| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # OpenOfficeConverter - helper plugin that does office document conversion
|
|---|
| 4 | # using jodconverter combined with OpenOffice
|
|---|
| 5 | #
|
|---|
| 6 | # A component of the Greenstone digital library software
|
|---|
| 7 | # from the New Zealand Digital Library Project at the
|
|---|
| 8 | # University of Waikato, New Zealand.
|
|---|
| 9 | #
|
|---|
| 10 | # Copyright (C) 2010 New Zealand Digital Library Project
|
|---|
| 11 | #
|
|---|
| 12 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 13 | # it under the terms of the GNU General Public License as published by
|
|---|
| 14 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 15 | # (at your option) any later version.
|
|---|
| 16 | #
|
|---|
| 17 | # This program is distributed in the hope that it will be useful,
|
|---|
| 18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 20 | # GNU General Public License for more details.
|
|---|
| 21 | #
|
|---|
| 22 | # You should have received a copy of the GNU General Public License
|
|---|
| 23 | # along with this program; if not, write to the Free Software
|
|---|
| 24 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 25 | #
|
|---|
| 26 | ###########################################################################
|
|---|
| 27 | package OpenOfficeConverter;
|
|---|
| 28 |
|
|---|
| 29 | use ConvertBinaryFile;
|
|---|
| 30 | use BaseMediaConverter;
|
|---|
| 31 |
|
|---|
| 32 | use strict;
|
|---|
| 33 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 34 |
|
|---|
| 35 | use gsprintf 'gsprintf';
|
|---|
| 36 |
|
|---|
| 37 | # these two variables mustn't be initialised here or they will get stuck
|
|---|
| 38 | # at those values.
|
|---|
| 39 | our $openoffice_conversion_available;
|
|---|
| 40 | our $no_openoffice_conversion_reason;
|
|---|
| 41 |
|
|---|
| 42 | BEGIN {
|
|---|
| 43 | @OpenOfficeConverter::ISA = ('BaseMediaConverter');
|
|---|
| 44 |
|
|---|
| 45 | # Check that OpenOffice and jodconverter are installed and available on
|
|---|
| 46 | # the path
|
|---|
| 47 | $openoffice_conversion_available = 1;
|
|---|
| 48 | $no_openoffice_conversion_reason = "";
|
|---|
| 49 |
|
|---|
| 50 | if (! defined $ENV{'GEXT_OPENOFFICE'}) {
|
|---|
| 51 | $openoffice_conversion_available = 0;
|
|---|
| 52 | $no_openoffice_conversion_reason = "gextopenofficenotinstalled";
|
|---|
| 53 | }
|
|---|
| 54 | else {
|
|---|
| 55 | my $gextoo_home = $ENV{'GEXT_OPENOFFICE'};
|
|---|
| 56 | my $jodjar = &util::filename_cat($gextoo_home,"lib","java","jodconverter.jar");
|
|---|
| 57 |
|
|---|
| 58 | if (!-e $jodjar) {
|
|---|
| 59 | #print STDERR "Failed to find $jodjar\n";
|
|---|
| 60 | $openoffice_conversion_available = 0;
|
|---|
| 61 | $no_openoffice_conversion_reason = "gextjodconverternotinstalled";
|
|---|
| 62 | }
|
|---|
| 63 | else {
|
|---|
| 64 | # test to see if soffice is in path
|
|---|
| 65 | my $cmd = "soffice -headless 2>&1"; # for linux and mac
|
|---|
| 66 |
|
|---|
| 67 | # The port that jodconverter uses to connect to soffice.
|
|---|
| 68 | # We set this to 8100 (else jodconverter would have used
|
|---|
| 69 | # 2002 as the port number, which can clash with some programs
|
|---|
| 70 | # such as LogMeIn. Users can specify another port by setting
|
|---|
| 71 | # the environment variable in setup.bat or <ext>-setup.pl.
