| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # BasPlug.pm -- base class for all the import plugins
|
|---|
| 4 | # A component of the Greenstone digital library software
|
|---|
| 5 | # from the New Zealand Digital Library Project at the
|
|---|
| 6 | # University of Waikato, New Zealand.
|
|---|
| 7 | #
|
|---|
| 8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
|---|
| 9 | #
|
|---|
| 10 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 11 | # it under the terms of the GNU General Public License as published by
|
|---|
| 12 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 13 | # (at your option) any later version.
|
|---|
| 14 | #
|
|---|
| 15 | # This program is distributed in the hope that it will be useful,
|
|---|
| 16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 18 | # GNU General Public License for more details.
|
|---|
| 19 | #
|
|---|
| 20 | # You should have received a copy of the GNU General Public License
|
|---|
| 21 | # along with this program; if not, write to the Free Software
|
|---|
| 22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 23 | #
|
|---|
| 24 | ###########################################################################
|
|---|
| 25 |
|
|---|
| 26 | package BasPlug;
|
|---|
| 27 |
|
|---|
| 28 | use parsargv;
|
|---|
| 29 | use multiread;
|
|---|
| 30 | use cnseg;
|
|---|
| 31 | use acronym;
|
|---|
| 32 | use textcat;
|
|---|
| 33 | use strict;
|
|---|
| 34 | use doc;
|
|---|
| 35 |
|
|---|
| 36 | sub print_general_usage {
|
|---|
| 37 | my ($plugin_name) = @_;
|
|---|
| 38 |
|
|---|
| 39 | print STDERR "\n usage: plugin $plugin_name [options]\n\n";
|
|---|
| 40 | print STDERR " -input_encoding The encoding of the source documents. Documents will be\n";
|
|---|
| 41 | print STDERR " converted from these encodings and stored internally as\n";
|
|---|
| 42 | print STDERR " utf8. The default input_encoding is ascii. Accepted values\n";
|
|---|
| 43 | print STDERR " are:\n";
|
|---|
| 44 | print STDERR " iso_8859_1 (extended ascii)\n";
|
|---|
| 45 | print STDERR " Latin1 (the same as iso-8859-1)\n";
|
|---|
| 46 | print STDERR " ascii (7 bit ascii -- may be faster than Latin1 as no\n";
|
|---|
| 47 | print STDERR " conversion is neccessary)\n";
|
|---|
| 48 | print STDERR " gb (GB or GBK simplified Chinese)\n";
|
|---|
| 49 | print STDERR " iso_8859_6 (8 bit Arabic)\n";
|
|---|
| 50 | print STDERR " windows_1256 (Windows codepage 1256 (Arabic))\n";
|
|---|
| 51 | print STDERR " Arabic (the same as windows_1256)\n";
|
|---|
| 52 | print STDERR " utf8 (either utf8 or unicode -- automatically detected)\n";
|
|---|
| 53 | print STDERR " unicode (just unicode -- doesn't currently do endian\n";
|
|---|
| 54 | print STDERR " detection)\n";
|
|---|
| 55 | print STDERR " -process_exp A perl regular expression to match against filenames.\n";
|
|---|
| 56 | print STDERR " Matching filenames will be processed by this plugin.\n";
|
|---|
| 57 | print STDERR " Each plugin has its own default process_exp. e.g HTMLPlug\n";
|
|---|
| 58 | print STDERR " defaults to '(?i)\.html?\$' i.e. all documents ending in\n";
|
|---|
| 59 | print STDERR " .htm or .html (case-insensitive).\n";
|
|---|
| 60 | print STDERR " -block_exp Files matching this regular expression will be blocked from\n";
|
|---|
| 61 | print STDERR " being passed to any further plugins in the list. This has no\n";
|
|---|
| 62 | print STDERR " real effect other than to prevent lots of warning messages\n";
|
|---|
| 63 | print STDERR " about input files you don't care about. Each plugin may or may\n";
|
|---|
| 64 | print STDERR " not have a default block_exp. e.g. by default HTMLPlug blocks\n";
|
|---|
| 65 | print STDERR " any files with .gif, .jpg, .jpeg, .png, .pdf, .rtf or .css\n";
|
|---|
| 66 | print STDERR " file extensions.\n";
|
|---|
| 67 | print STDERR " -extract_acronyms Extract acronyms from within text and set as metadata\n\n";
|
|---|
| 68 | print STDERR " -extract_langauge Identify the language of the text and set as metadata\n\n";
|
|---|
| 69 | }
|
|---|
| 70 |
|
|---|
