| 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 | # suppress the annoying "subroutine redefined" warning that various
|
|---|
| 29 | # plugins cause under perl 5.6
|
|---|
| 30 | $SIG{__WARN__} = sub {warn($_[0]) unless ($_[0] =~ /Subroutine\s+\S+\sredefined/)};
|
|---|
| 31 |
|
|---|
| 32 | use Kea;
|
|---|
| 33 | use parsargv;
|
|---|
| 34 | use multiread;
|
|---|
| 35 | use encodings;
|
|---|
| 36 | use cnseg;
|
|---|
| 37 | use acronym;
|
|---|
| 38 | use textcat;
|
|---|
| 39 | use doc;
|
|---|
| 40 | use diagnostics;
|
|---|
| 41 | use DateExtract;
|
|---|
| 42 | use ghtml;
|
|---|
| 43 |
|
|---|
| 44 | sub print_general_usage {
|
|---|
| 45 | my ($plugin_name) = @_;
|
|---|
| 46 |
|
|---|
| 47 | print STDERR "\n usage: plugin $plugin_name [options]\n\n";
|
|---|
| 48 |
|
|---|
| 49 | print STDERR " -process_exp A perl regular expression to match against filenames.\n";
|
|---|
| 50 | print STDERR " Matching filenames will be processed by this plugin.\n";
|
|---|
| 51 | print STDERR " Each plugin has its own default process_exp. e.g HTMLPlug\n";
|
|---|
| 52 | print STDERR " defaults to '(?i)\.html?\$' i.e. all documents ending in\n";
|
|---|
| 53 | print STDERR " .htm or .html (case-insensitive).\n\n";
|
|---|
| 54 |
|
|---|
| 55 | print STDERR " -block_exp Files matching this regular expression will be blocked from\n";
|
|---|
| 56 | print STDERR " being passed to any later plugins in the list. This has no\n";
|
|---|
| 57 | print STDERR " real effect other than to prevent lots of warning messages\n";
|
|---|
| 58 | print STDERR " about input files you don't care about. Each plugin might\n";
|
|---|
| 59 | print STDERR " have a default block_exp. e.g. by default HTMLPlug blocks\n";
|
|---|
| 60 | print STDERR " any files with .gif, .jpg, .jpeg, .png or .css\n";
|
|---|
| 61 | print STDERR " file extensions.\n\n";
|
|---|
| 62 |
|
|---|
| 63 |
|
|---|
| 64 | print STDERR " -input_encoding The encoding of the source documents. Documents will be\n";
|
|---|
| 65 | print STDERR " converted from these encodings and stored internally as\n";
|
|---|
| 66 | print STDERR " utf8. The default input_encoding is 'auto'. Accepted values\n";
|
|---|
| 67 | print STDERR " are:\n";
|
|---|
| 68 |
|
|---|
| 69 | print STDERR " auto: Use text categorization algorithm to automatically\n";
|
|---|
| 70 | print STDERR " identify the encoding of each source document. This\n";
|
|---|
| 71 | print STDERR " will be slower than explicitly setting the encoding\n";
|
|---|
| 72 | print STDERR " but will work where more than one encoding is used\n";
|
|---|
| 73 | print STDERR " within the same collection.\n";
|
|---|
| 74 |
|
|---|
| 75 | print STDERR " ascii: Plain 7 bit ascii. This may be a bit faster than\n";
|
|---|
| 76 | print STDERR " using iso_8859_1. Beware of using this on a collection\n";
|
|---|
| 77 | print STDERR " of documents that may contain characters outside the\n";
|
|---|
| 78 | print STDERR " plain 7 bit ascii set though (e.g. German or French\n";
|
|---|
| 79 | print STDERR " documents containing accents), use iso_8859_1 instead.\n";
|
|---|
| 80 |
|
|---|
| 81 | print STDERR " utf8: either utf8 or unicode -- automatically detected\n";
|
|---|
| 82 | print STDERR " unicode: just unicode\n";
|
|---|
| 83 |
|
|---|
| 84 | my $e = $encodings::encodings;
|
|---|
| 85 | foreach my $enc (sort {$e->{$a}->{'name'} cmp $e->{$b}->{'name'}} keys (%$e)) {
|
|---|
| 86 | print STDERR " $enc: $e->{$enc}->{'name'}\n";
|
|---|
| 87 | }
|
|---|
| 88 | print STDERR "\n";
|
|---|
| 89 | print STDERR " -default_encoding Use this encoding if -input_encoding is set to 'auto' and\n";
|
|---|
| 90 | print STDERR " the text categorization algorithm fails to extract the\n";
|
|---|
| 91 | print STDERR " encoding or extracts an encoding unsupported by Greenstone.\n";
|
|---|
| 92 | print STDERR " The default is iso_8859_1.\n\n";
