| 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 | use parsargv;
|
|---|
| 60 |
|
|---|
| 61 |
|
|---|
| 62 | sub BEGIN {
|
|---|
| 63 | @UnknownPlug::ISA = ('BasPlug');
|
|---|
| 64 | }
|
|---|
| 65 |
|
|---|
| 66 | my $arguments =
|
|---|
| 67 | [ { 'name' => "assoc_field",
|
|---|
| 68 | 'desc' => "{UnknownPlug.assoc_field}",
|
|---|
| 69 | 'type' => "string",
|
|---|
| 70 | 'deft' => "",
|
|---|
| 71 | 'reqd' => "no" },
|
|---|
| 72 | { 'name' => "file_format",
|
|---|
| 73 | 'desc' => "{UnknownPlug.file_format}",
|
|---|
| 74 | 'type' => "string",
|
|---|
| 75 | 'deft' => "",
|
|---|
| 76 | 'reqd' => "no" },
|
|---|
| 77 | { 'name' => "mime_type",
|
|---|
| 78 | 'desc' => "{UnknownPlug.mime_type}",
|
|---|
| 79 | 'type' => "string",
|
|---|
| 80 | 'deft' => "",
|
|---|
| 81 | 'reqd' => "no" },
|
|---|
| 82 | { 'name' => "process_extension",
|
|---|
| 83 | 'desc' => "{UnknownPlug.process_extension}",
|
|---|
| 84 | 'type' => "string",
|
|---|
| 85 | 'deft' => "",
|
|---|
| 86 | 'reqd' => "no" } ];
|
|---|
| 87 |
|
|---|
| 88 | my $options = { 'name' => "UnknownPlug",
|
|---|
| 89 | 'desc' => "{UnknownPlug.desc}",
|
|---|
| 90 | 'abstract' => "no",
|
|---|
| 91 | 'inherits' => "yes",
|
|---|
| 92 | 'args' => $arguments };
|
|---|
| 93 |
|
|---|
| 94 |
|
|---|
| 95 | sub new {
|
|---|
| 96 | my ($class) = @_;
|
|---|
| 97 | my $self = new BasPlug ($class, @_);
|
|---|
| 98 | $self->{'plugin_type'} = "UnknownPlug";
|
|---|
| 99 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
|---|
| 100 | my $option_list = $self->{'option_list'};
|
|---|
| 101 | push( @{$option_list}, $options );
|
|---|
| 102 |
|
|---|
| 103 | if (!parsargv::parse(\@_,
|
|---|
| 104 | q^assoc_field/.*/^, \$self->{'assoc_field'},
|
|---|
| 105 | q^file_format/.*/^, \$self->{'file_format'},
|
|---|
| 106 | q^mime_type/.*/^, \$self->{'mime_type'},
|
|---|
| 107 | q^process_extension/.*/^, \$self->{'process_extension'},
|
|---|
| 108 | "allow_extra_options")) {
|
|---|
| 109 | print STDERR "\nIncorrect options passed to UnknownPlug, check your collect.cfg configuration file\n";
|
|---|
| 110 | $self->print_txt_usage(""); # Use default resource bundle
|
|---|
| 111 | die "\n";
|
|---|
| 112 | }
|
|---|
| 113 |
|
|---|
| 114 | # "-process_extension" is a simpler alternative to -process_exp for non-regexp people
|
|---|
| 115 | if (!$self->{'process_exp'} && $self->{'process_extension'}) {
|
|---|
| 116 | $self->{'process_exp'} = "\\." . $self->{'process_extension'} . "\$";
|
|---|
| 117 | }
|
|---|
| 118 |
|
|---|
| 119 | return bless $self, $class;
|
|---|
| 120 | }
|
|---|
| 121 |
|
|---|
| 122 | sub get_default_process_exp {
|
|---|
| 123 | return '';
|
|---|
| 124 | }
|
|---|
| 125 |
|
|---|
| 126 |
|
|---|
| 127 | # Associate the unknown file with the new document
|
|---|
| 128 |
|
|---|
| 129 | sub associate_unknown_file {
|
|---|
| 130 | my $self = shift (@_);
|
|---|
| 131 | my $filename = shift (@_); # filename with full path
|
|---|
| 132 | my $file = shift (@_); # filename without path
|
|---|
| 133 | my $doc_obj = shift (@_);
|
|---|
| 134 |
|
|---|
| 135 | my $verbosity = $self->{'verbosity'};
|
|---|
| 136 | my $outhandle = $self->{'outhandle'};
|
|---|
| 137 |
|
|---|
| 138 | # check the filename is okay
|
|---|
| 139 | return 0 if ($file eq "" || $filename eq "");
|
|---|
| 140 |
|
|---|
| 141 | # Add the image metadata
|
|---|
| 142 | my $url = $file;
|
|---|
| 143 | $url =~ s/ /%20/g;
|
|---|
| 144 |
|
|---|
| 145 | # Add the file as an associated file ...
