1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # UnknownPlug.pm -- Plugin for files you know about but Greenstone doesn't
|
---|
4 | #
|
---|
5 | # A component of the Greenstone digital library software from the New
|
---|
6 | # Zealand Digital Library Project at the University of Waikato, New
|
---|
7 | # Zealand.
|
---|
8 | #
|
---|
9 | # Copyright (C) 2001 Gordon W. Paynter
|
---|
10 | # Copyright (C) 2001 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, but
|
---|
18 | # WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
---|
20 | # 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 |
|
---|
28 | # UnknownPlug - a plugin for unknown files
|
---|
29 |
|
---|
30 | # This is a simple Plugin for importing files in formats that
|
---|
31 | # Greenstone doesn't know anything about. A fictional document will
|
---|
32 | # be created for every such file, and the file itself will be passed
|
---|
33 | # to Greenstone as the "associated file" of the document.
|
---|
34 |
|
---|
35 | # Here's an example where it is useful: I have a collection of
|
---|
36 | # pictures that include a couple of quicktime movie files with names
|
---|
37 | # like DCP_0163.MOV. Rather than write a new plugin for quicktime
|
---|
38 | # movies, I add this line to the collection configuration file:
|
---|
39 |
|
---|
40 | # plugin UnknownPlug -process_exp "*.MOV" -assoc_field "movie"
|
---|
41 |
|
---|
42 | # A document is created for each movie, with the associated movie
|
---|
43 | # file's name in the "movie" metadata field. In the collection's
|
---|
44 | # format strings, I use the {If} macro to output different text for
|
---|
45 | # each type of file, like this:
|
---|
46 |
|
---|
47 | # {If}{[movie],<HTML for displaying movie>}{If}{[Image],<HTML for displaying image>}
|
---|
48 |
|
---|
49 | # You can also add extra metadata, such as the Title, Subject, and
|
---|
50 | # Duration, with metadata.xml files and RecPlug. (If you want to use
|
---|
51 | # UnknownPlug with more than one type of file, you will have to add
|
---|
52 | # some sort of distinguishing metadata in this way.)
|
---|
53 |
|
---|
54 |
|
---|
55 |
|
---|
56 | package UnknownPlug;
|
---|
57 |
|
---|
58 | use BasPlug;
|
---|
59 |
|
---|
60 | use strict;
|
---|
61 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
62 |
|
---|
63 | sub BEGIN {
|
---|
64 | @UnknownPlug::ISA = ('BasPlug');
|
---|
65 | }
|
---|
66 |
|
---|
67 | my $arguments =
|
---|
68 | [ { 'name' => "assoc_field",
|
---|
69 | 'desc' => "{UnknownPlug.assoc_field}",
|
---|
70 | 'type' => "string",
|
---|
71 | 'deft' => "",
|
---|
72 | 'reqd' => "no" },
|
---|
73 | { 'name' => "file_format",
|
---|
74 | 'desc' => "{UnknownPlug.file_format}",
|
---|
75 | 'type' => "string",
|
---|
76 | 'deft' => "",
|
---|
77 | 'reqd' => "no" },
|
---|
78 | { 'name' => "mime_type",
|
---|
79 | 'desc' => "{UnknownPlug.mime_type}",
|
---|
80 | 'type' => "string",
|
---|
81 | 'deft' => "",
|
---|
82 | 'reqd' => "no" },
|
---|
83 | { 'name' => "srcicon",
|
---|
84 | 'desc' => "{UnknownPlug.srcicon}",
|
---|
85 | 'type' => "string",
|
---|
86 | 'deft' => "iconunknown",
|
---|
87 | 'reqd' => "no" },
|
---|
88 | { 'name' => "process_extension",
|
---|
89 | 'desc' => "{UnknownPlug.process_extension}",
|
---|
90 | 'type' => "string",
|
---|
91 | 'deft' => "",
|
---|