|
|---|
| 93 |
|
|---|
| 94 | print STDERR " -extract_language Identify the language of each document and set 'Language'\n";
|
|---|
| 95 | print STDERR " metadata. Note that this will be done automatically if\n";
|
|---|
| 96 | print STDERR " -input_encoding is 'auto'.\n\n";
|
|---|
| 97 | print STDERR " -default_language If Greenstone fails to work out what language a document is\n";
|
|---|
| 98 | print STDERR " the 'Language' metadata element will be set to this value.\n";
|
|---|
| 99 | print STDERR " The default is 'en' (ISO 639 language symbols are used:\n";
|
|---|
| 100 | print STDERR " en = English). Note that if -input_encoding is not set to\n";
|
|---|
| 101 | print STDERR " 'auto' and -extract_language is not set, all documents will\n";
|
|---|
| 102 | print STDERR " have their 'Language' metadata set to this value.\n\n";
|
|---|
| 103 |
|
|---|
| 104 | print STDERR " -extract_acronyms Extract acronyms from within text and set as metadata\n";
|
|---|
| 105 |
|
|---|
| 106 | print STDERR " -markup_acronyms Add acronym metadata into document text\n\n";
|
|---|
| 107 |
|
|---|
| 108 | print STDERR " -first Comma separated list of first sizes to extract from the\n";
|
|---|
| 109 | print STDERR " text into a metadata field. The field is called 'FirstNNN'.\n\n";
|
|---|
| 110 |
|
|---|
| 111 | print STDERR " -extract_email Extract email addresses as metadata\n\n";
|
|---|
| 112 |
|
|---|
| 113 | print STDERR " -extract_date Extract dates pertaining to the content of documents about\n";
|
|---|
| 114 | print STDERR " history\n";
|
|---|
| 115 | print STDERR " -maximum_date The maximum historical date to be used as metadata (in a\n";
|
|---|
| 116 | print STDERR " Common Era date, such as 1950)\n";
|
|---|
| 117 | print STDERR " -maximum_century The maximum named century to be extracted as historical\n";
|
|---|
| 118 | print STDERR " metadata (e.g. 14 will extract all references up to the\n";
|
|---|
| 119 | print STDERR " 14th century)\n";
|
|---|
| 120 | print STDERR " -no_bibliography Do not try and block bibliographic dates when extracting\n";
|
|---|
| 121 | print STDERR " historical dates.\n\n";
|
|---|
| 122 | }
|
|---|
| 123 |
|
|---|
| 124 | # print_usage should be overridden for any sub-classes having
|
|---|
| 125 | # their own plugin specific options
|
|---|
| 126 | sub print_usage {
|
|---|
| 127 | print STDERR "\nThis plugin has no plugin specific options\n\n";
|
|---|
| 128 | }
|
|---|
| 129 |
|
|---|
| 130 | sub new {
|
|---|
| 131 | my $class = shift (@_);
|
|---|
| 132 | my $plugin_name = shift (@_);
|
|---|
| 133 | my $self = {};
|
|---|
| 134 |
|
|---|
| 135 | my $enc = "^(";
|
|---|
| 136 | map {$enc .= "$_|";} keys %$encodings::encodings;
|
|---|
| 137 | my $denc = $enc . "ascii|utf8|unicode)\$";
|
|---|
| 138 | $enc .= "ascii|utf8|unicode|auto)\$";
|
|---|
| 139 |
|
|---|
| 140 | $self->{'outhandle'} = STDERR;
|
|---|
| 141 | my $year = (localtime)[5]+1900;
|
|---|
| 142 |
|
|---|
| 143 | $self->{'textcat'} = new textcat();
|
|---|
| 144 |
|
|---|
| 145 | $self->{'num_processed'} = 0;
|
|---|
| 146 | $self->{'num_not_processed'} = 0;
|
|---|
| 147 | $self->{'num_blocked'} = 0;
|
|---|
| 148 | $self->{'num_archives'} = 0;
|
|---|
| 149 |
|
|---|
| 150 | # general options available to all plugins
|
|---|
| 151 | if (!parsargv::parse(\@_,
|
|---|
| 152 | q^process_exp/.*/^, \$self->{'process_exp'},
|
|---|
| 153 | q^block_exp/.*/^, \$self->{'block_exp'},
|
|---|
| 154 | q^extract_acronyms^, \$self->{'extract_acronyms'},
|
|---|
| 155 | q^extract_keyphrases^, \$self->{'kea'}, #with extra options
|
|---|
| 156 | q^extract_keyphrase_options/.*/^, \$self->{'kea_options'}, #no extra options
|
|---|
| 157 | qq^input_encoding/$enc/auto^, \$self->{'input_encoding'},
|
|---|
| 158 | qq^default_encoding/$denc/iso_8859_1^, \$self->{'default_encoding'},
|
|---|
| 159 | q^extract_email^, \$self->{'extract_email'},
