| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # ReadTxtFile.pm -- base class for import plugins that have plain text files
|
|---|
| 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) 1999-2005 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 ReadTextFile;
|
|---|
| 27 |
|
|---|
| 28 | use strict;
|
|---|
| 29 | no strict 'subs';
|
|---|
| 30 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 31 |
|
|---|
| 32 | use Encode;
|
|---|
| 33 |
|
|---|
| 34 | use multiread;
|
|---|
| 35 | use encodings;
|
|---|
| 36 | use unicode;
|
|---|
| 37 | use textcat;
|
|---|
| 38 | use doc;
|
|---|
| 39 | use ghtml;
|
|---|
| 40 | use gsprintf 'gsprintf';
|
|---|
| 41 | use util;
|
|---|
| 42 | use FileUtils;
|
|---|
| 43 |
|
|---|
| 44 | use AutoExtractMetadata;
|
|---|
| 45 |
|
|---|
| 46 | sub BEGIN {
|
|---|
| 47 | @ReadTextFile::ISA = ( 'AutoExtractMetadata' );
|
|---|
| 48 | }
|
|---|
| 49 |
|
|---|
| 50 | my $encoding_plus_auto_list =
|
|---|
| 51 | [ { 'name' => "auto",
|
|---|
| 52 | 'desc' => "{ReadTextFile.input_encoding.auto}" } ];
|
|---|
| 53 | push(@{$encoding_plus_auto_list},@{$BasePlugin::encoding_list});
|
|---|
| 54 |
|
|---|
| 55 | my $arguments =
|
|---|
| 56 | [ { 'name' => "input_encoding",
|
|---|
| 57 | 'desc' => "{ReadTextFile.input_encoding}",
|
|---|
| 58 | 'type' => "enum",
|
|---|
| 59 | 'list' => $encoding_plus_auto_list,
|
|---|
| 60 | 'reqd' => "no" ,
|
|---|
| 61 | 'deft' => "auto" } ,
|
|---|
| 62 | { 'name' => "default_encoding",
|
|---|
| 63 | 'desc' => "{ReadTextFile.default_encoding}",
|
|---|
| 64 | 'type' => "enum",
|
|---|
| 65 | 'list' => $BasePlugin::encoding_list,
|
|---|
| 66 | 'reqd' => "no",
|
|---|
| 67 | 'deft' => "utf8" },
|
|---|
| 68 | { 'name' => "extract_language",
|
|---|
| 69 | 'desc' => "{ReadTextFile.extract_language}",
|
|---|
| 70 | 'type' => "flag",
|
|---|
| 71 | 'reqd' => "no" },
|
|---|
| 72 | { 'name' => "default_language",
|
|---|
| 73 | 'desc' => "{ReadTextFile.default_language}",
|
|---|
| 74 | 'type' => "string",
|
|---|
| 75 | 'deft' => "en",
|
|---|
| 76 | 'reqd' => "no" }
|
|---|
| 77 | ];
|
|---|
| 78 |
|
|---|
| 79 |
|
|---|
| 80 | my $options = { 'name' => "ReadTextFile",
|
|---|
| 81 | 'desc' => "{ReadTextFile.desc}",
|
|---|
| 82 | 'abstract' => "yes",
|
|---|
| 83 | 'inherits' => "no",
|
|---|
| 84 | 'args' => $arguments };
|
|---|
| 85 |
|
|---|
| 86 |
|
|---|
| 87 |
|
|---|
| 88 | sub new {
|
|---|
| 89 | my $class = shift (@_);
|
|---|
| 90 | my ($pluginlist,$inputargs,$hashArgOptLists, $auxiliary) = @_;
|
|---|
| 91 | push(@$pluginlist, $class);
|
|---|
| 92 |
|
|---|
| 93 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 94 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 95 |
|
|---|
| 96 | my $self = new AutoExtractMetadata($pluginlist, $inputargs, $hashArgOptLists, $auxiliary);
|
|---|
| 97 |
|
|---|
| 98 | return bless $self, $class;
|
|---|
| 99 |
|
|---|
| 100 | }
|
|---|
| 101 |
|
|---|
| 102 |
|
|---|
| 103 |
|
|---|
| 104 | # The ReadTextFile read_into_doc_obj() function. This function does all the
|
|---|
| 105 | # right things to make general options work for a given plugin. It reads in
|
|---|
| 106 | # a file and sets up a slew of metadata all saved in doc_obj, which
|
|---|
| 107 | # it then returns as part of a tuple (process_status,doc_obj)
|
|---|
| 108 | #
|
|---|
| 109 | # Much of this functionality used to reside in read, but it was broken
|
|---|
| 110 | # down into a supporting routine to make the code more flexible.
