tailname$suffix to $convert_to format with extension $convert_to_ext\n";
|
|---|
| 214 | }
|
|---|
| 215 |
|
|---|
| 216 | # The command to be executed must be provided the input filename and output file/dir name
|
|---|
| 217 | # input filename = tmp_filename
|
|---|
| 218 | # 1. We now work out the output filename. Code for it comes from
|
|---|
| 219 | # ConvertBinaryFile::tmp_area_convert_file(), but slightly modified
|
|---|
| 220 |
|
|---|
| 221 | my $output_type=$self->{'convert_to'};
|
|---|
| 222 |
|
|---|
| 223 | # store the *actual* output type and return the output filename
|
|---|
| 224 | # it's possible we requested conversion to html, but only to text succeeded
|
|---|
| 225 | #$self->{'convert_to_ext'} = $output_type;
|
|---|
| 226 | if ($output_type =~ /html/i) {
|
|---|
| 227 | $self->{'converted_to'} = "HTML";
|
|---|
| 228 | } elsif ($output_type =~ /te?xt/i) {
|
|---|
| 229 | $self->{'converted_to'} = "Text";
|
|---|
| 230 | } elsif ($output_type =~ /item/i || $output_type =~ /^pagedimg/){
|
|---|
| 231 | $self->{'converted_to'} = "PagedImage";
|
|---|
| 232 | }
|
|---|
| 233 |
|
|---|
| 234 | my $output_filename = $tmp_filename;
|
|---|
| 235 | my $output_dirname;
|
|---|
| 236 | if ($output_type =~ /item/i || $output_type =~ /^pagedimg/) {
|
|---|
| 237 | # running under windows
|
|---|
| 238 | if ($ENV{'GSDLOS'} =~ /^windows$/i) {
|
|---|
| 239 | $output_dirname = $tmp_dirname . "\\$utf8_tailname\\";
|
|---|
| 240 | } else {
|
|---|
| 241 | $output_dirname = $tmp_dirname . "\/$utf8_tailname\/";
|
|---|
| 242 | }
|
|---|
| 243 | $output_filename = $output_dirname . $utf8_tailname . ".item";
|
|---|
| 244 | } else {
|
|---|
| 245 | $output_filename =~ s/$lc_suffix$/.$output_type/;
|
|---|
| 246 | }
|
|---|
| 247 |
|
|---|
| 248 |
|
|---|
| 249 | # 2. Execute the conversion command and get the type of the result,
|
|---|
| 250 | # making sure the converter gives us the appropriate output type
|
|---|
| 251 |
|
|---|
| 252 | # On Linux: if the program isn't installed, $? tends to come back with 127, in any case neither 0 nor 1.
|
|---|
| 253 | # On Windows: echo %ERRORLEVEL% ends up as 9009 if the program is not installed.
|
|---|
| 254 | # If running the command returns 0, let's assume success and so the act of running the command
|
|---|
| 255 | # should produce either a text file or output to stdout.
|
|---|
| 256 |
|
|---|
| 257 | my $plugin_name = $self->{'plugin_type'}; # inherited from BaseImporter
|
|---|
| 258 |
|
|---|
| 259 | my $cmd = $self->{'exec_cmd'};
|
|---|
| 260 | if(!$cmd) { # empty string for instance
|
|---|
| 261 | print $outhandle "$plugin_name Conversion error: a command to execute is required, cmd provided is |$cmd|\n";
|
|---|
| 262 | return "";
|
|---|
| 263 | }
|
|---|
| 264 |
|
|---|
| 265 | # replace remaining occurrences of placeholders in cmd string, the ones that change per input file
|
|---|
| 266 | #$cmd =~ s@\"@\\"@g;
|
|---|
| 267 | $cmd =~ s@%%INPUT_FILE@\"$tmp_filename\"@g; # refer to the softlink
|
|---|
| 268 | if(defined $output_dirname) {
|
|---|
| 269 | $cmd =~ s@%%OUTPUT@\"$output_dirname\"@g;
|
|---|
| 270 | } else {
|
|---|
| 271 | $cmd =~ s@%%OUTPUT@\"$output_filename\"@g;
|
|---|
| 272 | }
|
|---|
| 273 |
|