|
|---|
| 160 | q^markup_acronyms^, \$self->{'markup_acronyms'},
|
|---|
| 161 | q^default_language/.{2}/en^, \$self->{'default_language'},
|
|---|
| 162 | q^first/.*/^, \$self->{'first'},
|
|---|
| 163 | q^extract_date^, \$self->{'date_extract'},
|
|---|
| 164 | qq^maximum_date/\\d{4}/$year^, \$self->{'max_year'},
|
|---|
| 165 | q^no_bibliography^, \$self->{'no_biblio'},
|
|---|
| 166 | qq^maximum_century/-?\\d{1,2}( ?B\\.C\\.E\\.)?/-1^, \$self->{'max_century'},
|
|---|
| 167 | "allow_extra_options")) {
|
|---|
| 168 |
|
|---|
| 169 | print STDERR "\nThe $plugin_name plugin uses an incorrect general option (general options are those\n";
|
|---|
| 170 | print STDERR "available to all plugins). Check your collect.cfg configuration file.\n";
|
|---|
| 171 | &print_general_usage($plugin_name);
|
|---|
| 172 | die "\n";
|
|---|
| 173 | }
|
|---|
| 174 |
|
|---|
| 175 | return bless $self, $class;
|
|---|
| 176 | }
|
|---|
| 177 |
|
|---|
| 178 | # initialize BasPlug options
|
|---|
| 179 | # if init() is overridden in a sub-class, remember to call BasPlug::init()
|
|---|
| 180 | sub init {
|
|---|
| 181 | my $self = shift (@_);
|
|---|
| 182 | my ($verbosity, $outhandle, $failhandle) = @_;
|
|---|
| 183 |
|
|---|
| 184 | # verbosity is passed through from the processor
|
|---|
| 185 | $self->{'verbosity'} = $verbosity;
|
|---|
| 186 |
|
|---|
| 187 | # as are the outhandle and failhandle
|
|---|
| 188 | $self->{'outhandle'} = $outhandle if defined $outhandle;
|
|---|
| 189 | $self->{'failhandle'} = $failhandle;
|
|---|
| 190 |
|
|---|
| 191 | # set process_exp and block_exp to defaults unless they were
|
|---|
| 192 | # explicitly set
|
|---|
| 193 |
|
|---|
| 194 | if ((!$self->is_recursive()) and
|
|---|
| 195 | (!defined $self->{'process_exp'}) || ($self->{'process_exp'} eq "")) {
|
|---|
| 196 |
|
|---|
| 197 | $self->{'process_exp'} = $self->get_default_process_exp ();
|
|---|
| 198 | if ($self->{'process_exp'} eq "") {
|
|---|
| 199 | warn ref($self) . " Warning: Non-recursive plugin has no process_exp\n";
|
|---|
| 200 | }
|
|---|
| 201 | }
|
|---|
| 202 |
|
|---|
| 203 | if ((!defined $self->{'block_exp'}) || ($self->{'block_exp'} eq "")) {
|
|---|
| 204 | $self->{'block_exp'} = $self->get_default_block_exp ();
|
|---|
| 205 | }
|
|---|
| 206 | }
|
|---|
| 207 |
|
|---|
| 208 | sub begin {
|
|---|
| 209 | my $self = shift (@_);
|
|---|
| 210 | my ($pluginfo, $base_dir, $processor, $maxdocs) = @_;
|
|---|
| 211 | $self->initialise_extractors();
|
|---|
| 212 | }
|
|---|
| 213 |
|
|---|
| 214 | sub end {
|
|---|
| 215 | my ($self) = @_;
|
|---|
| 216 | $self->finalise_extractors();
|
|---|
| 217 | }
|
|---|
| 218 |
|
|---|
| 219 | # this function should be overridden to return 1
|
|---|
| 220 | # in recursive plugins
|
|---|
| 221 | sub is_recursive {
|
|---|
| 222 | my $self = shift (@_);
|
|---|
| 223 |
|
|---|
| 224 | return 0;
|
|---|
| 225 | }
|
|---|
| 226 |
|
|---|
| 227 | sub get_default_block_exp {
|
|---|
| 228 | my $self = shift (@_);
|
|---|
| 229 |
|
|---|
| 230 | return "";
|
|---|
| 231 | }
|
|---|
| 232 |
|
|---|
| 233 | sub get_default_process_exp {
|
|---|
| 234 | my $self = shift (@_);
|
|---|
| 235 |
|
|---|
| 236 | return "";
|
|---|
| 237 | }
|
|---|
| 238 |
|
|---|
| 239 | # The BasPlug read() function. This function does all the right things
|
|---|
| 240 | # to make general options work for a given plugin. It calls the process()
|
|---|
| 241 | # function which does all the work specific to a plugin (like the old
|
|---|
| 242 | # read functions used to do). Most plugins should define their own
|
|---|
| 243 | # process() function and let this read() function keep control.
|
|---|
| 244 | #
|
|---|
| 245 | # recursive plugins (e.g. RecPlug) and specialized plugins like those
|
|---|
| 246 | # capable of processing many documents within a single file (e.g.