|
|---|
| 111 | #
|
|---|
| 112 | # recursive plugins (e.g. RecPlug) and specialized plugins like those
|
|---|
| 113 | # capable of processing many documents within a single file (e.g.
|
|---|
| 114 | # GMLPlug) will normally want to implement their own version of
|
|---|
| 115 | # read_into_doc_obj()
|
|---|
| 116 | #
|
|---|
| 117 | # Note that $base_dir might be "" and that $file might
|
|---|
| 118 | # include directories
|
|---|
| 119 | sub read_into_doc_obj {
|
|---|
| 120 | my $self = shift (@_);
|
|---|
| 121 | my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
|
|---|
| 122 |
|
|---|
| 123 | my $outhandle = $self->{'outhandle'};
|
|---|
| 124 | # should we move this to read? What about secondary plugins?
|
|---|
| 125 | print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
|
|---|
| 126 | my $pp_file = &util::prettyprint_file($base_dir,$file,$gli);
|
|---|
| 127 | print $outhandle "$self->{'plugin_type'} processing $pp_file\n"
|
|---|
| 128 | if $self->{'verbosity'} > 1;
|
|---|
| 129 |
|
|---|
| 130 | my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
|
|---|
| 131 |
|
|---|
| 132 | # Do encoding stuff
|
|---|
| 133 | my ($language, $content_encoding) = $self->textcat_get_language_encoding ($filename_full_path);
|
|---|
| 134 | if ($self->{'verbosity'} > 2) {
|
|---|
| 135 | print $outhandle "ReadTextFile: reading $file as ($content_encoding,$language)\n";
|
|---|
| 136 | }
|
|---|
| 137 |
|
|---|
| 138 | # create a new document
|
|---|
| 139 | my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
|
|---|
| 140 | my $top_section = $doc_obj->get_top_section();
|
|---|
| 141 |
|
|---|
| 142 | # this should look at the plugin option too...
|
|---|
| 143 | $doc_obj->add_utf8_metadata($top_section, "Plugin", "$self->{'plugin_type'}");
|
|---|
| 144 | $doc_obj->add_utf8_metadata($top_section, "FileSize", &FileUtils::fileSize($filename_full_path));
|
|---|
| 145 |
|
|---|
| 146 | my $plugin_filename_encoding = $self->{'filename_encoding'};
|
|---|
| 147 | my $filename_encoding = $self->deduce_filename_encoding($file,$metadata,$plugin_filename_encoding);
|
|---|
| 148 | $self->set_Source_metadata($doc_obj, $filename_full_path, $filename_encoding);
|
|---|
| 149 |
|
|---|
| 150 | $doc_obj->add_utf8_metadata($top_section, "Language", $language);
|
|---|
| 151 | $doc_obj->add_utf8_metadata($top_section, "Encoding", $content_encoding);
|
|---|
| 152 |
|
|---|
| 153 | # read in file ($text will be in utf8)
|
|---|
| 154 | my $text = "";
|
|---|
| 155 | $self->read_file ($filename_full_path, $content_encoding, $language, \$text);
|
|---|
| 156 |
|
|---|
| 157 | if (!length ($text)) {
|
|---|
| 158 | if ($gli) {
|
|---|
| 159 | print STDERR "<ProcessingError n='$file' r='File contains no text'>\n";
|
|---|
| 160 | }
