[1410] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # ConvertToPlug.pm -- plugin that inherits from HTML or TEXT Plug, depending
|
---|
| 4 | # on plugin argument convert_to
|
---|
| 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) 1999 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 |
|
---|
| 28 | # The plugin is inherited by such plugins as WordPlug and PDFPlug.
|
---|
| 29 | # It facilitates the conversion of these document types to either HTML
|
---|
| 30 | # or TEXT by setting up variable that instruct ConvertToBasPlug
|
---|
| 31 | # how to work.
|
---|
| 32 |
|
---|
| 33 | # It works by dynamically inheriting HTMLPlug or TEXTPlug based on
|
---|
| 34 | # the plugin argument 'convert_to'. If the argument is not present,
|
---|
| 35 | # the default is to inherit HTMLPlug.
|
---|
| 36 |
|
---|
| 37 |
|
---|
| 38 | package ConvertToPlug;
|
---|
| 39 |
|
---|
[1446] | 40 | use BasPlug;
|
---|
[1410] | 41 | use HTMLPlug;
|
---|
| 42 | use TEXTPlug;
|
---|
| 43 |
|
---|
| 44 | sub BEGIN {
|
---|
[1446] | 45 | @ISA = ('HTMLPlug');
|
---|
| 46 | # @ISA = ('HTMLPlug', 'TEXTPlug');
|
---|
| 47 | # @ISA = ('BasPlug'); #, 'HTMLPlug', 'TEXTPlug');
|
---|
[1410] | 48 | }
|
---|
| 49 |
|
---|
| 50 | use strict;
|
---|
| 51 |
|
---|
| 52 | sub print_usage {
|
---|
| 53 | my ($plugin_name) = @_;
|
---|
[1741] | 54 |
|
---|
| 55 | # for when this function is called directly by pluginfo.pl
|
---|
| 56 | if (ref ($plugin_name)) {
|
---|
| 57 | $plugin_name = ref ($plugin_name);
|
---|
| 58 | }
|
---|
[1410] | 59 |
|
---|
| 60 | print STDERR "\n usage: plugin $plugin_name [options]\n\n";
|
---|
| 61 | print STDERR " options:\n";
|
---|
| 62 | print STDERR " -convert_to (html|text) plugin converts to TEXT or HTML\n";
|
---|
| 63 | print STDERR " (default html)\n";
|
---|
| 64 | }
|
---|
| 65 |
|
---|
| 66 | sub parse_args
|
---|
| 67 | {
|
---|
| 68 | my $class = shift (@_);
|
---|
| 69 | my ($args) = @_;
|
---|
| 70 |
|
---|
| 71 | my $plugin_name = $class;
|
---|
| 72 | $plugin_name =~ s/\.pm$//;
|
---|
[1415] | 73 |
|
---|
[1410] | 74 | my $generate_format;
|
---|
| 75 | if (!parsargv::parse($args,
|
---|
| 76 | q^convert_to/(html|text)/html^, \$generate_format,
|
---|
| 77 | "allow_extra_options")) {
|
---|
| 78 |
|
---|
| 79 | print STDERR "\nIncorrect options passed to $plugin_name, ";
|
---|
| 80 | print STDERR "check your collect.cfg configuration file\n";
|
---|
| 81 | &print_usage($plugin_name);
|
---|
| 82 | die "\n";
|
---|
| 83 | }
|
---|
| 84 |
|
---|
| 85 | return ($plugin_name,$generate_format);
|
---|
| 86 | }
|
---|
| 87 |
|
---|
| 88 | sub new {
|
---|
| 89 | my $class = shift (@_);
|
---|
| 90 | my ($plugin_name,$generate_format) = $class->parse_args(\@_);
|
---|
| 91 | my $self;
|
---|
| 92 |
|
---|
| 93 | if ($generate_format eq "text")
|
---|
| 94 | {
|
---|
| 95 | $self = new TEXTPlug ($class, @_);
|
---|
[1435] | 96 | $self->{'convert_to'} = "TEXT";
|
---|
| 97 | $self->{'convert_to_ext'} = "txt";
|
---|
[1410] | 98 | }
|
---|
| 99 | else
|
---|
| 100 | {
|
---|
| 101 | $self = new HTMLPlug ($class, @_);
|
---|
[1435] | 102 | $self->{'convert_to'} = "HTML";
|
---|
| 103 | $self->{'convert_to_ext'} = "html";
|
---|
[1446] | 104 |
|
---|
| 105 | $self->{'rename_assoc_files'} = 1;
|
---|
| 106 | $self->{'metadata_fields'} .= ",GENERATOR";
|
---|
[1410] | 107 | }
|
---|
| 108 |
|
---|
| 109 | return bless $self, $class;
|
---|
| 110 | }
|
---|
| 111 |
|
---|
| 112 |
|
---|
[1435] | 113 |
|
---|
[1446] | 114 | # Run conversion utility on the input file.