| 71 | # print_usage should be overridden for any sub-classes having
|
|---|
| 72 | # their own plugin specific options
|
|---|
| 73 | sub print_usage {
|
|---|
| 74 | print STDERR "\nThis plugin has no plugin specific options\n\n";
|
|---|
| 75 |
|
|---|
| 76 | }
|
|---|
| 77 |
|
|---|
| 78 | sub new {
|
|---|
| 79 | my $class = shift (@_);
|
|---|
| 80 | my $plugin_name = shift (@_);
|
|---|
| 81 |
|
|---|
| 82 | my $self = {};
|
|---|
| 83 | my $encodings = "^(iso_8859_1|Latin1|ascii|gb|iso_8859_6|windows_1256|Arabic|utf8|unicode)\$";
|
|---|
| 84 |
|
|---|
| 85 | # general options available to all plugins
|
|---|
| 86 | if (!parsargv::parse(\@_,
|
|---|
| 87 | qq^input_encoding/$encodings/ascii^, \$self->{'input_encoding'},
|
|---|
| 88 | q^process_exp/.*/^, \$self->{'process_exp'},
|
|---|
| 89 | q^block_exp/.*/^, \$self->{'block_exp'},
|
|---|
| 90 | q^extract_acronyms^, \$self->{'extract_acronyms'},
|
|---|
| 91 | q^extract_language^, \$self->{'extract_language'},
|
|---|
| 92 | "allow_extra_options")) {
|
|---|
| 93 |
|
|---|
| 94 | print STDERR "\nThe $plugin_name plugin uses an incorrect general option (general options are those\n";
|
|---|
| 95 | print STDERR "available to all plugins). Check your collect.cfg configuration file.\n";
|
|---|
| 96 | &print_general_usage($plugin_name);
|
|---|
| 97 | die "\n";
|
|---|
| 98 | }
|
|---|
| 99 |
|
|---|
| 100 | return bless $self, $class;
|
|---|
| 101 | }
|
|---|
| 102 |
|
|---|
| 103 | # initialize BasPlug options
|
|---|
| 104 | # if init() is overridden in a sub-class, remember to call BasPlug::init()
|
|---|
| 105 | sub init {
|
|---|
| 106 | my $self = shift (@_);
|
|---|
| 107 | my ($verbosity) = @_;
|
|---|
| 108 |
|
|---|
| 109 | # verbosity is passed through from the processor
|
|---|
| 110 | $self->{'verbosity'} = $verbosity;
|
|---|
| 111 |
|
|---|
| 112 | # set process_exp and block_exp to defaults unless they were
|
|---|
| 113 | # explicitly set
|
|---|
| 114 |
|
|---|
| 115 | if ((!$self->is_recursive()) and
|
|---|
| 116 | (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
|
|---|
| 117 |
|
|---|
| 118 | $self->{'process_exp'} = $self->get_default_process_exp ();
|
|---|
| 119 | if ($self->{'process_exp'} eq "") {
|
|---|
| 120 | warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
|
|---|
| 121 | }
|
|---|
| 122 | }
|
|---|
| 123 |
|
|---|
| 124 | if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
|
|---|
| 125 | $self->{'block_exp'} = $self->get_default_block_exp ();
|
|---|
| 126 | }
|
|---|
| 127 |
|
|---|
| 128 | # handle input_encoding aliases
|
|---|
| 129 | $self->{'input_encoding'} = "iso_8859_1" if $self->{'input_encoding'} eq "Latin1";
|
|---|
| 130 | $self->{'input_encoding'} = "windows_1256" if $self->{'input_encoding'} eq "Arabic";
|
|---|
| 131 | }
|
|---|
| 132 |
|
|---|
| 133 | sub begin {
|
|---|
| 134 | my $self = shift (@_);
|
|---|
| 135 | my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
|
|---|
| 136 | }
|
|---|
| 137 |
|
|---|
| 138 | sub end {
|
|---|
| 139 | my ($self) = @_;
|
|---|
| 140 | }
|
|---|
| 141 |
|
|---|
| 142 | # this function should be overridden to return 1
|
|---|
| 143 | # in recursive plugins
|
|---|
| 144 | sub is_recursive {
|
|---|
| 145 | my $self = shift (@_);
|
|---|
| 146 |
|
|---|
| 147 | return 0;
|
|---|
| 148 | }
|
|---|
| 149 |
|
|---|
| 150 | sub get_default_block_exp {
|
|---|
| 151 | my $self = shift (@_);
|
|---|
| 152 |
|
|---|
| 153 | return "";
|
|---|
| 154 | }
|
|---|
| 155 |
|
|---|
| 156 | sub get_default_process_exp {
|
|---|
| 157 | my $self = shift (@_);
|
|---|
| 158 |
|
|---|
| 159 | return "";
|
|---|
| 160 | }
|
|---|
| 161 |
|
|---|
| 162 | # The BasPlug read() function. This function does all the right things
|
|---|
| 163 | # to make general options work for a given plugin. It calls the process()
|
|---|
| 164 | # function which does all the work specific to a plugin (like the old
|
|---|
| 165 | # read functions used to do). Most plugins should define their own
|
|---|
| 166 | # process() function and let this read() function keep control.