|
|---|
| 247 | # GMLPlug) should normally implement their own version of read()
|
|---|
| 248 | #
|
|---|
| 249 | # Return number of files processed, undef if can't process
|
|---|
| 250 | # Note that $base_dir might be "" and that $file might
|
|---|
| 251 | # include directories
|
|---|
| 252 |
|
|---|
| 253 | sub read {
|
|---|
| 254 | my $self = shift (@_);
|
|---|
| 255 |
|
|---|
| 256 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs) = @_;
|
|---|
| 257 |
|
|---|
| 258 | if ($self->is_recursive()) {
|
|---|
| 259 | die "BasPlug::read function must be implemented in sub-class for recursive plugins\n";
|
|---|
| 260 | }
|
|---|
| 261 |
|
|---|
| 262 | my $outhandle = $self->{'outhandle'};
|
|---|
| 263 |
|
|---|
| 264 | my $filename = $file;
|
|---|
| 265 | $filename = &util::filename_cat ($base_dir, $file) if $base_dir =~ /\w/;
|
|---|
| 266 |
|
|---|
| 267 | if ($self->{'block_exp'} ne "" && $filename =~ /$self->{'block_exp'}/) {
|
|---|
| 268 | $self->{'num_blocked'} ++;
|
|---|
| 269 | return 0;
|
|---|
| 270 | }
|
|---|
| 271 | if ($filename !~ /$self->{'process_exp'}/ || !-f $filename) {
|
|---|
| 272 | return undef;
|
|---|
| 273 | }
|
|---|
| 274 | my $plugin_name = ref ($self);
|
|---|
| 275 | $file =~ s/^[\/\\]+//; # $file often begins with / so we'll tidy it up
|
|---|
| 276 |
|
|---|
| 277 | my ($language, $encoding);
|
|---|
| 278 | if ($self->{'input_encoding'} eq "auto") {
|
|---|
| 279 | # use textcat to automatically work out the input encoding and language
|
|---|
| 280 | ($language, $encoding) = $self->get_language_encoding ($filename);
|
|---|
| 281 |
|
|---|
| 282 | } elsif ($self->{'extract_language'}) {
|
|---|
| 283 | # use textcat to get language metadata
|
|---|
| 284 | ($language, $extracted_encoding) = $self->get_language_encoding ($filename);
|
|---|
| 285 | $encoding = $self->{'input_encoding'};
|
|---|
| 286 |
|
|---|
| 287 | if ($extracted_encoding ne $encoding && $self->{'verbosity'}) {
|
|---|
| 288 | print $outhandle "$plugin_name: WARNING: $file was read using $encoding encoding but ";
|
|---|
| 289 | print $outhandle "appears to be encoded as $extracted_encoding.\n";
|
|---|
| 290 | }
|
|---|
| 291 |
|
|---|
| 292 | } else {
|
|---|
| 293 | $language = $self->{'default_language'};
|
|---|
| 294 | $encoding = $self->{'input_encoding'};
|
|---|
| 295 | }
|
|---|
| 296 |
|
|---|
| 297 | # create a new document
|
|---|
| 298 | my $doc_obj = new doc ($filename, "indexed_doc");
|
|---|
| 299 | $doc_obj->set_OIDtype ($processor->{'OIDtype'});
|
|---|
| 300 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Language", $language);
|
|---|
| 301 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Encoding", $encoding);
|
|---|
| 302 | my ($filemeta) = $file =~ /([^\\\/]+)$/;
|
|---|
| 303 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Source", &ghtml::dmsafe($filemeta));
|
|---|
| 304 |
|
|---|
| 305 | # read in file ($text will be in utf8)
|
|---|
| 306 | my $text = "";
|
|---|
| 307 | $self->read_file ($filename, $encoding, $language, \$text);
|
|---|
| 308 |
|
|---|
| 309 | if (!length ($text)) {
|
|---|
| 310 | print $outhandle "$plugin_name: ERROR: $file contains no text\n" if $self->{'verbosity'};
|
|---|
| 311 |
|
|---|
| 312 | my $failhandle = $self->{'failhandle'};
|
|---|
| 313 | print $failhandle "$file: " . ref($self) . ": file contains no text\n";
|
|---|
| 314 | $self->{'num_not_processed'} ++;
|
|---|
| 315 |
|
|---|
| 316 | return 0;
|
|---|
| 317 | }
|
|---|
| 318 |
|
|---|
| 319 | # include any metadata passed in from previous plugins
|
|---|
| 320 | # note that this metadata is associated with the top level section
|
|---|
| 321 | $self->extra_metadata ($doc_obj, $doc_obj->get_top_section(), $metadata);
|
|---|
| 322 |
|
|---|
| 323 | # do plugin specific processing of doc_obj
|
|---|
| 324 | return undef unless defined ($self->process (\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj));
|
|---|
| 325 |
|
|---|
| 326 | # do any automatic metadata extraction
|
|---|
| 327 | $self->auto_extract_metadata ($doc_obj);
|
|---|
| 328 |
|
|---|
| 329 | # add an OID
|
|---|
| 330 | $doc_obj->set_OID();
|
|---|
| 331 |
|
|---|
| 332 | # process the document
|
|---|
| 333 | $processor->process($doc_obj);
|
|---|
| 334 |
|
|---|
| 335 | $self->{'num_processed'} ++;
|
|---|
| 336 |
|
|---|
| 337 | return 1; # processed the file
|
|---|
| 338 | }
|