|
|---|
| 161 | gsprintf($outhandle, "$self->{'plugin_type'}: {ReadTextFile.file_has_no_text}\n", $filename_full_path) if $self->{'verbosity'};
|
|---|
| 162 |
|
|---|
| 163 | my $failhandle = $self->{'failhandle'};
|
|---|
| 164 | gsprintf($failhandle, "$file: " . ref($self) . ": {ReadTextFile.empty_file}\n");
|
|---|
| 165 | # print $failhandle "$file: " . ref($self) . ": file contains no text\n";
|
|---|
| 166 | $self->{'num_not_processed'} ++;
|
|---|
| 167 |
|
|---|
| 168 | return (0,undef); # what should we return here?? error but don't want to pass it on
|
|---|
| 169 | }
|
|---|
| 170 |
|
|---|
| 171 | # do plugin specific processing of doc_obj
|
|---|
| 172 | unless (defined ($self->process (\$text, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli))) {
|
|---|
| 173 | $text = '';
|
|---|
| 174 | undef $text;
|
|---|
| 175 | print STDERR "<ProcessingError n='$file'>\n" if ($gli);
|
|---|
| 176 | return (-1,undef);
|
|---|
| 177 | }
|
|---|
| 178 | $text='';
|
|---|
| 179 | undef $text;
|
|---|
| 180 |
|
|---|
| 181 | # include any metadata passed in from previous plugins
|
|---|
| 182 | # note that this metadata is associated with the top level section
|
|---|
| 183 | $self->add_associated_files($doc_obj, $filename_full_path);
|
|---|
| 184 | $self->extra_metadata ($doc_obj, $top_section, $metadata);
|
|---|
| 185 |
|
|---|
| 186 | # do any automatic metadata extraction
|
|---|
| 187 | $self->auto_extract_metadata ($doc_obj);
|
|---|
| 188 |
|
|---|
| 189 |
|
|---|
| 190 | # if we haven't found any Title so far, assign one
|
|---|
| 191 | $self->title_fallback($doc_obj,$top_section,$filename_no_path);
|
|---|
| 192 |
|
|---|
| 193 | $self->add_OID($doc_obj);
|
|---|
| 194 |
|
|---|
| 195 | return (1,$doc_obj);
|
|---|
| 196 | }
|
|---|
| 197 |
|
|---|
| 198 | # uses the multiread package to read in the entire file pointed to
|
|---|
| 199 | # by filename and loads the resulting text into $$textref. Input text
|
|---|
| 200 | # may be in any of the encodings handled by multiread, output text
|
|---|
| 201 | # will be in utf8
|
|---|
| 202 | sub read_file {
|
|---|
| 203 | my $self = shift (@_);
|
|---|
| 204 | my ($filename, $encoding, $language, $textref) = @_;
|
|---|
| 205 |
|
|---|
| 206 | if (!&FileUtils::canRead($filename))
|
|---|
| 207 | {
|
|---|
| 208 | my $outhandle = $self->{'outhandle'};
|
|---|
| 209 | gsprintf($outhandle, "{ReadTextFile.read_denied}\n", $filename) if $self->{'verbosity'};
|
|---|
| 210 | # print $outhandle "Read permission denied for $filename\n" if $self->{'verbosity'};
|
|---|
| 211 | return;
|
|---|
| 212 | }
|
|---|
| 213 |
|
|---|
| 214 | $$textref = "";
|
|---|
| 215 | # use the util::file_* functions to provide some extra smarts to the open
|
|---|