|---|
| 274 | # Allow the user to use %%GSDL(3|3SRC)HOME and replace them here with the
|
|---|
| 275 | # OS-specific $GSDL(3|3SRC)HOME or %GSDL(3|3SRC)HOME%
|
|---|
| 276 | $cmd =~ s@%%GSDLHOME@\"$ENV{'GSDLHOME'}\"@g;
|
|---|
| 277 | $cmd =~ s@%%GSDL3HOME@\"$ENV{'GSDL3HOME'}\"@g;
|
|---|
| 278 | $cmd =~ s@%%GSDL3SRCHOME@\"$ENV{'GSDL3SRCHOME'}\"@g;
|
|---|
| 279 |
|
|---|
| 280 | # Some debugging
|
|---|
| 281 | if ($self->{'verbosity'} > 2) {
|
|---|
| 282 | print STDERR "$plugin_name: executing conversion cmd \n|$cmd|\n";
|
|---|
| 283 | print STDERR " on infile |$tmp_filename|\n";
|
|---|
| 284 | print STDERR " to produce expected $output_filename\n";
|
|---|
| 285 | }
|
|---|
| 286 |
|
|---|
| 287 | # Run the command at last
|
|---|
| 288 | my $status = system($cmd);
|
|---|
| 289 |
|
|---|
| 290 | if($status == 127 || $status == 9009) { # means the cmd isn't recognised on Unix and Windows, respectively
|
|---|
| 291 | print $outhandle "$plugin_name Conversion error: cmd unrecognised, may not be installed (got $status when running $cmd)\n";
|
|---|
| 292 | return "";
|
|---|
| 293 | }
|
|---|
| 294 |
|
|---|
| 295 | if($status != 0) {
|
|---|
| 296 | print $outhandle "$plugin_name Conversion error: conversion failed with exit value $status\n";
|
|---|
| 297 | return "";
|
|---|
| 298 | }
|
|---|
| 299 |
|
|---|
| 300 | # remove symbolic link to original file
|
|---|
| 301 | &FileUtils::removeFiles($tmp_filename);
|
|---|
| 302 |
|
|---|
| 303 |
|
|---|
| 304 | if(defined $output_dirname && ! -d $output_dirname) {
|
|---|
| 305 | print $outhandle "$plugin_name Conversion error: Output directory $output_dirname doesn't exist\n";
|
|---|
| 306 | return "";
|
|---|
| 307 | }
|
|---|
| 308 | elsif (! -f $output_filename) {
|
|---|
| 309 | print $outhandle "$plugin_name Conversion error: Output file $output_filename doesn't exist\n";
|
|---|
| 310 | return "";
|
|---|
| 311 | }
|
|---|
| 312 |
|
|---|
| 313 | # else, conversion success
|
|---|
| 314 |
|
|---|
| 315 | # if multiple images were generated by running the conversion
|
|---|
| 316 | if ($self->{'convert_to'} =~ /^pagedimg/) {
|
|---|
| 317 | my $item_filename = $self->generate_item_file($output_filename);
|
|---|
| 318 |
|
|---|
| 319 | if (!-e $item_filename) {
|
|---|
| 320 | print $outhandle "$plugin_name Conversion error: Item file $item_filename was not generated\n";
|
|---|
| 321 | return "";
|
|---|
| 322 | }
|
|---|
| 323 | $output_filename = $item_filename;
|
|---|
| 324 | }
|
|---|
| 325 |
|
|---|
| 326 | $self->{'output_dirname'} = $output_dirname;
|
|---|
| 327 | $self->{'output_filename'} = $output_filename;
|
|---|
| 328 |
|
|---|
| 329 | return $output_filename;
|
|---|
| 330 |
|
|---|
| 331 | }
|
|---|
| 332 |
|
|---|
| 333 |
|
|---|
| 334 | # use the read_into_doc_obj inherited from ConvertBinaryFile "to call secondary plugin stuff"
|
|---|
| 335 | sub read_into_doc_obj {
|
|---|
| 336 | my $self = shift (@_);
|
|---|
| 337 | $self->ConvertBinaryFile::read_into_doc_obj(@_);
|
|---|
| 338 | }
|
|---|
| 339 |
|
|---|
| 340 | sub process {
|
|---|
| 341 | my $self = shift (@_);
|
|---|
| 342 | $self->UnknownPlugin::process(@_);
|
|---|
| 343 | }
|
|---|
| 344 |
|
|---|
| 345 |
|
|---|
| 346 | 1;
|
|---|