|
---|
| 115 | #
|
---|
| 116 | # The conversion takes place in a collection specific 'tmp' directory so
|
---|
| 117 | # that we don't accidentally damage the input.
|
---|
| 118 | #
|
---|
| 119 | # The desired output type is indicated by $output_ext. This is usually
|
---|
| 120 | # something like "html" or "word", but can be "best" (or the empty string)
|
---|
| 121 | # to indicate that the conversion utility should do the best it can.
|
---|
[1435] | 122 |
|
---|
| 123 | sub tmp_area_convert_file {
|
---|
| 124 | my $self = shift (@_);
|
---|
| 125 | my ($output_ext,$input_filename, $textref) = @_;
|
---|
| 126 |
|
---|
[1446] | 127 | my $convert_to = $self->{'convert_to'};
|
---|
| 128 |
|
---|
[1435] | 129 | # softlink to collection tmp dir
|
---|
| 130 | my $colname = &util::use_collection();
|
---|
| 131 | my $tmp_dirname
|
---|
| 132 | = &util::filename_cat($ENV{'GSDLHOME'},"collect",$colname,"tmp");
|
---|
| 133 | &util::mk_dir($tmp_dirname) if (!-e $tmp_dirname);
|
---|
| 134 |
|
---|
| 135 | # derive tmp filename from input filename
|
---|
| 136 | my ($tailname,$dirname,$suffix)
|
---|
[1705] | 137 | = File::Basename::fileparse($input_filename,'\.[^\.]+$');
|
---|
| 138 |
|
---|
[1435] | 139 | my $tmp_filename = &util::filename_cat($tmp_dirname,"$tailname$suffix");
|
---|
| 140 | # Remove any white space from filename -- no risk of name collision, and
|
---|
[1446] | 141 | # makes later conversion by utils simpler
|
---|
[1435] | 142 | $tmp_filename =~ s/\s+//g;
|
---|
| 143 |
|
---|
| 144 | &util::soft_link($input_filename,$tmp_filename);
|
---|
| 145 |
|
---|
| 146 | my $verbosity = $self->{'verbosity'};
|
---|
| 147 | if ($verbosity>0)
|
---|
| 148 | {
|
---|
[1446] | 149 | print STDERR "Converting $tailname$suffix to $convert_to format\n";
|
---|
[1435] | 150 | }
|
---|
| 151 |
|
---|
[1446] | 152 | # Execute the conversion command and get the type of the result,
|
---|
| 153 | # making sure the converter gives us the appropriate output type
|
---|
| 154 | my $output_type = lc($convert_to);
|
---|
| 155 | my $cmd = "gsConvert.pl -verbose $verbosity -output $output_type \"$tmp_filename\"";
|
---|
| 156 | $output_type = `$cmd`;
|
---|
| 157 |
|
---|
| 158 | # Check STDERR here
|
---|
| 159 |
|
---|
| 160 | chomp $output_type;
|
---|
| 161 | if ($output_type eq "fail") {
|
---|
| 162 | print STDERR "Could not convert $tailname$suffix to $convert_to format\n";
|
---|
[1691] | 163 | return "";
|
---|
| 164 | ### exit 1;
|
---|
[1435] | 165 | }
|
---|
| 166 |
|
---|
| 167 | # remove symbolic link to original file
|
---|
| 168 | &util::rm($tmp_filename);
|
---|
| 169 |
|
---|
[1446] | 170 | # store the *actual* output type and return the output filename
|
---|
| 171 | $self->{'convert_to_ext'} = $output_type;
|
---|
| 172 | my $output_filename = $tmp_filename;
|
---|
| 173 | $output_filename =~ s/$suffix$/.$output_type/;
|
---|
| 174 |
|
---|
[1435] | 175 | return $output_filename;
|
---|
| 176 | }
|
---|
| 177 |
|
---|
| 178 |
|
---|
| 179 | # Remove collection specific tmp directory and all its contents.