|
|---|
| 167 | #
|
|---|
| 168 | # recursive plugins (e.g. RecPlug) and specialized plugins like those
|
|---|
| 169 | # capable of processing many documents within a single file (e.g.
|
|---|
| 170 | # GMLPlug) should normally implement their own version of read()
|
|---|
| 171 | #
|
|---|
| 172 | # Return number of files processed, undef if can't process
|
|---|
| 173 | # Note that $base_dir might be "" and that $file might
|
|---|
| 174 | # include directories
|
|---|
| 175 |
|
|---|
| 176 | sub read {
|
|---|
| 177 | my $self = shift (@_);
|
|---|
| 178 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs) = @_;
|
|---|
| 179 |
|
|---|
| 180 | if ($self->is_recursive()) {
|
|---|
| 181 | die "BasPlug::read function must be implemented in sub-class for recursive plugins\n";
|
|---|
| 182 | }
|
|---|
| 183 |
|
|---|
| 184 | my $filename = &util::filename_cat($base_dir, $file);
|
|---|
| 185 | return 0 if $self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/;
|
|---|
| 186 | if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
|
|---|
| 187 | return undef;
|
|---|
| 188 | }
|
|---|
| 189 | my $plugin_name = ref ($self);
|
|---|
| 190 | $file =~ s/^[\/\\]+//; # $file often begins with / so we'll tidy it up
|
|---|
| 191 |
|
|---|
| 192 | # create a new document
|
|---|
| 193 | my $doc_obj = new doc ($file, "indexed_doc");
|
|---|
| 194 |
|
|---|
| 195 | # read in file ($text will be in utf8)
|
|---|
| 196 | my $text = "";
|
|---|
| 197 | $self->read_file ($filename, \$text);
|
|---|
| 198 |
|
|---|
| 199 | if ($text !~ /\w/) {
|
|---|
| 200 | print STDERR "$plugin_name: ERROR: $file contains no text\n" if $self->{'verbosity'};
|
|---|
| 201 | return 0;
|
|---|
| 202 | }
|
|---|
| 203 |
|
|---|
| 204 | # include any metadata passed in from previous plugins
|
|---|
| 205 | # note that this metadata is associated with the top level section
|
|---|
| 206 | $self->extra_metadata ($doc_obj, $doc_obj->get_top_section(), $metadata);
|
|---|
| 207 |
|
|---|
| 208 | # do plugin specific processing of doc_obj
|
|---|
| 209 | return undef unless defined ($self->process (\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj));
|
|---|
| 210 |
|
|---|
| 211 | # do any automatic metadata extraction
|
|---|
| 212 | $self->auto_extract_metadata ($doc_obj);
|
|---|
| 213 |
|
|---|
| 214 | # add an OID
|
|---|
| 215 | $doc_obj->set_OID();
|
|---|
| 216 |
|
|---|
| 217 | # process the document
|
|---|
| 218 | $processor->process($doc_obj);
|
|---|
| 219 |
|
|---|
| 220 | return 1; # processed the file
|
|---|
| 221 | }
|
|---|
| 222 |
|
|---|