| 216 | # command in order to support protocols like http: or hdfs:
|
|---|
| 217 | my $fh;
|
|---|
| 218 | if (!&FileUtils::openFileHandle($filename, '<', \$fh))
|
|---|
| 219 | {
|
|---|
| 220 | gsprintf(STDERR, "ReadTextFile::read_file {ReadTextFile.could_not_open_for_reading} ($!)\n", $filename);
|
|---|
| 221 | die "\n";
|
|---|
| 222 | }
|
|---|
| 223 |
|
|---|
| 224 | if ($encoding eq "ascii")
|
|---|
| 225 | {
|
|---|
| 226 | # fast file 'slurp' implementation
|
|---|
| 227 | sysread($fh, $$textref, &FileUtils::fileSize($filename));
|
|---|
| 228 | # the old slow way of reading in a file
|
|---|
| 229 | #undef $/;
|
|---|
| 230 | #$$textref = <FILE>;
|
|---|
| 231 | #$/ = "\n";
|
|---|
| 232 | }
|
|---|
| 233 | else
|
|---|
| 234 | {
|
|---|
| 235 | my $reader = new multiread();
|
|---|
| 236 | $reader->set_handle ($fh);
|
|---|
| 237 | $reader->set_encoding ($encoding);
|
|---|
| 238 | $reader->read_file ($textref);
|
|---|
| 239 | }
|
|---|
| 240 |
|
|---|
| 241 | # At this point $$textref is a binary byte string
|
|---|
| 242 | # => turn it into a Unicode aware string, so full
|
|---|
| 243 | # Unicode aware pattern matching can be used.
|
|---|
| 244 | # For instance: 's/\x{0101}//g' or '[[:upper:]]'
|
|---|
| 245 | #
|
|---|
| 246 | $$textref = decode("utf8", $$textref);
|
|---|
| 247 |
|
|---|
| 248 | &FileUtils::closeFileHandle($filename, \$fh);
|
|---|
| 249 | }
|
|---|
| 250 |
|
|---|
| 251 |
|
|---|
| 252 | # Not currently used
|
|---|
| 253 | sub UNUSED_read_file_usingPerlsEncodeModule {
|
|---|
| 254 | ##sub read_file {
|
|---|
| 255 | my $self = shift (@_);
|
|---|
| 256 | my ($filename, $encoding, $language, $textref) = @_;
|
|---|
| 257 |
|
|---|
| 258 | if (!-r $filename)
|
|---|
| 259 | {
|
|---|
| 260 | my $outhandle = $self->{'outhandle'};
|
|---|
| 261 | gsprintf($outhandle, "{ReadTextFile.read_denied}\n", $filename) if $self->{'verbosity'};
|
|---|
| 262 | # print $outhandle "Read permission denied for $filename\n" if $self->{'verbosity'};
|
|---|
| 263 | return;
|
|---|
| 264 | }
|
|---|
| 265 | $$textref = "";
|
|---|
| 266 | if (!open (FILE, $filename)) {
|
|---|
| 267 | gsprintf(STDERR, "ReadTextFile::read_file {ReadTextFile.could_not_open_f
|
|---|
| 268 | or_reading} ($!)\n", $filename);
|
|---|
| 269 | die "\n";
|
|---|
| 270 | }
|
|---|
| 271 |
|
|---|
| 272 | my $store_slash = $/;
|
|---|
| 273 | undef $/;
|
|---|
| 274 | my $text = <FILE>;
|
|---|
| 275 | $/ = $store_slash;
|
|---|
| 276 |
|
|---|
| 277 | $$textref = decode($encoding,$text);
|
|---|
| 278 |
|
|---|
| 279 | close FILE;
|
|---|
| 280 | }
|
|---|
| 281 |
|
|---|
| 282 |
|
|---|
| 283 | sub read_file_no_decoding {
|
|---|
| 284 | my $self = shift (@_);
|
|---|
| 285 | my ($filename, $textref) = @_;
|
|---|
| 286 |
|
|---|
| 287 | if (!-r $filename)
|
|---|