|---|
| 339 |
|
|---|
| 340 | # returns undef if file is rejected by the plugin
|
|---|
| 341 | sub process {
|
|---|
| 342 | my $self = shift (@_);
|
|---|
| 343 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj) = @_;
|
|---|
| 344 |
|
|---|
| 345 | die "Basplug::process function must be implemented in sub-class\n";
|
|---|
| 346 |
|
|---|
| 347 | return undef; # never gets here
|
|---|
| 348 | }
|
|---|
| 349 |
|
|---|
| 350 | # uses the multiread package to read in the entire file pointed to
|
|---|
| 351 | # by filename and loads the resulting text into $$textref. Input text
|
|---|
| 352 | # may be in any of the encodings handled by multiread, output text
|
|---|
| 353 | # will be in utf8
|
|---|
| 354 | sub read_file {
|
|---|
| 355 | my $self = shift (@_);
|
|---|
| 356 | my ($filename, $encoding, $language, $textref) = @_;
|
|---|
| 357 |
|
|---|
| 358 | if (!-r $filename)
|
|---|
| 359 | {
|
|---|
| 360 | my $outhandle = $self->{'outhandle'};
|
|---|
| 361 | print $outhandle "Read permission denied for $filename\n" if $self->{'verbosity'};
|
|---|
| 362 | return;
|
|---|
| 363 | }
|
|---|
| 364 |
|
|---|
| 365 | $$textref = "";
|
|---|
| 366 |
|
|---|
| 367 | open (FILE, $filename) || die "BasPlug::read_file could not open $filename for reading ($!)\n";
|
|---|
| 368 |
|
|---|
| 369 | if ($encoding eq "ascii") {
|
|---|
| 370 | undef $/;
|
|---|
| 371 | $$textref = <FILE>;
|
|---|
| 372 | $/ = "\n";
|
|---|
| 373 | } else {
|
|---|
| 374 | my $reader = new multiread();
|
|---|
| 375 | $reader->set_handle ('BasPlug::FILE');
|
|---|
| 376 | $reader->set_encoding ($encoding);
|
|---|
| 377 | $reader->read_file ($textref);
|
|---|
| 378 |
|
|---|
| 379 | if ($language eq "zh") {
|
|---|
| 380 | # segment the Chinese words
|
|---|
| 381 | $$textref = &cnseg::segment($$textref);
|
|---|
| 382 | }
|
|---|
| 383 | }
|
|---|
| 384 |
|
|---|
| 385 | close FILE;
|
|---|
| 386 | }
|
|---|
| 387 |
|
|---|
| 388 | # Uses textcat to work out the encoding and language of the text in
|
|---|
| 389 | # $filename. All html tags are removed before processing.
|
|---|
| 390 | # returns an array containing "language" and "encoding"
|
|---|
| 391 | sub get_language_encoding {
|
|---|
| 392 | my $self = shift (@_);
|
|---|
| 393 | my ($filename) = @_;
|
|---|
| 394 | my $outhandle = $self->{'outhandle'};
|
|---|
| 395 |
|
|---|
| 396 | # read in file
|
|---|
| 397 | open (FILE, $filename) || die "BasPlug::get_language_encoding could not open $filename for reading ($!)\n";
|
|---|
| 398 | undef $/;
|
|---|
| 399 | my $text = <FILE>;
|
|---|
| 400 | $/ = "\n";
|
|---|
| 401 | close FILE;
|
|---|
| 402 |
|
|---|
| 403 | # remove <title>stuff</title> -- as titles tend often to be in English
|
|---|
| 404 | # for foreign language documents
|
|---|
| 405 | $text =~ s/<title>.*?<\/title>//i;
|
|---|
| 406 |
|
|---|
| 407 | # remove all HTML tags
|
|---|
| 408 | $text =~ s/<[^>]*>//sg;
|
|---|
| 409 |
|
|---|
| 410 | # get the language/encoding
|
|---|
| 411 | my $results = $self->{'textcat'}->classify(\$text);
|
|---|
| 412 |
|
|---|
| 413 | # if textcat returns 3 or less possibilities we'll use the
|
|---|
| 414 | # first one in the list - otherwise use the defaults
|
|---|
| 415 | if (scalar @$results > 3) {
|
|---|
| 416 |
|
|---|
| 417 | if ($self->{'input_encoding'} ne 'auto') {
|
|---|
| 418 | if ($self->{'extract_language'} && $self->{'verbosity'}) {
|
|---|
| 419 | print $outhandle "BasPlug: WARNING: language could not be extracted from $filename - ";
|
|---|
| 420 | print $outhandle "defaulting to $self->{'default_language'}\n";
|
|---|
| 421 | }
|
|---|
| 422 | return ($self->{'default_language'}, $self->{'input_encoding'});
|
|---|
| 423 |
|
|---|
| 424 | } else {
|
|---|
| 425 | if ($self->{'verbosity'}) {
|
|---|
| 426 | print $outhandle "BASPlug: WARNING: language/encoding could not be extracted from $filename - ";
|
|---|
| 427 | print $outhandle "defaulting to $self->{'default_language'}/$self->{'default_encoding'}\n";
|
|---|
| 428 | }
|
|---|
| 429 | return ($self->{'default_language'}, $self->{'default_encoding'});
|
|---|
| 430 | }
|
|---|
| 431 | }
|
|---|
| 432 |
|
|---|
| 433 | # format language/encoding
|
|---|
| 434 | my ($language, $encoding) = $results->[0] =~ /^([^-]*)(?:-(.*))?$/;
|
|---|
| 435 | if (!defined $language) {
|
|---|
| 436 | if ($self->{'verbosity'}) {
|
|---|