| 223 | # returns undef if file is rejected by the plugin
|
|---|
| 224 | sub process {
|
|---|
| 225 | my $self = shift (@_);
|
|---|
| 226 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
|---|
| 227 |
|
|---|
| 228 | die "Basplug::process function must be implemented in sub-class\n";
|
|---|
| 229 |
|
|---|
| 230 | return undef; # never gets here
|
|---|
| 231 | }
|
|---|
| 232 |
|
|---|
| 233 | # uses the multiread package to read in the entire file pointed to
|
|---|
| 234 | # by filename and loads the resulting text into $$textref. Input text
|
|---|
| 235 | # may be in any of the encodings handled by multiread, output text
|
|---|
| 236 | # will be in utf8
|
|---|
| 237 | sub read_file {
|
|---|
| 238 | my $self = shift (@_);
|
|---|
| 239 | my ($filename, $textref) = @_;
|
|---|
| 240 |
|
|---|
| 241 | $$textref = "";
|
|---|
| 242 |
|
|---|
| 243 | open (FILE, $filename) || die "BasPlug::read_file could not open $filename for reading ($!)\n";
|
|---|
| 244 |
|
|---|
| 245 | if ($self->{'input_encoding'} eq "ascii") {
|
|---|
| 246 | undef $/;
|
|---|
| 247 | $$textref = <FILE>;
|
|---|
| 248 | $/ = "\n";
|
|---|
| 249 | } else {
|
|---|
| 250 | my $reader = new multiread();
|
|---|
| 251 | $reader->set_handle ('BasPlug::FILE');
|
|---|
| 252 | $reader->set_encoding ($self->{'input_encoding'});
|
|---|
| 253 | $reader->read_file ($textref);
|
|---|
| 254 |
|
|---|
| 255 | if ($self->{'input_encoding'} eq "gb") {
|
|---|
| 256 | # segment the Chinese words
|
|---|
| 257 | $$textref = &cnseg::segment($$textref);
|
|---|
| 258 | }
|
|---|
| 259 | }
|
|---|
| 260 |
|
|---|
| 261 | close FILE;
|
|---|
| 262 | }
|
|---|
| 263 |
|
|---|
| 264 | # add any extra metadata that's been passed around from one
|
|---|
| 265 | # plugin to another.
|
|---|
| 266 | # extra_metadata uses add_utf8_metadata so it expects metadata values
|
|---|
| 267 | # to already be in utf8
|
|---|
| 268 | sub extra_metadata {
|
|---|
| 269 | my $self = shift (@_);
|
|---|
| 270 | my ($doc_obj, $cursection, $metadata) = @_;
|
|---|
| 271 |
|
|---|
| 272 | foreach my $field (keys(%$metadata)) {
|
|---|
| 273 | # $metadata->{$field} may be an array reference
|
|---|
| 274 | if (ref ($metadata->{$field}) eq "ARRAY") {
|
|---|
| 275 | map {
|
|---|
| 276 | $doc_obj->add_utf8_metadata ($cursection, $field, $_);
|
|---|
| 277 | } @{$metadata->{$field}};
|
|---|
| 278 | } else {
|
|---|
| 279 | $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
|
|---|
| 280 | }
|
|---|
| 281 | }
|
|---|
| 282 | }
|
|---|
| 283 |
|
|---|
| 284 | # extract acronyms (and hopefully other stuff soon too).
