1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # PowerPointPlugin.pm -- plugin for importing Microsoft PowerPoint files.
|
---|
4 | # (basic version supports versions 95 and 97)
|
---|
5 | # (through OpenOffice extension, supports all contemporary formats)
|
---|
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) 2002 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 PowerPointPlugin;
|
---|
30 |
|
---|
31 | use strict;
|
---|
32 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
33 | no strict 'subs';
|
---|
34 |
|
---|
35 | use gsprintf 'gsprintf';
|
---|
36 |
|
---|
37 | use AutoLoadConverters;
|
---|
38 | use ConvertBinaryFile;
|
---|
39 |
|
---|
40 | sub BEGIN {
|
---|
41 | @PowerPointPlugin::ISA = ('ConvertBinaryFile', 'AutoLoadConverters');
|
---|
42 | }
|
---|
43 |
|
---|
44 | my $openoffice_available = 0;
|
---|
45 |
|
---|
46 | my $windows_convert_to_list =
|
---|
47 | [ { 'name' => "auto",
|
---|
48 | 'desc' => "{ConvertBinaryFile.convert_to.auto}" },
|
---|
49 | { 'name' => "html",
|
---|
50 | 'desc' => "{ConvertBinaryFile.convert_to.html}" },
|
---|
51 | { 'name' => "text",
|
---|
52 | 'desc' => "{ConvertBinaryFile.convert_to.text}" },
|
---|
53 | { 'name' => "pagedimg_jpg",
|
---|
54 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_jpg}" },
|
---|
55 | { 'name' => "pagedimg_gif",
|
---|
56 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_gif}" },
|
---|
57 | { 'name' => "pagedimg_png",
|
---|
58 | 'desc' => "{ConvertBinaryFile.convert_to.pagedimg_png}" }
|
---|
59 | ];
|
---|
60 |
|
---|
61 | my $openoffice_convert_to_list =
|
---|
62 | [ { 'name' => "auto",
|
---|
63 | 'desc' => "{ConvertBinaryFile.convert_to.auto}" },
|
---|
64 | { 'name' => "html",
|
---|
65 | 'desc' => "{PowerPointPlugin.convert_to.oo_html}" },
|
---|
66 | { 'name' => "text",
|
---|
67 | 'desc' => "{ConvertBinaryFile.convert_to.text}" },
|
---|
68 | { 'name' => "pagedimg",
|
---|
69 | 'desc' => "{PowerPointPlugin.convert_to.pagedimg}" }
|
---|
70 | ];
|
---|
71 |
|
---|
72 | my $arguments =
|
---|
73 | [ { 'name' => "process_exp",
|
---|
74 | 'desc' => "{BasePlugin.process_exp}",
|
---|
75 | 'type' => "regexp",
|
---|
76 | 'reqd' => "no",
|
---|
77 | 'deft' => "&get_default_process_exp()", # delayed (see below)
|
---|
78 | }
|
---|
79 | ];
|
---|
80 |
|
---|
81 | my $opt_windows_args =
|
---|
82 | [ { 'name' => "convert_to",
|
---|
83 | 'desc' => "{ConvertBinaryFile.convert_to}",
|
---|
84 | 'type' => "enum",
|
---|
85 | 'reqd' => "yes",
|
---|
86 | 'list' => $windows_convert_to_list,
|
---|
87 | 'deft' => "html" },
|
---|
88 | { 'name' => "windows_scripting",
|
---|
89 | 'desc' => "{PowerPointPlugin.windows_scripting}",
|
---|
90 | 'type' => "flag",
|
---|
91 | 'reqd' => "no" }
|
---|
92 | ];
|
---|
93 |
|
---|
94 | my $opt_office_args =
|
---|
95 | [ { 'name' => "convert_to",
|
---|
96 | 'desc' => "{ConvertBinaryFile.convert_to}",
|
---|
97 | 'type' => "enum",
|
---|
98 | 'reqd' => "yes",
|
---|
99 | 'list' => $openoffice_convert_to_list,
|
---|
100 | 'deft' => "html" }
|
---|
101 | ];
|
---|
102 |
|
---|
103 | my $options = { 'name' => "PowerPointPlugin",
|
---|
104 | 'desc' => "{PowerPointPlugin.desc}",
|
---|
105 | 'abstract' => "no",
|
---|
106 | 'inherits' => "yes",
|
---|
107 | 'srcreplaceable' => "yes", # Source docs in PPT format can be replaced with GS-generated html
|
---|
108 | 'args' => $arguments };
|
---|
109 |
|
---|
110 | sub new {
|
---|
111 | my ($class) = shift (@_);
|
---|
112 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
---|
113 | push(@$pluginlist, $class);
|
---|
114 |
|
---|
115 | # this bit needs to happen later after the arguments array has been
|
---|
116 | # finished - used for parsing the input args.