92 | 'reqd' => "no" } ];
|
---|
93 |
|
---|
94 | my $options = { 'name' => "UnknownPlug",
|
---|
95 | 'desc' => "{UnknownPlug.desc}",
|
---|
96 | 'abstract' => "no",
|
---|
97 | 'inherits' => "yes",
|
---|
98 | 'args' => $arguments };
|
---|
99 |
|
---|
100 |
|
---|
101 | sub new {
|
---|
102 | my ($class) = shift (@_);
|
---|
103 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
---|
104 | push(@$pluginlist, $class);
|
---|
105 |
|
---|
106 | if(defined $arguments){ push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});}
|
---|
107 | if(defined $options) { push(@{$hashArgOptLists->{"OptList"}},$options)};
|
---|
108 |
|
---|
109 | my $self = new BasPlug($pluginlist, $inputargs, $hashArgOptLists);
|
---|
110 |
|
---|
111 | # "-process_extension" is a simpler alternative to -process_exp for non-regexp people
|
---|
112 | if (!$self->{'process_exp'} && $self->{'process_extension'}) {
|
---|
113 | $self->{'process_exp'} = "\\." . $self->{'process_extension'} . "\$";
|
---|
114 | }
|
---|
115 |
|
---|
116 | return bless $self, $class;
|
---|
117 | }
|
---|
118 |
|
---|
119 | sub get_default_process_exp {
|
---|
120 | return '';
|
---|
121 | }
|
---|
122 |
|
---|
123 |
|
---|
124 | # Associate the unknown file with the new document
|
---|
125 |
|
---|
126 | sub associate_unknown_file {
|
---|
127 | my $self = shift (@_);
|
---|
128 | my $filename = shift (@_); # filename with full path
|
---|
129 | my $file = shift (@_); # filename without path
|
---|
130 | my $doc_obj = shift (@_);
|
---|
131 |
|
---|
132 | my $verbosity = $self->{'verbosity'};
|
---|
133 | my $outhandle = $self->{'outhandle'};
|
---|
134 |
|
---|
135 | # check the filename is okay
|
---|
136 | return 0 if ($file eq "" || $filename eq "");
|
---|
137 |
|
---|
138 | # Add the image metadata
|
---|
139 | my $url = $file;
|
---|
140 | $url =~ s/ /%20/g;
|
---|
141 |
|
---|
142 | # Add the file as an associated file ...
|
---|
143 | my $section = $doc_obj->get_top_section();
|
---|
144 | my $file_format = $self->{'file_format'} || "unknown";
|
---|
145 | my $mime_type = $self->{'mime_type'} || "unknown/unknown";
|
---|
146 | my $assoc_field = $self->{'assoc_field'} || "unknown_file";
|
---|
147 |
|
---|
148 | $doc_obj->associate_file($filename, $file, $mime_type, $section);
|
---|
149 | $doc_obj->add_metadata ($section, "FileFormat", $file_format);
|
---|
150 | $doc_obj->add_metadata ($section, "MimeType", $mime_type);
|
---|
151 | $doc_obj->add_metadata ($section, $assoc_field, $file);
|
---|
152 |
|
---|
153 | $doc_obj->add_metadata ($section, "srclink",
|
---|
154 | "<a href=\"_httpprefix_/collect/[collection]/index/assoc/[assocfilepath]/[$assoc_field]\">");
|
---|
155 | #$doc_obj->add_metadata ($section, "srcicon", "_iconunknown_");
|
---|
156 | $doc_obj->add_metadata ($section, "srcicon", "_".$self->{'srcicon'}."_");
|
---|
157 | $doc_obj->add_metadata ($section, "/srclink", "</a>");
|
---|
158 |
|
---|
159 | return 1;
|
---|
160 | }
|
---|
161 |
|
---|
162 |
|
---|
163 |
|
---|
164 | # The UnknownPlug read() function. This function does all the right
|
---|
165 | # things to make general options work for a given plugin. UnknownPlug
|
---|
166 | # overrides read() because there is no need to read the actual text of
|
---|
167 | # the file in, because the contents of the file is not text...