| 437 | print $outhandle "BasPlug: WARNING: language could not be extracted from $filename - ";
|
|---|
| 438 | print $outhandle "defaulting to $self->{'default_language'}\n";
|
|---|
| 439 | }
|
|---|
| 440 | $language = $self->{'default_language'};
|
|---|
| 441 | }
|
|---|
| 442 | if (!defined $encoding) {
|
|---|
| 443 | if ($self->{'verbosity'}) {
|
|---|
| 444 | print $outhandle "BasPlug: WARNING: encoding could not be extracted from $filename - ";
|
|---|
| 445 | print $outhandle "defaulting to $self->{'default_encoding'}\n";
|
|---|
| 446 | }
|
|---|
| 447 | $encoding = $self->{'default_encoding'};
|
|---|
| 448 | }
|
|---|
| 449 |
|
|---|
| 450 | if ($encoding !~ /^(ascii|utf8|unicode)$/ &&
|
|---|
| 451 | !defined $encodings::encodings->{$encoding}) {
|
|---|
| 452 | if ($self->{'verbosity'}) {
|
|---|
| 453 | print $outhandle "BasPlug: WARNING: $filename appears to be encoded in an unsupported encoding ($encoding) - ";
|
|---|
| 454 | print $outhandle "using $self->{'default_encoding'}\n";
|
|---|
| 455 | }
|
|---|
| 456 | $encoding = $self->{'default_encoding'};
|
|---|
| 457 | }
|
|---|
| 458 |
|
|---|
| 459 | return ($language, $encoding);
|
|---|
| 460 | }
|
|---|
| 461 |
|
|---|
| 462 | # add any extra metadata that's been passed around from one
|
|---|
| 463 | # plugin to another.
|
|---|
| 464 | # extra_metadata uses add_utf8_metadata so it expects metadata values
|
|---|
| 465 | # to already be in utf8
|
|---|
| 466 | sub extra_metadata {
|
|---|
| 467 | my $self = shift (@_);
|
|---|
| 468 | my ($doc_obj, $cursection, $metadata) = @_;
|
|---|
| 469 |
|
|---|
| 470 | foreach my $field (keys(%$metadata)) {
|
|---|
| 471 | # $metadata->{$field} may be an array reference
|
|---|
| 472 | if (ref ($metadata->{$field}) eq "ARRAY") {
|
|---|
| 473 | map {
|
|---|
| 474 | $doc_obj->add_utf8_metadata ($cursection, $field, $_);
|
|---|
| 475 | } @{$metadata->{$field}};
|
|---|
| 476 | } else {
|
|---|
| 477 | $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field});
|
|---|
| 478 | }
|
|---|
| 479 | }
|
|---|
| 480 | }
|
|---|
| 481 |
|
|---|
| 482 | # initialise metadata extractors
|
|---|
| 483 | sub initialise_extractors {
|
|---|
| 484 | my $self = shift (@_);
|
|---|
| 485 |
|
|---|
| 486 | if ($self->{'extract_acronyms'} || $self->{'markup_acronyms'}) {
|
|---|
| 487 | &acronym::initialise_acronyms();
|
|---|
| 488 | }
|
|---|
| 489 | }
|
|---|
| 490 |
|
|---|
| 491 | # finalise metadata extractors
|
|---|
| 492 | sub finalise_extractors {
|
|---|
| 493 | my $self = shift (@_);
|
|---|
| 494 |
|
|---|
| 495 | if ($self->{'extract_acronyms'} || $self->{'markup_acronyms'}) {
|
|---|
| 496 | &acronym::finalise_acronyms();
|
|---|
| 497 | }
|
|---|
| 498 | }
|
|---|
| 499 |
|
|---|
| 500 | # FIRSTNNN: extract the first NNN characters as metadata
|
|---|
| 501 | sub extract_first_NNNN_characters {
|
|---|
| 502 | my $self = shift (@_);
|
|---|
| 503 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 504 |
|
|---|
| 505 | foreach my $size (split /,/, $self->{'first'}) {
|
|---|
| 506 | my $tmptext = $$textref;
|
|---|
| 507 | $tmptext =~ s/^\s+//;
|
|---|
| 508 | $tmptext =~ s/\s+$//;
|
|---|
| 509 | $tmptext =~ s/\s+/ /gs;
|
|---|
| 510 | $tmptext = substr ($tmptext, 0, $size);
|
|---|
| 511 | $tmptext =~ s/\s\S*$/…/;
|
|---|
| 512 | $doc_obj->add_utf8_metadata ($thissection, "First$size", $tmptext);
|
|---|
| 513 | }
|
|---|
| 514 | }
|
|---|
| 515 |
|
|---|
| 516 | sub extract_email {
|
|---|
| 517 | my $self = shift (@_);
|
|---|
| 518 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 519 | my $outhandle = $self->{'outhandle'};
|
|---|
| 520 |
|
|---|
| 521 | print $outhandle " extracting email addresses ...\n"
|
|---|
| 522 | if ($self->{'verbosity'} > 2);
|
|---|
| 523 |
|
|---|
| 524 | my @email = ($$textref =~ m/([-a-z0-9\.@+_=]+@(?:[-a-z0-9]+\.)+(?:com|org|edu|mil|int|net|[a-z][a-z]))/g);
|
|---|
| 525 | @email = sort @email;
|
|---|
| 526 |
|
|---|
| 527 | my @email2 = ();
|
|---|
| 528 | foreach my $address (@email) {
|
|---|
| 529 | if (!(join(" ",@email2) =~ m/$address/ )) {
|
|---|
| 530 | push @email2, $address;
|
|---|
| 531 | $doc_obj->add_utf8_metadata ($thissection, "emailAddress", $address);
|
|---|
| 532 | print $outhandle " extracting $address\n"
|
|---|
| 533 | if ($self->{'verbosity'} > 3);
|
|---|
| 534 | }
|
|---|
| 535 | }
|
|---|
| 536 | print $outhandle " done extracting email addresses.\n"
