| 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 | @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_type",
|
|---|
| 73 | 'desc' => "{UnknownPlug.file_type}",
|
|---|
| 74 | 'type' => "string",
|
|---|
| 75 | 'deft' => "",
|
|---|
| 76 | 'reqd' => "no" } ];
|
|---|
| 77 |
|
|---|
| 78 | my $options = { 'name' => "UnknownPlug",
|
|---|
| 79 | 'desc' => "{UnknownPlug.desc}",
|
|---|
| 80 | 'inherits' => "yes",
|
|---|
| 81 | 'args' => $arguments };
|
|---|
| 82 |
|
|---|
| 83 | # sub print_usage {
|
|---|
| 84 | # print STDERR "\n usage: plugin UnknownPlug [options]\n\n";
|
|---|
| 85 | # print STDERR " options:\n";
|
|---|
| 86 | # print STDERR " -assoc_field\t Name of the metadata field that will hold\n";
|
|---|
| 87 | # print STDERR "\t\t the associated file's name.\n";
|
|---|
| 88 | # print STDERR "\n";
|
|---|
| 89 | # print STDERR " -file_type\t Mime type of the file (e.g. image/gif)\n";
|
|---|
| 90 | # print STDERR "\n";
|
|---|
| 91 | # }
|
|---|
| 92 |
|
|---|
| 93 | sub new {
|
|---|
| 94 | my ($class) = @_;
|
|---|
| 95 | my $self = new BasPlug ($class, @_);
|
|---|
| 96 |
|
|---|
| 97 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
|---|
| 98 | my $option_list = $self->{'option_list'};
|
|---|
| 99 | push( @{$option_list}, $options );
|
|---|
| 100 |
|
|---|
| 101 | if (!parsargv::parse(\@_,
|
|---|
| 102 | q^assoc_field/.*/^, \$self->{'assoc_field'},
|
|---|
| 103 | q^file_type/.*/^, \$self->{'file_type'},
|
|---|
| 104 | "allow_extra_options")) {
|
|---|
| 105 | print STDERR "\nIncorrect options passed to UnknownPlug, check your collect.cfg configuration file\n";
|
|---|
| 106 | $self->print_txt_usage(""); # Use default resource bundle
|
|---|
| 107 | die "\n";
|
|---|
| 108 | }
|
|---|
| 109 |
|
|---|
| 110 |
|
|---|
| 111 | return bless $self, $class;
|
|---|
| 112 | }
|
|---|
| 113 |
|
|---|
| 114 | sub get_default_process_exp {
|
|---|
| 115 | return '';
|
|---|
| 116 | }
|
|---|
| 117 |
|
|---|
| 118 |
|
|---|
| 119 | # Associate the unknown file with the new document
|
|---|
| 120 |
|
|---|
| 121 | sub associate_unknown_file {
|
|---|
| 122 | my $self = shift (@_);
|
|---|
| 123 | my $filename = shift (@_); # filename with full path
|
|---|
| 124 | my $file = shift (@_); # filename without path
|
|---|
| 125 | my $doc_obj = shift (@_);
|
|---|
| 126 |
|
|---|
| 127 | my $verbosity = $self->{'verbosity'};
|
|---|
| 128 | my $outhandle = $self->{'outhandle'};
|
|---|
| 129 |
|
|---|
| 130 | # check the filename is okay
|
|---|
| 131 | return 0 if ($file eq "" || $filename eq "");
|
|---|
| 132 |
|
|---|
| 133 | # if ($filename =~ m/ /) {
|
|---|
| 134 | # print $outhandle "UnknownPlug: \"$filename\" contains a space. choking.\n";
|
|---|
| 135 | # return undef;
|
|---|
| 136 | # }
|
|---|
| 137 |
|
|---|
| 138 | # Escape spaces in the filename
|
|---|
| 139 | $filename =~ s/ /%20/g;
|
|---|
| 140 | $file =~ s/ /%20/g;
|
|---|
| 141 |
|
|---|
| 142 | # Add the file as an associated file ...
|
|---|
| 143 | my $section = $doc_obj->get_top_section();
|
|---|
| 144 | my $file_type = $self->{'file_type'} || "unknown/unknown";
|
|---|
| 145 | my $assoc_field = $self->{'assoc_field'} || "unknown_file";
|
|---|
| 146 |
|
|---|
| 147 | $doc_obj->associate_file($filename, $file, $mime_type, $section);
|
|---|
| 148 | $doc_obj->add_metadata ($section, $assoc_field, $file);
|
|---|
| 149 |
|
|---|
| 150 | return 1;
|
|---|
| 151 | }
|
|---|
| 152 |
|
|---|
| 153 |
|
|---|
| 154 |
|
|---|
| 155 | # The UnknownPlug read() function. This function does all the right
|
|---|
| 156 | # things to make general options work for a given plugin. UnknownPlug
|
|---|
| 157 | # overrides read() because there is no need to read the actual text of
|
|---|
| 158 | # the file in, because the contents of the file is not text...