|
---|
168 | #
|
---|
169 | #
|
---|
170 | # Return number of files processed, undef if can't process
|
---|
171 | #
|
---|
172 | # Note that $base_dir might be "" and that $file might include directories
|
---|
173 |
|
---|
174 | sub read {
|
---|
175 | my $self = shift (@_);
|
---|
176 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
|
---|
177 |
|
---|
178 | my $outhandle = $self->{'outhandle'};
|
---|
179 |
|
---|
180 | # Make sure we're processing the correct file
|
---|
181 | my ($block_status,$filename) = $self->read_block(@_);
|
---|
182 | return $block_status if ((!defined $block_status) || ($block_status==0));
|
---|
183 |
|
---|
184 | print STDERR "<Processing n='$file' p='UnknownPlug'>\n" if ($gli);
|
---|
185 | print $outhandle "UnknownPlug processing \"$filename\"\n"
|
---|
186 | if $self->{'verbosity'} > 1;
|
---|
187 |
|
---|
188 | #if there's a leading directory name, eat it...
|
---|
189 | $file =~ s/^.*[\/\\]//;
|
---|
190 |
|
---|
191 | # create a new document
|
---|
192 | my $doc_obj = new doc ($filename, "indexed_doc");
|
---|
193 | $doc_obj->set_OIDtype ($processor->{'OIDtype'}, $processor->{'OIDmetadata'});
|
---|
194 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
|
---|
195 | $doc_obj->add_metadata($doc_obj->get_top_section(), "Source", &ghtml::dmsafe($file)); # set the filename as Source metadata to be consistent with other plugins
|
---|
196 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "FileSize", (-s $filename));
|
---|
197 |
|
---|
198 | # URL metadata (even invalid ones) are used to support internal
|
---|
199 | # links, so even if 'file_is_url' is off, still need to store info
|
---|
200 |
|
---|
201 | my $web_url = "http://$file";
|
---|
202 | $doc_obj->add_metadata($doc_obj->get_top_section(), "URL", $web_url);
|
---|
203 |
|
---|
204 |
|
---|
205 | # associate the file with the document
|
---|
206 | if (associate_unknown_file($self, $filename, $file, $doc_obj) != 1)
|
---|
207 | {
|
---|
208 | if ($gli) {
|
---|
209 | print STDERR "<ProcessingError n='$file'>\n";
|
---|
210 | }
|
---|
211 | print $outhandle "UnknownPlug: couldn't process \"$filename\"\n";
|
---|
212 | return -1; # error during processing
|
---|
213 | }
|
---|
214 |
|
---|
215 | #create an empty text string so we don't break downstream plugins
|
---|
216 | my $text = &gsprintf::lookup_string("{BasPlug.dummy_text}");
|
---|
217 |
|
---|
218 | # include any metadata passed in from previous plugins
|
---|
219 | my $section = $doc_obj->get_top_section();
|
---|
220 | $self->extra_metadata ($doc_obj, $section, $metadata);
|
---|
221 |
|
---|
222 | $self->title_fallback($doc_obj,$section,$file);
|
---|
223 |
|
---|
224 | # do plugin specific processing of doc_obj
|
---|
225 | unless (defined ($self->process(\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj))) {
|
---|
226 | print STDERR "<ProcessingError n='$file'>\n" if ($gli);
|
---|
227 | return -1;
|
---|
228 | }
|
---|
229 |
|
---|
230 | # do any automatic metadata extraction
|
---|
231 | $self->auto_extract_metadata ($doc_obj);
|
---|
232 |
|
---|
233 | # add an OID
|
---|
234 | $doc_obj->set_OID();
|
---|
235 | $doc_obj->add_text($section, $text);
|
---|
236 |
|
---|
237 | # process the document
|
---|
238 | $processor->process($doc_obj);
|
---|
239 |
|
---|
240 | $self->{'num_processed'} ++;
|
---|
241 | return 1;
|
---|
242 | }
|
---|
243 |
|
---|
244 |
|
---|
245 | # UnknownPlug processing of doc_obj. In practice we don't need to do
|
---|
246 | # anything here because the read function takes care of everything.
|
---|
247 |
|
---|
248 | sub process {
|
---|
249 | my $self = shift (@_);
|
---|
250 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
---|
251 | my $outhandle = $self->{'outhandle'};
|
---|
252 |
|
---|
253 | return 1;
|
---|
254 | }
|
---|
255 |
|
---|
256 |
|
---|
257 | 1;
|
---|
258 |
|
---|
259 |
|
---|
260 |
|
---|
261 |
|
---|
262 |
|
---|
263 |
|
---|
264 |
|
---|
265 |
|
---|
266 |
|
---|
267 |
|
---|
268 |
|
---|