|
---|
117 | # push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
118 | # this one needs to go in first, to get the print info in the right order
|
---|
119 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
---|
120 |
|
---|
121 | if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
|
---|
122 | push(@$arguments,@$opt_windows_args);
|
---|
123 | }
|
---|
124 |
|
---|
125 | my $auto_converter_self = new AutoLoadConverters($pluginlist,$inputargs,$hashArgOptLists,["OpenOfficeConverter"],1);
|
---|
126 |
|
---|
127 | if ($auto_converter_self->{'openoffice_available'}) {
|
---|
128 | push (@$arguments,@$opt_office_args);
|
---|
129 | $openoffice_available = 1;
|
---|
130 | }
|
---|
131 | # TODO need to do the case where they are both enabled!!! what will the convert to list be???
|
---|
132 |
|
---|
133 | # evaluate the default for process_exp - it needs to be delayed till here so we know if openoffice is available or not. But needs to be done before parsing the args.
|
---|
134 | foreach my $a (@$arguments) {
|
---|
135 | if ($a->{'name'} eq "process_exp") {
|
---|
136 | my $eval_expr = $a->{'deft'};
|
---|
137 | $a->{'deft'} = eval "$eval_expr";
|
---|
138 | last;
|
---|
139 | }
|
---|
140 | }
|
---|
141 |
|
---|
142 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
143 |
|
---|
144 | my $cbf_self = new ConvertBinaryFile($pluginlist, $inputargs, $hashArgOptLists);
|
---|
145 | my $self = BasePlugin::merge_inheritance($auto_converter_self, $cbf_self);
|
---|
146 |
|
---|
147 | if ($self->{'info_only'}) {
|
---|
148 | # don't worry about any options etc
|
---|
149 | return bless $self, $class;
|
---|
150 | }
|
---|
151 |
|
---|
152 | $self = bless $self, $class;
|
---|
153 | $self->{'filename_extension'} = "ppt";
|
---|
154 | $self->{'file_type'} = "PPT";
|
---|
155 |
|
---|
156 | if ($self->{'convert_to'} eq "auto") {
|
---|
157 | if ($self->{'windows_scripting'}) {
|
---|
158 | $self->{'convert_to'} = "pagedimg_jpg";
|
---|
159 | }
|
---|
160 | else {
|
---|
161 | $self->{'convert_to'} = "html";
|
---|
162 | }
|
---|
163 | }
|
---|
164 |
|
---|
165 | my $outhandle = $self->{'outhandle'};
|
---|
166 |
|
---|
167 | # can't have windows_scripting and openoffice_conversion at the same time
|
---|
168 | if ($self->{'windows_scripting'} && $self->{'openoffice_conversion'}) {
|
---|
169 | print $outhandle "Warning: Cannot have -windows_scripting and -openoffice_conversion\n";
|
---|
170 | print $outhandle " on at the same time. Defaulting to -windows_scripting\n";
|
---|
171 | $self->{'openoffice_conversion'} = 0;
|
---|
172 | }
|
---|
173 |
|
---|
174 | #these are passed through to gsConvert.pl by ConvertBinaryFile.pm
|
---|
175 | $self->{'convert_options'} = "-windows_scripting" if $self->{'windows_scripting'};
|
---|
176 |
|
---|
177 | # set convert_to_plugin and convert_to_ext
|
---|
178 | $self->set_standard_convert_settings();
|
---|
179 |
|
---|
180 | my $secondary_plugin_name = $self->{'convert_to_plugin'};
|
---|
181 | my $secondary_plugin_options = $self->{'secondary_plugin_options'};
|
---|
182 |
|
---|
183 | if (!defined $secondary_plugin_options->{$secondary_plugin_name}) {
|
---|
184 | $secondary_plugin_options->{$secondary_plugin_name} = [];
|
---|
185 | }
|
---|
186 | my $specific_options = $secondary_plugin_options->{$secondary_plugin_name};
|
---|
187 |
|
---|
188 | push(@$specific_options, "-file_rename_method", "none");
|
---|
189 | push(@$specific_options, "-extract_language") if $self->{'extract_language'};
|
---|
190 |
|
---|
191 | if ($secondary_plugin_name eq "HTMLPlugin") {
|
---|
192 | push(@$specific_options, "-processing_tmp_files");
|
---|
193 | push(@$specific_options,"-metadata_fields","Title,GENERATOR,date,author<Creator>");
|
---|
194 | }
|
---|
195 | elsif ($secondary_plugin_name eq "PagedImagePlugin") {
|
---|
196 | push(@$specific_options, "-processing_tmp_files");
|
---|
197 | #is this true??