|
---|
| 180 |
|
---|
| 181 | sub cleanup_tmp_area {
|
---|
| 182 | my $self = shift (@_);
|
---|
| 183 |
|
---|
| 184 | my $colname = &util::use_collection();
|
---|
| 185 | my $tmp_dirname
|
---|
| 186 | = &util::filename_cat($ENV{'GSDLHOME'},"collect",$colname,"tmp");
|
---|
| 187 | &util::rm_r($tmp_dirname);
|
---|
| 188 | &util::mk_dir($tmp_dirname);
|
---|
| 189 | }
|
---|
| 190 |
|
---|
| 191 |
|
---|
| 192 |
|
---|
[1420] | 193 | # Override BasPlug read_file
|
---|
| 194 |
|
---|
| 195 | sub read_file {
|
---|
| 196 | my $self = shift (@_);
|
---|
| 197 | my ($src_filename, $textref) = @_;
|
---|
| 198 |
|
---|
| 199 | my $output_ext = $self->{'convert_to_ext'};
|
---|
| 200 | my $conv_filename = $self->tmp_area_convert_file($output_ext,$src_filename);
|
---|
[1691] | 201 | if ("$conv_filename" eq "") {return "";} # allows continue on errors
|
---|
[1420] | 202 | $self->{'conv_filename'} = $conv_filename;
|
---|
| 203 |
|
---|
| 204 | BasPlug::read_file($self,$conv_filename,$textref);
|
---|
| 205 | }
|
---|
| 206 |
|
---|
| 207 |
|
---|
| 208 | # Override BasPlug read
|
---|
| 209 |
|
---|
| 210 | sub read {
|
---|
| 211 | my $self = shift (@_);
|
---|
| 212 |
|
---|
| 213 | my $ret_val = BasPlug::read($self,@_);
|
---|
| 214 |
|
---|
| 215 | $self->cleanup_tmp_area();
|
---|
| 216 |
|
---|
| 217 | return $ret_val;
|
---|
| 218 | }
|
---|
| 219 |
|
---|
| 220 |
|
---|
[1410] | 221 | # do plugin specific processing of doc_obj for HTML type
|
---|
| 222 | sub process_type {
|
---|
| 223 | my $self = shift (@_);
|
---|
| 224 | my ($doc_ext, $textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
---|
| 225 |
|
---|
| 226 | my $conv_filename = $self->{'conv_filename'};
|
---|
| 227 | my $tmp_dirname = File::Basename::dirname($conv_filename);
|
---|
| 228 | my $tmp_tailname = File::Basename::basename($conv_filename);
|
---|
[1929] | 229 |
|
---|
[1410] | 230 | my $convert_to = $self->{'convert_to'};
|
---|
| 231 | my $ret_val;
|
---|
| 232 |
|
---|
| 233 | if ($convert_to eq "TEXT")
|
---|
| 234 | {
|
---|
[1446] | 235 |
|
---|
[1410] | 236 | $ret_val = TEXTPlug::process($self,$textref,$pluginfo,
|
---|
| 237 | $tmp_dirname,$tmp_tailname,
|
---|
| 238 | $metadata,$doc_obj);
|
---|
| 239 | }
|
---|
| 240 | else
|
---|
| 241 | {
|
---|
| 242 | $ret_val = HTMLPlug::process($self,$textref,$pluginfo,
|
---|
| 243 | $tmp_dirname,$tmp_tailname,
|
---|
| 244 | $metadata,$doc_obj);
|
---|
| 245 | }
|
---|
| 246 |
|
---|
| 247 | # associate original file with doc object
|
---|
| 248 | my $cursection = $doc_obj->get_top_section();
|
---|
| 249 | my $filename = &util::filename_cat($base_dir,$file);
|
---|
| 250 | $doc_obj->associate_file($filename, "doc.$doc_ext", undef, $cursection);
|
---|
| 251 |
|
---|
[1435] | 252 | my $doclink = "<a href=_httpcollection_/index/assoc/[archivedir]/doc.$doc_ext>";
|
---|
| 253 | $doc_obj->add_utf8_metadata ($cursection, "srclink", $doclink);
|
---|
[1410] | 254 | $doc_obj->add_utf8_metadata ($cursection, "srcicon", "_icon".$doc_ext."_");
|
---|
| 255 | $doc_obj->add_utf8_metadata ($cursection, "/srclink", "</a>");
|
---|
| 256 |
|
---|
| 257 | return $ret_val;
|
---|
| 258 | }
|
---|
| 259 |
|
---|
| 260 | 1;
|
---|