|
|---|
| 72 | if(!defined $ENV{'JODCONVERTER_PORT'}) {
|
|---|
| 73 | $ENV{'JODCONVERTER_PORT'} = "8100";
|
|---|
| 74 | }
|
|---|
| 75 |
|
|---|
| 76 | if ($ENV{'GSDLOS'} =~ m/^windows$/) {
|
|---|
| 77 | if(!defined $ENV{'SOFFICE_HOST'}) {
|
|---|
| 78 | $ENV{'SOFFICE_HOST'} = "localhost";
|
|---|
| 79 | }
|
|---|
| 80 | if(!defined $ENV{'SOFFICE_PORT'}) {
|
|---|
| 81 | $ENV{'SOFFICE_PORT'} = "8200";
|
|---|
| 82 | }
|
|---|
| 83 | # important to have this set when the Greenstone server
|
|---|
| 84 | # and open office is on a remote windows machine
|
|---|
| 85 | if(!defined $ENV{'SOFFICE_HOME'}) {
|
|---|
| 86 | $ENV{'SOFFICE_HOME'} = &util::filename_cat($ENV{'ProgramFiles'},"OpenOffice.org 3");
|
|---|
| 87 | }
|
|---|
| 88 | my $ooffice_dir_guess =
|
|---|
| 89 | &util::filename_cat($ENV{'SOFFICE_HOME'},"program");
|
|---|
| 90 | if (-d $ooffice_dir_guess) {
|
|---|
| 91 | &util::envvar_append("PATH",$ooffice_dir_guess);
|
|---|
| 92 | }
|
|---|
| 93 |
|
|---|
| 94 | # for windows, when working on a remote system, want to be able to start OO if
|
|---|
| 95 | # not already running. We'll use the uno socket method to do so. Else client-gli
|
|---|
| 96 | # tends to hang, waiting for the prompt to return after OO has been started up
|
|---|
| 97 | # (which doesn't happen, so need to Ctrl-C GLI and run it again for it to work).
|
|---|
| 98 | $cmd = "start \"soffice process\" soffice \"-accept=socket,host=$ENV{'SOFFICE_HOST'},port=$ENV{'SOFFICE_PORT'};urp;StarOffice.ServiceManager\" -headless 2>&1";
|
|---|
| 99 | $cmd .= " >nul";
|
|---|
| 100 | #print STDERR "***** Tried to start-up OpenOffice with:\n$cmd\n";
|
|---|
| 101 | }
|
|---|
| 102 | else {
|
|---|
| 103 | # Windows seems to launch OpenOffice as a service (i.e.
|
|---|
| 104 | # automatically puts it in the background).
|
|---|
| 105 | # For Unix putting it in the background needs to be done
|
|---|
| 106 | # explicitly
|
|---|
| 107 |
|
|---|
| 108 | $cmd .= " >/dev/null &";
|
|---|
| 109 | }
|
|---|
| 110 |
|
|---|
| 111 | my $status = system($cmd);
|
|---|
| 112 | if ($status != 0) {
|
|---|
| 113 | #print STDERR "Failed to run: $cmd\n";
|
|---|
| 114 | #print STDERR "$!\n";
|
|---|
| 115 | $openoffice_conversion_available = 0;
|
|---|
| 116 | $no_openoffice_conversion_reason = "openofficenotinstalled";
|
|---|
| 117 | }
|
|---|
| 118 | }
|
|---|
| 119 | }
|
|---|
| 120 | }
|
|---|
| 121 |
|
|---|
| 122 | my $arguments = [
|
|---|
| 123 | { 'name' => "openoffice_port",
|
|---|
| 124 | 'desc' => "{OpenOfficeConverter.openoffice_port}",
|
|---|
| 125 | 'type' => "int",
|
|---|
| 126 | 'deft' => "8100",
|
|---|
| 127 | 'range' => "81,",
|
|---|
| 128 | 'reqd' => "no" },
|
|---|
| 129 | ];
|
|---|
| 130 |
|
|---|
| 131 |
|
|---|
| 132 | my $options = { 'name' => "OpenOfficeConverter",
|
|---|
| 133 | 'desc' => "{OpenOfficeConverter.desc}",
|
|---|
| 134 | 'abstract' => "yes",
|
|---|
| 135 | 'inherits' => "yes",
|
|---|
| 136 | 'args' => $arguments };
|
|---|
| 137 |
|
|---|
| 138 | sub new {
|
|---|
| 139 | my ($class) = shift (@_);
|
|---|
| 140 | my ($pluginlist,$inputargs,$hashArgOptLists,$auxilary) = @_;
|
|---|
| 141 | push(@$pluginlist, $class);
|
|---|
| 142 |
|
|---|
| 143 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 144 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 145 |
|
|---|
| 146 | my $self = new BaseMediaConverter($pluginlist, $inputargs,
|
|---|
| 147 | $hashArgOptLists, $auxilary);
|
|---|
| 148 |
|
|---|
| 149 | if ($self->{'info_only'}) {
|
|---|
| 150 | # don't worry about any options etc
|
|---|
| 151 | return bless $self, $class;
|
|---|
| 152 | }
|
|---|
| 153 | if (!$openoffice_conversion_available) {
|
|---|
| 154 | $self->{'no_openoffice_conversion_reason'} = $no_openoffice_conversion_reason;
|
|---|
| 155 |
|
|---|
| 156 | my $outhandle = $self->{'outhandle'};
|
|---|
| 157 | &gsprintf($outhandle, "OpenOfficeConverter: {OpenOfficeConverter.noconversionavailable} ({OpenOfficeConverter.$no_openoffice_conversion_reason})\n");
|
|---|
| 158 | }
|
|---|
| 159 |
|
|---|