|
|---|
| 285 | sub auto_extract_metadata {
|
|---|
| 286 | my $self = shift (@_);
|
|---|
| 287 | my ($doc_obj) = @_;
|
|---|
| 288 |
|
|---|
| 289 | if ($self->{'extract_acronyms'}) {
|
|---|
| 290 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 291 | while (defined $thissection) {
|
|---|
| 292 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 293 | $self->extract_acronyms (\$text, $doc_obj, $thissection) if $text =~ /./;
|
|---|
| 294 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 295 | }
|
|---|
| 296 | }
|
|---|
| 297 |
|
|---|
| 298 | if ($self->{'extract_language'}) {
|
|---|
| 299 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 300 | while (defined $thissection) {
|
|---|
| 301 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 302 | $self->extract_language (\$text, $doc_obj, $thissection) if $text =~ /./;
|
|---|
| 303 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 304 | }
|
|---|
| 305 | }
|
|---|
| 306 |
|
|---|
| 307 | }
|
|---|
| 308 |
|
|---|
| 309 |
|
|---|
| 310 | # Identify the language of a section and add it to the metadata
|
|---|
| 311 | sub extract_language {
|
|---|
| 312 | my $self = shift (@_);
|
|---|
| 313 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 314 |
|
|---|
| 315 | # remove all HTML tags
|
|---|
| 316 | my $text = $$textref;
|
|---|
| 317 | $text =~ s/<P[^>]*>/\n/sgi;
|
|---|
| 318 | $text =~ s/<H[^>]*>/\n/sgi;
|
|---|
| 319 | $text =~ s/<[^>]*>//sgi;
|
|---|
| 320 | $text =~ tr/\n/\n/s;
|
|---|
| 321 |
|
|---|
| 322 | # get the language
|
|---|
| 323 | my @results = textcat::classify($text);
|
|---|
| 324 | @results = ("unknown") if ($#results > 2);
|
|---|
| 325 |
|
|---|
| 326 | my $language = join(" or ", @results);
|
|---|
| 327 | $doc_obj->add_utf8_metadata($thissection, "Language", $language);
|
|---|
| 328 | print "Language: $language\n";
|
|---|
| 329 |
|
|---|
| 330 | }
|
|---|
| 331 |
|
|---|
| 332 | # extract acronyms from a section in a document. progress is
|
|---|
| 333 | # reported to STDERR based on the verbosity. both the Acronym
|
|---|
| 334 | # and the AcronymKWIC metadata items are created.
|
|---|
| 335 |
|
|---|
| 336 | sub extract_acronyms {
|
|---|
| 337 | my $self = shift (@_);
|
|---|
| 338 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 339 |
|
|---|
| 340 | print STDERR " checking for acronyms ...\n"
|
|---|
| 341 | if ($self->{'verbosity'} >= 2);
|
|---|
| 342 |
|
|---|
| 343 | my $acro_array = &acronym::acronyms($textref);
|
|---|
| 344 |
|
|---|
| 345 | foreach my $acro (@$acro_array) {
|
|---|
| 346 |
|
|---|
| 347 | #check that this is the first time ...
|
|---|
| 348 | my $seen_before = "false";
|
|---|
| 349 | my $previous_data = $doc_obj->get_metadata($thissection, "Acronym");
|
|---|
| 350 | foreach my $thisAcro (@$previous_data) {
|
|---|
| 351 | if ($thisAcro eq $acro->to_string())
|
|---|
| 352 | {
|
|---|
| 353 | $seen_before = "true";
|
|---|
| 354 | print STDERR " not adding ". $acro->to_string() . "\n"
|
|---|
| 355 | if ($self->{'verbosity'} >= 2);
|
|---|
| 356 | }
|
|---|
| 357 | }
|
|---|
| 358 |
|
|---|
| 359 | if ($seen_before eq "false")
|
|---|
| 360 | {
|
|---|
| 361 | #do the normal acronym
|
|---|
| 362 | $doc_obj->add_utf8_metadata($thissection, "Acronym", $acro->to_string());
|
|---|
| 363 | print STDERR " adding ". $acro->to_string() . "\n"
|
|---|
| 364 | if ($self->{'verbosity'} >= 1);
|
|---|
| 365 |
|
|---|
| 366 | # do the KWIC (Key Word In Context) acronym
|
|---|
| 367 | my @kwic = $acro->to_string_kwic();
|
|---|
| 368 | foreach my $kwic (@kwic) {
|
|---|
| 369 | $doc_obj->add_utf8_metadata($thissection, "AcronymKWIC", $kwic);
|
|---|
| 370 | print STDERR " adding ". $kwic . "\n"
|
|---|
| 371 | if ($self->{'verbosity'} >= 2);
|
|---|
| 372 | }
|
|---|
| 373 | }
|
|---|
| 374 | }
|
|---|
| 375 | print STDERR " checked for acronyms.\n"
|
|---|
| 376 | if ($self->{'verbosity'} >= 2);
|
|---|
| 377 | }
|
|---|
| 378 |
|
|---|
| 379 | 1;
|
|---|