|
|---|
| 537 | if ($self->{'verbosity'} > 2);
|
|---|
| 538 |
|
|---|
| 539 | }
|
|---|
| 540 |
|
|---|
| 541 | # extract metadata
|
|---|
| 542 | sub auto_extract_metadata {
|
|---|
| 543 |
|
|---|
| 544 |
|
|---|
| 545 | my $self = shift (@_);
|
|---|
| 546 | my ($doc_obj) = @_;
|
|---|
| 547 |
|
|---|
| 548 | if ($self->{'extract_email'}) {
|
|---|
| 549 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 550 | while (defined $thissection) {
|
|---|
| 551 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 552 | $self->extract_email (\$text, $doc_obj, $thissection) if $text =~ /./;
|
|---|
| 553 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 554 | }
|
|---|
| 555 | }
|
|---|
| 556 |
|
|---|
| 557 |
|
|---|
| 558 | #adding kea keyphrases
|
|---|
| 559 | if ($self->{'kea'}) {
|
|---|
| 560 |
|
|---|
| 561 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 562 | my $text = "";
|
|---|
| 563 | my @list;
|
|---|
| 564 |
|
|---|
| 565 | while (defined $thissection) { #loop through sections to gather whole doc
|
|---|
| 566 | my $sectiontext = $doc_obj->get_text($thissection);
|
|---|
| 567 | $text = $text.$sectiontext;
|
|---|
| 568 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 569 | }
|
|---|
| 570 |
|
|---|
| 571 | if($self->{'kea_options'}) { #if kea options flag is set, call Kea with specified options
|
|---|
| 572 | @list = &Kea::extract_KeyPhrases ($text, $self->{'kea_options'});
|
|---|
| 573 | } else { #otherwise call Kea with no options
|
|---|
| 574 | @list = &Kea::extract_KeyPhrases ($text);
|
|---|
| 575 | }
|
|---|
| 576 |
|
|---|
| 577 | if(@list){ #if a list of kea keyphrases was returned (ie not empty)
|
|---|
| 578 | my $keyphrases = $list[0]; #first arg is keyphrase list
|
|---|
| 579 | my $stems = $list[1]; #second arg is stemmed keyphrase list
|
|---|
| 580 | print STDERR "keyphrases: $keyphrases\n";
|
|---|
| 581 | print STDERR "stems: $stems\n";
|
|---|
| 582 | $thissection = $doc_obj->get_top_section(); #add metadata to top section
|
|---|
| 583 | $doc_obj->add_metadata($thissection, "kea", $keyphrases);
|
|---|
| 584 | $doc_obj->add_metadata($thissection, "stems", $stems);
|
|---|
| 585 | }
|
|---|
| 586 | } #end of kea
|
|---|
| 587 |
|
|---|
| 588 | if ($self->{'first'}) {
|
|---|
| 589 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 590 | while (defined $thissection) {
|
|---|
| 591 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 592 | $self->extract_first_NNNN_characters (\$text, $doc_obj, $thissection) if $text =~ /./;
|
|---|
| 593 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 594 | }
|
|---|
| 595 | }
|
|---|
| 596 |
|
|---|
| 597 | if ($self->{'extract_acronyms'}) {
|
|---|
| 598 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 599 | while (defined $thissection) {
|
|---|
| 600 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 601 | $self->extract_acronyms (\$text, $doc_obj, $thissection) if $text =~ /./;
|
|---|
| 602 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 603 | }
|
|---|
| 604 | }
|
|---|
| 605 |
|
|---|
| 606 | if ($self->{'markup_acronyms'}) {
|
|---|
| 607 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 608 | while (defined $thissection) {
|
|---|
| 609 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 610 | $text = $self->markup_acronyms ($text, $doc_obj, $thissection);
|
|---|
| 611 | $doc_obj->delete_text($thissection);
|
|---|
| 612 | $doc_obj->add_text($thissection, $text);
|
|---|
| 613 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 614 | }
|
|---|
| 615 | }
|
|---|
| 616 |
|
|---|
| 617 | if($self->{'date_extract'}) {
|
|---|
| 618 | my $thissection = $doc_obj->get_top_section();
|
|---|
| 619 | while (defined $thissection) {
|
|---|
| 620 |
|
|---|
| 621 | my $text = $doc_obj->get_text($thissection);
|
|---|
| 622 | &DateExtract::get_date_metadata($text, $doc_obj,
|
|---|
| 623 | $thissection,
|
|---|
| 624 | $self->{'no_biblio'},
|
|---|
| 625 | $self->{'max_year'},
|
|---|
| 626 | $self->{'max_century'});
|
|---|
| 627 | $thissection = $doc_obj->get_next_section ($thissection);
|
|---|
| 628 | }
|
|---|
| 629 | }
|
|---|
| 630 | }
|
|---|
| 631 |
|
|---|
| 632 | # extract acronyms from a section in a document. progress is
|
|---|
| 633 | # reported to outhandle based on the verbosity. both the Acronym
|
|---|
| 634 | # and the AcronymKWIC metadata items are created.