| 160 | $self->{'openoffice_conversion_available'} = $openoffice_conversion_available;
|
|---|
| 161 |
|
|---|
| 162 | return bless $self, $class;
|
|---|
| 163 |
|
|---|
| 164 | }
|
|---|
| 165 |
|
|---|
| 166 | sub init {
|
|---|
| 167 | my $self = shift(@_);
|
|---|
| 168 | my ($verbosity, $outhandle, $failhandle) = @_;
|
|---|
| 169 |
|
|---|
| 170 | if ($openoffice_conversion_available) {
|
|---|
| 171 | my $oo_port = $self->{'openoffice_port'};
|
|---|
| 172 |
|
|---|
| 173 | my $launch_cmd = "soffice";
|
|---|
| 174 | $launch_cmd .= " \"-accept=socket,host=localhost,port=$oo_port;urp;StarOffice.ServiceManager\"";
|
|---|
| 175 | $launch_cmd .= " -headless";
|
|---|
| 176 | $self->{'openoffice_launch_cmd'} = $launch_cmd;
|
|---|
| 177 | }
|
|---|
| 178 |
|
|---|
| 179 | $self->{'ootmp_file_paths'} = ();
|
|---|
| 180 | }
|
|---|
| 181 |
|
|---|
| 182 | sub deinit {
|
|---|
| 183 | my $self = shift(@_);
|
|---|
| 184 |
|
|---|
| 185 | $self->clean_up_temporary_files();
|
|---|
| 186 | }
|
|---|
| 187 |
|
|---|
| 188 |
|
|---|
| 189 | sub convert {
|
|---|
| 190 | my $self = shift(@_);
|
|---|
| 191 | my $source_file_full_path = shift(@_);
|
|---|
| 192 | my $target_file_type = shift(@_);
|
|---|
| 193 | my $convert_options = shift(@_) || "";
|
|---|
| 194 | my $convert_id = shift(@_) || "";
|
|---|
| 195 | my $cache_mode = shift(@_) || "";
|
|---|
| 196 |
|
|---|
| 197 | return (0,undef,undef) unless $openoffice_conversion_available;
|
|---|
| 198 | # check the filename
|
|---|
| 199 | return (0,undef,undef) if ( !-f $source_file_full_path);
|
|---|
| 200 |
|
|---|
| 201 | my $outhandle = $self->{'outhandle'};
|
|---|
| 202 | my $verbosity = $self->{'verbosity'};
|
|---|
| 203 |
|
|---|
| 204 | my $source_file_no_path = &File::Basename::basename($source_file_full_path);
|
|---|
| 205 | # Determine the full name and path of the output file
|
|---|
| 206 | my $target_file_path;
|
|---|
| 207 | if ($self->{'enable_cache'}) {
|
|---|
| 208 | $self->init_cache_for_file($source_file_full_path);
|
|---|
| 209 | my $cache_dir = $self->{'cached_dir'};
|
|---|
| 210 | my $file_root = $self->{'cached_file_root'};
|
|---|
| 211 | $file_root .= "_$convert_id" if ($convert_id ne "");
|
|---|
| 212 | my $target_file = "$file_root.$target_file_type";
|
|---|
| 213 | $target_file_path = &util::filename_cat($cache_dir,$target_file);
|
|---|
| 214 | }
|
|---|
| 215 | else {
|
|---|
| 216 | $target_file_path = &util::get_timestamped_tmp_filename_in_collection($source_file_full_path, $target_file_type);
|
|---|
| 217 | push(@{$self->{'ootmp_file_paths'}}, $target_file_path);
|
|---|
| 218 | }
|
|---|
| 219 |
|
|---|
| 220 | # Generate and run the convert command
|
|---|
| 221 |
|
|---|
| 222 | my $gextoo_home = $ENV{'GEXT_OPENOFFICE'};
|
|---|
| 223 | my $jodjar = &util::filename_cat($gextoo_home,"lib","java","jodconverter.jar");
|
|---|
| 224 |
|
|---|
| 225 | # debugging: print out all the env vars, when import's verbosity is high
|
|---|
| 226 | #foreach my $key (sort(keys %ENV)) {
|
|---|
| 227 | # print $outhandle "$key = $ENV{$key}\n";
|
|---|
| 228 | #}
|
|---|
| 229 |
|
|---|
| 230 | # *If* SOFFICE_HOME\program was not added to the PATH, *another* way is to insert:
|
|---|
| 231 | # -Doffice.home=C:\\PROGRA~1\\OPENOF~1.ORG
|
|---|
| 232 | # into the command to launch jodconverter so that, when using the remote GS server
|
|---|
| 233 | # on Windows, jodconverter can find the openoffice installation:
|
|---|
| 234 | my $office_short_path = Win32::GetShortPathName($ENV{'SOFFICE_HOME'});
|
|---|
| 235 | my $convert_cmd = "java -Doffice.home=\"".$office_short_path."\" -jar \"$jodjar\" --port $ENV{'JODCONVERTER_PORT'}";
|
|---|
| 236 |
|
|---|
| 237 | # commented out the following line since jodjar always assumes that soffice
|
|---|
| 238 | # is in C:\Prog Files\OpenOffice org 3 rather than checking for any SOFFICE var
|
|---|
| 239 | # the user may have set. So need to run with the -Doffice.home flag that's set above
|
|---|
| 240 | # rather than run without it as was being done below.