| 288 | {
|
|---|
| 289 | my $outhandle = $self->{'outhandle'};
|
|---|
| 290 | gsprintf($outhandle, "{ReadTextFile.read_denied}\n", $filename) if $self->{'verbosity'};
|
|---|
| 291 | # print $outhandle "Read permission denied for $filename\n" if $self->{'verbosity'};
|
|---|
| 292 | return;
|
|---|
| 293 | }
|
|---|
| 294 | $$textref = "";
|
|---|
| 295 | if (!open (FILE, $filename)) {
|
|---|
| 296 | gsprintf(STDERR, "ReadTextFile::read_file {ReadTextFile.could_not_open_for_reading} ($!)\n", $filename);
|
|---|
| 297 | die "\n";
|
|---|
| 298 | }
|
|---|
| 299 |
|
|---|
| 300 | my $reader = new multiread();
|
|---|
| 301 | $reader->set_handle ('ReadTextFile::FILE');
|
|---|
| 302 | $reader->read_file_no_decoding ($textref);
|
|---|
| 303 |
|
|---|
| 304 | $self->{'reader'} = $reader;
|
|---|
| 305 |
|
|---|
| 306 | close FILE;
|
|---|
| 307 | }
|
|---|
| 308 |
|
|---|
| 309 |
|
|---|
| 310 | sub decode_text {
|
|---|
| 311 | my $self = shift (@_);
|
|---|
| 312 | my ($raw_text, $encoding, $language, $textref) = @_;
|
|---|
| 313 |
|
|---|
| 314 | my $reader = $self->{'reader'};
|
|---|
| 315 | if (!defined $reader) {
|
|---|
| 316 | gsprintf(STDERR, "ReadTextFile::decode_text needs to call ReadTextFile::read_file_no_decoding first\n");
|
|---|
| 317 | }
|
|---|
| 318 | else {
|
|---|
| 319 | $reader->set_encoding($encoding);
|
|---|
| 320 | $reader->decode_text($raw_text,$textref);
|
|---|
| 321 |
|
|---|
| 322 | # At this point $$textref is a binary byte string
|
|---|
| 323 | # => turn it into a Unicode aware string, so full
|
|---|
| 324 | # Unicode aware pattern matching can be used.
|
|---|
| 325 | # For instance: 's/\x{0101}//g' or '[[:upper:]]'
|
|---|
| 326 |
|
|---|
| 327 | $$textref = decode("utf8",$$textref);
|
|---|
| 328 | }
|
|---|
| 329 | }
|
|---|
| 330 |
|
|---|
| 331 |
|
|---|
| 332 | sub textcat_get_language_encoding {
|
|---|
| 333 | my $self = shift (@_);
|
|---|
| 334 | my ($filename) = @_;
|
|---|
| 335 |
|
|---|
| 336 | my ($language, $encoding, $extracted_encoding);
|
|---|
| 337 | if ($self->{'input_encoding'} eq "auto") {
|
|---|
| 338 | # use textcat to automatically work out the input encoding and language
|
|---|
| 339 | ($language, $encoding) = $self->get_language_encoding ($filename);
|
|---|
| 340 | } elsif ($self->{'extract_language'}) {
|
|---|
| 341 | # use textcat to get language metadata
|
|---|
| 342 | ($language, $extracted_encoding) = $self->get_language_encoding ($filename);
|
|---|
| 343 | $encoding = $self->{'input_encoding'};
|
|---|
| 344 | # don't print this message for english... english in utf8 is identical
|
|---|
| 345 | # to english in iso-8859-1 (except for some punctuation). We don't have
|
|---|
| 346 | # a language model for en_utf8, so textcat always says iso-8859-1!