|
|---|
| 159 | #
|
|---|
| 160 | #
|
|---|
| 161 | # Return number of files processed, undef if can't process
|
|---|
| 162 | #
|
|---|
| 163 | # Note that $base_dir might be "" and that $file might include directories
|
|---|
| 164 |
|
|---|
| 165 | sub read {
|
|---|
| 166 | my $self = shift (@_);
|
|---|
| 167 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs) = @_;
|
|---|
| 168 |
|
|---|
| 169 | my $outhandle = $self->{'outhandle'};
|
|---|
| 170 |
|
|---|
| 171 | # Make sure we're processing the correct file
|
|---|
| 172 | my $filename = &util::filename_cat($base_dir, $file);
|
|---|
| 173 | return 0 if $self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/;
|
|---|
| 174 | if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
|
|---|
| 175 | return undef;
|
|---|
| 176 | }
|
|---|
| 177 | print $outhandle "UnknownPlug processing \"$filename\"\n"
|
|---|
| 178 | if $self->{'verbosity'} > 1;
|
|---|
| 179 |
|
|---|
| 180 | #if there's a leading directory name, eat it...
|
|---|
| 181 | $file =~ s/^.*[\/\\]//;
|
|---|
| 182 |
|
|---|
| 183 | # create a new document
|
|---|
| 184 | my $doc_obj = new doc ($filename, "indexed_doc");
|
|---|
| 185 | $doc_obj->set_OIDtype ($processor->{'OIDtype'});
|
|---|
| 186 |
|
|---|
| 187 | # associate the file with the document
|
|---|
| 188 | if (associate_unknown_file($self, $filename, $file, $doc_obj) != 1)
|
|---|
| 189 | {
|
|---|
| 190 | print "UnknownPlug: couldn't process \"$filename\"\n";
|
|---|
| 191 | return 0;
|
|---|
| 192 | }
|
|---|
| 193 |
|
|---|
| 194 | #create an empty text string so we don't break downstream plugins
|
|---|
| 195 | my $text = "Dummy text inserted by UnknownPlug...";
|
|---|
| 196 |
|
|---|
| 197 | # include any metadata passed in from previous plugins
|
|---|
| 198 | my $section = $doc_obj->get_top_section();
|
|---|
| 199 | $self->extra_metadata ($doc_obj, $section, $metadata);
|
|---|
| 200 |
|
|---|
| 201 | # do plugin specific processing of doc_obj
|
|---|
| 202 | return undef unless defined ($self->process (\$text, $pluginfo, $base_dir,
|
|---|
| 203 | $file, $metadata, $doc_obj));
|
|---|
| 204 |
|
|---|
| 205 | # do any automatic metadata extraction
|
|---|
| 206 | $self->auto_extract_metadata ($doc_obj);
|
|---|
| 207 |
|
|---|
| 208 | # add an OID
|
|---|
| 209 | $doc_obj->set_OID();
|
|---|
| 210 | $doc_obj->add_text($section, $text);
|
|---|
| 211 |
|
|---|
| 212 | # process the document
|
|---|
| 213 | $processor->process($doc_obj);
|
|---|
| 214 |
|
|---|
| 215 | return 1;
|
|---|
| 216 | }
|
|---|
| 217 |
|
|---|
| 218 |
|
|---|
| 219 | # UnknownPlug processing of doc_obj. In practice we don't need to do
|
|---|
| 220 | # anything here because the read function takes care of everything.
|
|---|
| 221 |
|
|---|
| 222 | sub process {
|
|---|
| 223 | my $self = shift (@_);
|
|---|
| 224 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
|---|
| 225 | my $outhandle = $self->{'outhandle'};
|
|---|
| 226 |
|
|---|
| 227 | return 1;
|
|---|
| 228 | }
|
|---|
| 229 |
|
|---|
| 230 |
|
|---|
| 231 | 1;
|
|---|
| 232 |
|
|---|
| 233 |
|
|---|
| 234 |
|
|---|
| 235 |
|
|---|
| 236 |
|
|---|
| 237 |
|
|---|
| 238 |
|
|---|
| 239 |
|
|---|
| 240 |
|
|---|
| 241 |
|
|---|
| 242 |
|
|---|