|
|---|
| 146 | my $section = $doc_obj->get_top_section();
|
|---|
| 147 | my $file_format = $self->{'file_format'} || "unknown";
|
|---|
| 148 | my $mime_type = $self->{'mime_type'} || "unknown/unknown";
|
|---|
| 149 | my $assoc_field = $self->{'assoc_field'} || "unknown_file";
|
|---|
| 150 |
|
|---|
| 151 | $doc_obj->associate_file($filename, $file, $mime_type, $section);
|
|---|
| 152 | $doc_obj->add_metadata ($section, "FileFormat", $file_format);
|
|---|
| 153 | $doc_obj->add_metadata ($section, $assoc_field, $file);
|
|---|
| 154 |
|
|---|
| 155 | $doc_obj->add_metadata ($section, "srclink",
|
|---|
| 156 | "<a href=\"_httpcollection_/index/assoc/[assocfilepath]/[$assoc_field]\">");
|
|---|
| 157 | $doc_obj->add_metadata ($section, "srcicon", "_iconunknown_");
|
|---|
| 158 | $doc_obj->add_metadata ($section, "/srclink", "</a>");
|
|---|
| 159 |
|
|---|
| 160 | return 1;
|
|---|
| 161 | }
|
|---|
| 162 |
|
|---|
| 163 |
|
|---|
| 164 |
|
|---|
| 165 | # The UnknownPlug read() function. This function does all the right
|
|---|
| 166 | # things to make general options work for a given plugin. UnknownPlug
|
|---|
| 167 | # overrides read() because there is no need to read the actual text of
|
|---|
| 168 | # the file in, because the contents of the file is not text...
|
|---|
| 169 | #
|
|---|
| 170 | #
|
|---|
| 171 | # Return number of files processed, undef if can't process
|
|---|
| 172 | #
|
|---|
| 173 | # Note that $base_dir might be "" and that $file might include directories
|
|---|
| 174 |
|
|---|
| 175 | sub read {
|
|---|
| 176 | my $self = shift (@_);
|
|---|
| 177 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
|
|---|
| 178 |
|
|---|
| 179 | my $outhandle = $self->{'outhandle'};
|
|---|
| 180 |
|
|---|
| 181 | # Make sure we're processing the correct file
|
|---|
| 182 | my $filename = &util::filename_cat($base_dir, $file);
|
|---|
| 183 | return 0 if $self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/;
|
|---|
| 184 | if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
|
|---|
| 185 | return undef;
|
|---|
| 186 | }
|
|---|
| 187 | print STDERR "<Processing n='$file' p='UnknownPlug'>\n" if ($gli);
|
|---|
| 188 | print $outhandle "UnknownPlug processing \"$filename\"\n"
|
|---|
| 189 | if $self->{'verbosity'} > 1;
|
|---|
| 190 |
|
|---|
| 191 | #if there's a leading directory name, eat it...