|
|---|
| 241 | # my $convert_cmd = "java -jar \"$jodjar\" --port $ENV{'JODCONVERTER_PORT'}";
|
|---|
| 242 | $convert_cmd .= " \"$source_file_full_path\" \"$target_file_path\"";
|
|---|
| 243 |
|
|---|
| 244 | if ($verbosity>2) {
|
|---|
| 245 | print $outhandle "Convert command: $convert_cmd\n";
|
|---|
| 246 | }
|
|---|
| 247 |
|
|---|
| 248 | my $print_info = { 'message_prefix' => "OpenOffice Conversion",
|
|---|
| 249 | 'message' => "Converting $source_file_no_path to: $target_file_type" };
|
|---|
| 250 | $print_info->{'cache_mode'} = $cache_mode if ($cache_mode ne "");
|
|---|
| 251 |
|
|---|
| 252 | my ($regenerated,$result,$had_error)
|
|---|
| 253 | = $self->autorun_general_cmd($convert_cmd,$source_file_full_path, $target_file_path,$print_info);
|
|---|
| 254 | if ($had_error) {
|
|---|
| 255 | return (0, $result,$target_file_path);
|
|---|
| 256 | }
|
|---|
| 257 | return (1, $result,$target_file_path);
|
|---|
| 258 | }
|
|---|
| 259 |
|
|---|
| 260 |
|
|---|
| 261 | sub convert_without_result {
|
|---|
| 262 | my $self = shift(@_);
|
|---|
| 263 |
|
|---|
| 264 | my $source_file_path = shift(@_);
|
|---|
| 265 | my $target_file_type = shift(@_);
|
|---|
| 266 | my $convert_options = shift(@_) || "";
|
|---|
| 267 | my $convert_id = shift(@_) || "";
|
|---|
| 268 |
|
|---|
| 269 | return $self->convert($source_file_path,$target_file_type,
|
|---|
| 270 | $convert_options,$convert_id,"without_result");
|
|---|
| 271 | }
|
|---|
| 272 |
|
|---|
| 273 |
|
|---|
| 274 | sub tmp_area_convert_fileXX {
|
|---|
| 275 | my $self = shift (@_);
|
|---|
| 276 | my ($output_ext, $input_filename, $textref) = @_;
|
|---|
| 277 |
|
|---|
| 278 | my $outhandle = $self->{'outhandle'};
|
|---|
| 279 | my $convert_to = $self->{'convert_to'};
|
|---|
| 280 | my $failhandle = $self->{'failhandle'};
|
|---|
| 281 | my $convert_to_ext = $self->{'convert_to_ext'};
|
|---|
| 282 |
|
|---|
| 283 | # derive tmp filename from input filename
|
|---|
| 284 | my ($tailname, $dirname, $suffix)
|
|---|
| 285 | = &File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
|
|---|
| 286 | }
|
|---|
| 287 |
|
|---|
| 288 |
|
|---|
| 289 | sub clean_up_temporary_files {
|
|---|
| 290 | my $self = shift(@_);
|
|---|
| 291 |
|
|---|
| 292 | foreach my $ootmp_file_path (@{$self->{'ootmp_file_paths'}}) {
|
|---|
| 293 | if (-e $ootmp_file_path) {
|
|---|
| 294 | &util::rm($ootmp_file_path);
|
|---|
| 295 | }
|
|---|
| 296 | }
|
|---|
| 297 |
|
|---|
| 298 | $self->{'ootmp_file_paths'} = ();
|
|---|
| 299 | }
|
|---|
| 300 |
|
|---|
| 301 |
|
|---|
| 302 |
|
|---|
| 303 | 1;
|
|---|