|
|---|
| 635 |
|
|---|
| 636 | sub extract_acronyms {
|
|---|
| 637 | my $self = shift (@_);
|
|---|
| 638 | my ($textref, $doc_obj, $thissection) = @_;
|
|---|
| 639 | my $outhandle = $self->{'outhandle'};
|
|---|
| 640 |
|
|---|
| 641 | print $outhandle " extracting acronyms ...\n"
|
|---|
| 642 | if ($self->{'verbosity'} > 2);
|
|---|
| 643 |
|
|---|
| 644 | my $acro_array = &acronym::acronyms($textref);
|
|---|
| 645 |
|
|---|
| 646 | foreach my $acro (@$acro_array) {
|
|---|
| 647 |
|
|---|
| 648 | #check that this is the first time ...
|
|---|
| 649 | my $seen_before = "false";
|
|---|
| 650 | my $previous_data = $doc_obj->get_metadata($thissection, "Acronym");
|
|---|
| 651 | foreach my $thisAcro (@$previous_data) {
|
|---|
| 652 | if ($thisAcro eq $acro->to_string()) {
|
|---|
| 653 | $seen_before = "true";
|
|---|
| 654 | print $outhandle " already seen ". $acro->to_string() . "\n"
|
|---|
| 655 | if ($self->{'verbosity'} >= 4);
|
|---|
| 656 | }
|
|---|
| 657 | }
|
|---|
| 658 |
|
|---|
| 659 | if ($seen_before eq "false") {
|
|---|
| 660 | #write it to the file ...
|
|---|
| 661 | $acro->write_to_file();
|
|---|
| 662 |
|
|---|
| 663 | #do the normal acronym
|
|---|
| 664 | $doc_obj->add_utf8_metadata($thissection, "Acronym", $acro->to_string());
|
|---|
| 665 | print $outhandle " adding ". $acro->to_string() . "\n"
|
|---|
| 666 | if ($self->{'verbosity'} > 3);
|
|---|
| 667 |
|
|---|
| 668 | }
|
|---|
| 669 | }
|
|---|
| 670 | print $outhandle " done extracting acronyms. \n"
|
|---|
| 671 | if ($self->{'verbosity'} > 2);
|
|---|
| 672 | }
|
|---|
| 673 |
|
|---|
| 674 | sub markup_acronyms {
|
|---|
| 675 | my $self = shift (@_);
|
|---|
| 676 | my ($text, $doc_obj, $thissection) = @_;
|
|---|
| 677 | my $outhandle = $self->{'outhandle'};
|
|---|
| 678 |
|
|---|
| 679 | print $outhandle " marking up acronyms ...\n"
|
|---|
| 680 | if ($self->{'verbosity'} > 2);
|
|---|
| 681 |
|
|---|
| 682 | #self is passed in to check for verbosity ...
|
|---|
| 683 | $text = &acronym::markup_acronyms($text, $self);
|
|---|
| 684 |
|
|---|
| 685 | print $outhandle " done marking up acronyms. \n"
|
|---|
| 686 | if ($self->{'verbosity'} > 2);
|
|---|
| 687 |
|
|---|
| 688 | return $text;
|
|---|
| 689 | }
|
|---|
| 690 |
|
|---|
| 691 | sub compile_stats {
|
|---|
| 692 | my $self = shift(@_);
|
|---|
| 693 | my ($stats) = @_;
|
|---|
| 694 |
|
|---|
| 695 | $stats->{'num_processed'} += $self->{'num_processed'};
|
|---|
| 696 | $stats->{'num_not_processed'} += $self->{'num_not_processed'};
|
|---|
| 697 | $stats->{'num_archives'} += $self->{'num_archives'};
|
|---|
| 698 |
|
|---|
| 699 | }
|
|---|
| 700 |
|
|---|
| 701 | 1;
|
|---|