|
|---|
| 192 | $file =~ s/^.*[\/\\]//;
|
|---|
| 193 |
|
|---|
| 194 | # create a new document
|
|---|
| 195 | my $doc_obj = new doc ($filename, "indexed_doc");
|
|---|
| 196 | $doc_obj->set_OIDtype ($processor->{'OIDtype'});
|
|---|
| 197 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
|
|---|
| 198 | $doc_obj->add_metadata($doc_obj->get_top_section(), "Source", &ghtml::dmsafe($file)); # set the filename as Source metadata to be consistent with other plugins
|
|---|
| 199 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "FileSize", (-s $filename));
|
|---|
| 200 |
|
|---|
| 201 | # associate the file with the document
|
|---|
| 202 | if (associate_unknown_file($self, $filename, $file, $doc_obj) != 1)
|
|---|
| 203 | {
|
|---|
| 204 | if ($gli) {
|
|---|
| 205 | print STDERR "<ProcessingError n='$file'>\n";
|
|---|
| 206 | }
|
|---|
| 207 | print $outhandle "UnknownPlug: couldn't process \"$filename\"\n";
|
|---|
| 208 | return -1; # error during processing
|
|---|
| 209 | }
|
|---|
| 210 |
|
|---|
| 211 | #create an empty text string so we don't break downstream plugins
|
|---|
| 212 | my $text = &gsprintf::lookup_string("{BasPlug.dummy_text}");
|
|---|
| 213 |
|
|---|
| 214 | # include any metadata passed in from previous plugins
|
|---|
| 215 | my $section = $doc_obj->get_top_section();
|
|---|
| 216 | $self->extra_metadata ($doc_obj, $section, $metadata);
|
|---|
| 217 |
|
|---|
| 218 | $self->title_fallback($doc_obj,$section,$file);
|
|---|
| 219 |
|
|---|
| 220 | # do plugin specific processing of doc_obj
|
|---|
| 221 | unless (defined ($self->process(\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj))) {
|
|---|
| 222 | print STDERR "<ProcessingError n='$file'>\n" if ($gli);
|
|---|
| 223 | return -1;
|
|---|
| 224 | }
|
|---|
| 225 |
|
|---|
| 226 | # do any automatic metadata extraction
|
|---|
| 227 | $self->auto_extract_metadata ($doc_obj);
|
|---|
| 228 |
|
|---|
| 229 | # add an OID
|
|---|
| 230 | $doc_obj->set_OID();
|
|---|
| 231 | $doc_obj->add_text($section, $text);
|
|---|
| 232 |
|
|---|
| 233 | # process the document
|
|---|
| 234 | $processor->process($doc_obj);
|
|---|
| 235 |
|
|---|
| 236 | $self->{'num_processed'} ++;
|
|---|
| 237 | return 1;
|
|---|
| 238 | }
|
|---|
| 239 |
|
|---|
| 240 |
|
|---|
| 241 | # UnknownPlug processing of doc_obj. In practice we don't need to do
|
|---|
| 242 | # anything here because the read function takes care of everything.
|
|---|
| 243 |
|
|---|
| 244 | sub process {
|
|---|
| 245 | my $self = shift (@_);
|
|---|
| 246 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
|---|
| 247 | my $outhandle = $self->{'outhandle'};
|
|---|
| 248 |
|
|---|
| 249 | return 1;
|
|---|
| 250 | }
|
|---|
| 251 |
|
|---|
| 252 |
|
|---|
| 253 | 1;
|
|---|
| 254 |
|
|---|
| 255 |
|
|---|
| 256 |
|
|---|
| 257 |
|
|---|
| 258 |
|
|---|
| 259 |
|
|---|
| 260 |
|
|---|
| 261 |
|
|---|
| 262 |
|
|---|
| 263 |
|
|---|
| 264 |
|
|---|