|
---|
198 | push(@$specific_options,"-input_encoding", "utf8");
|
---|
199 | }
|
---|
200 |
|
---|
201 | $self = bless $self, $class;
|
---|
202 |
|
---|
203 | $self->load_secondary_plugins($class,$secondary_plugin_options,$hashArgOptLists);
|
---|
204 | return $self;
|
---|
205 | }
|
---|
206 |
|
---|
207 | sub get_default_process_exp {
|
---|
208 | my $self = shift (@_);
|
---|
209 |
|
---|
210 | if ($openoffice_available) {
|
---|
211 | return q^(?i)\.(ppt|pptx|odp)$^;
|
---|
212 | }
|
---|
213 |
|
---|
214 | return q^(?i)\.ppt$^;
|
---|
215 | }
|
---|
216 |
|
---|
217 | sub init {
|
---|
218 | my $self = shift (@_);
|
---|
219 |
|
---|
220 | # ConvertBinaryFile init
|
---|
221 | $self->SUPER::init(@_);
|
---|
222 | $self->AutoLoadConverters::init();
|
---|
223 |
|
---|
224 | }
|
---|
225 |
|
---|
226 | sub begin {
|
---|
227 | my $self = shift (@_);
|
---|
228 |
|
---|
229 | $self->AutoLoadConverters::begin();
|
---|
230 | $self->SUPER::begin(@_);
|
---|
231 |
|
---|
232 | }
|
---|
233 |
|
---|
234 | sub deinit {
|
---|
235 | my $self = shift (@_);
|
---|
236 |
|
---|
237 | $self->AutoLoadConverters::deinit();
|
---|
238 | $self->SUPER::deinit(@_);
|
---|
239 |
|
---|
240 | }
|
---|
241 |
|
---|
242 | # override AutoLoadConverters version, as we need to do more stuff once its converted if we are converting to item file
|
---|
243 | sub tmp_area_convert_file {
|
---|
244 | my $self = shift (@_);
|
---|
245 | my ($output_ext, $input_filename, $textref) = @_;
|
---|
246 |
|
---|
247 | if ($self->{'openoffice_conversion'}) {
|
---|
248 | if ($self->{'convert_to'} eq "pagedimg") {
|
---|
249 | $output_ext = "html"; # first convert to html
|
---|
250 | }
|
---|
251 | my ($result, $result_str, $new_filename) = $self->OpenOfficeConverter::convert($input_filename, $output_ext);
|
---|
252 | if ($result == 0) {
|
---|
253 | my $outhandle=$self->{'outhandle'};
|
---|
254 | print $outhandle "OpenOfficeConverter Conversion error\n";
|
---|
255 | print $outhandle $result_str;
|
---|
256 | return "";
|
---|
257 |
|
---|
258 | }
|
---|
259 | #print STDERR "result = $result\n";
|
---|
260 | if ($self->{'convert_to'} eq "pagedimg") {
|
---|
261 | #my $item_filename = $self->generate_item_file($new_filename);
|
---|
262 | #return $item_filename;
|
---|
263 | return "/research/kjdon/home/gsdl/collect/openoffice/test.item";
|
---|
264 | }
|
---|
265 | return $new_filename;
|
---|
266 |
|
---|
267 | }
|
---|
268 | else {
|
---|
269 | return $self->ConvertBinaryFile::tmp_area_convert_file(@_);
|
---|
270 | }
|
---|
271 | # get tmp filename
|
---|
272 | }
|
---|
273 |
|
---|
274 | # override default read in some situations, as the conversion of ppt to html results in many files, and we want them all to be processed.
|
---|
275 | sub read_XX {
|
---|
276 | my $self = shift (@_);
|
---|
277 | my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
|
---|
278 |
|
---|
279 | # can we process this file??
|
---|
280 | my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
|
---|
281 |
|
---|
282 | return undef unless $self->can_process_this_file($filename_full_path);
|
---|
283 |
|
---|
284 | my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
|
---|
285 |
|
---|
286 | if ((defined $process_status) && ($process_status == 1)) {
|
---|
287 |
|
---|
288 | # process the document
|
---|
289 | $processor->process($doc_obj);
|
---|
290 |
|
---|
291 | $self->{'num_processed'} ++;
|
---|
292 | undef $doc_obj;
|
---|
293 | }
|
---|
294 | # delete any temp files that we may have created
|
---|
295 | $self->clean_up_after_doc_obj_processing();
|
---|
296 |
|
---|
297 |
|
---|
298 | # if process_status == 1, then the file has been processed.
|
---|
299 | return $process_status;
|
---|
300 |
|
---|
301 | }
|
---|
302 |
|
---|
303 | 1;
|
---|
304 |
|
---|