|
|---|
| 347 | if ($extracted_encoding ne $encoding && $language ne "en" && $self->{'verbosity'}) {
|
|---|
| 348 | my $plugin_name = ref ($self);
|
|---|
| 349 | my $outhandle = $self->{'outhandle'};
|
|---|
| 350 | gsprintf($outhandle, "$plugin_name: {ReadTextFile.wrong_encoding}\n", $filename, $encoding, $extracted_encoding);
|
|---|
| 351 | }
|
|---|
| 352 | } else {
|
|---|
| 353 | $language = $self->{'default_language'};
|
|---|
| 354 | $encoding = $self->{'input_encoding'};
|
|---|
| 355 | }
|
|---|
| 356 |
|
|---|
| 357 | # print STDERR "**** language encoding of contents of file $filename:\n\t****$language $encoding\n";
|
|---|
| 358 |
|
|---|
| 359 | return ($language, $encoding);
|
|---|
| 360 | }
|
|---|
| 361 |
|
|---|
| 362 |
|
|---|
| 363 | # Uses textcat to work out the encoding and language of the text in
|
|---|
| 364 | # $filename. All html tags are removed before processing.
|
|---|
| 365 | # returns an array containing "language" and "encoding"
|
|---|
| 366 | sub get_language_encoding {
|
|---|
| 367 | my $self = shift (@_);
|
|---|
| 368 | my ($filename) = @_;
|
|---|
| 369 | my $outhandle = $self->{'outhandle'};
|
|---|
| 370 | my $unicode_format = "";
|
|---|
| 371 | my $best_language = "";
|
|---|
| 372 | my $best_encoding = "";
|
|---|
| 373 |
|
|---|
| 374 | # read in file
|
|---|
| 375 | my $fh;
|
|---|
| 376 | if (!&FileUtils::openFileHandle($filename, '<', \$fh))
|
|---|
| 377 | {
|
|---|
| 378 | gsprintf(STDERR, "ReadTextFile::get_language_encoding {ReadTextFile.could_not_open_for_reading} ($!)\n", $filename);
|
|---|
| 379 | # this is a pretty bad error, but try to continue anyway
|
|---|
| 380 | return ($self->{'default_language'}, $self->{'input_encoding'});
|
|---|
| 381 | }
|
|---|
| 382 | my $text;
|
|---|
| 383 | sysread($fh, $text, &FileUtils::fileSize($filename));
|
|---|
| 384 | #undef $/;
|
|---|
| 385 | #my $text = <FILE>;
|
|---|
| 386 | #$/ = "\n";
|
|---|
| 387 | close $fh;
|
|---|
| 388 |
|
|---|
| 389 | # check if first few bytes have a Byte Order Marker
|
|---|
| 390 | my $bom=substr($text,0,2); # check 16bit unicode
|
|---|
| 391 | if ($bom eq "\xff\xfe") { # little endian 16bit unicode
|
|---|
| 392 | $unicode_format="unicode";
|
|---|
| 393 | } elsif ($bom eq "\xfe\xff") { # big endian 16bit unicode
|
|---|
| 394 | $unicode_format="unicode";
|
|---|
| 395 | } else {
|
|---|
| 396 | $bom=substr($text,0,3); # check utf-8
|
|---|
| 397 | if ($bom eq "\xef\xbb\xbf") { # utf-8 coded FEFF bom
|
|---|
| 398 | $unicode_format="utf8";
|
|---|
| 399 | # } elsif ($bom eq "\xef\xbf\xbe") { # utf-8 coded FFFE bom. Error!?
|
|---|
| 400 | # $unicode_format="utf8";
|
|---|
| 401 | }
|
|---|
| 402 | }
|
|---|
| 403 |
|
|---|
| 404 | my $found_html_encoding = 0;
|
|---|
| 405 | # handle html files specially
|
|---|
| 406 | # XXX this doesn't match plugins derived from HTMLPlug (except ConvertTo)
|
|---|
| 407 | if (ref($self) eq 'HTMLPlugin' ||
|
|---|
| 408 | (exists $self->{'converted_to'} && $self->{'converted_to'} eq 'HTML')){
|
|---|
| 409 |
|
|---|
| 410 | # remove comments in head, including multiline ones, so that we don't match on
|
|---|
| 411 | # inactive tags (those that are nested inside comments)
|
|---|
| 412 | my ($head) = ($text =~ m/<head>(.*)<\/head>/si);
|
|---|
| 413 | $head = "" unless defined $head; # some files are not proper HTML eg php files
|
|---|
| 414 | $head =~ s/<!--.*?-->//sg;
|
|---|
| 415 |
|
|---|
| 416 | # remove <title>stuff</title> -- as titles tend often to be in English
|
|---|
| 417 | # for foreign language documents
|
|---|
| 418 | $text =~ s!<title>.*?</title>!!si;
|
|---|
| 419 |
|
|---|
| 420 | # see if this html file specifies its encoding
|
|---|
| 421 | if ($text =~ /^<\?xml.*encoding="(.+?)"/) {
|
|---|
| 422 | $best_encoding = $1;
|
|---|
| 423 | }
|
|---|
| 424 | # check the meta http-equiv charset tag
|
|---|
| 425 | elsif ($head =~ m/<meta http-equiv.*content-type.*charset=(.+?)\"/si) {
|
|---|
| 426 | $best_encoding = $1;
|
|---|
| 427 | }
|
|---|
| 428 | if ($best_encoding) { # we extracted an encoding
|
|---|
| 429 | $best_encoding =~ s/-+/_/g;
|
|---|
| 430 | $best_encoding = lc($best_encoding); # lowercase
|
|---|
| 431 | if ($best_encoding eq "utf_8") { $best_encoding = "utf8" }
|
|---|
| 432 | $found_html_encoding = 1;
|
|---|
| 433 | # We shouldn't be modifying this here!!
|
|---|
| 434 | #$self->{'input_encoding'} = $best_encoding;
|
|---|
| 435 | }
|
|---|
| 436 |
|
|---|
| 437 | # remove all HTML tags
|
|---|
| 438 | $text =~ s/<[^>]*>//sg;
|
|---|
| 439 | }
|
|---|
| 440 |
|
|---|
| 441 | # don't need to do textcat if we know the encoding now AND don't need to extract language
|
|---|
| 442 | if($found_html_encoding && !$self->{'extract_language'}) { # encoding specified in html file
|
|---|
| 443 | $best_language = $self->{'default_language'};
|
|---|
| 444 | }
|
|---|
| 445 |
|
|---|
| 446 | else { # need to use textcat to get either the language, or get both language and encoding
|
|---|
| 447 | $self->{'textcat'} = new textcat() if (!defined($self->{'textcat'}));
|
|---|
| 448 |
|
|---|
| 449 | if($found_html_encoding) { # know encoding, find language by limiting search to known encoding
|
|---|
| 450 | my $results = $self->{'textcat'}->classify_contents_for_encoding(\$text, $filename, $best_encoding);
|
|---|
| 451 |
|
|---|
| 452 | my $language;
|
|---|
| 453 | ($language) = $results->[0] =~ m/^([^-]*)(?:-(?:.*))?$/ if (scalar @$results > 0);
|
|---|
| 454 |
|
|---|
| 455 | if (!defined $language || scalar @$results > 3) {
|
|---|
| 456 | # if there were too many results even when restricting results by encoding,
|
|---|
| 457 | # or if there were no results, use default language with the known encoding
|
|---|
| 458 | $best_language = $self->use_default_language($filename);
|
|---|
| 459 | }
|
|---|
| 460 | else { # fewer than 3 results means textcat is more certain, use the first result
|
|---|
| 461 | $best_language = $language;
|
|---|
| 462 | }
|
|---|
| 463 | }
|
|---|
| 464 | else { # don't know encoding or language yet, therefore we use textcat
|
|---|
| 465 | my $results = $self->{'textcat'}->classify_contents(\$text, $filename);
|
|---|
| 466 |
|
|---|
| 467 | # if textcat returns 3 or less possibilities we'll use the first one in the list
|
|---|
| 468 | if (scalar @$results <= 3) { # results will be > 0 when we don't constrain textcat by an encoding
|
|---|
| 469 | my ($language, $encoding) = $results->[0] =~ m/^([^-]*)(?:-(.*))?$/;
|
|---|
| 470 |
|
|---|
| 471 | $language = $self->use_default_language($filename) unless defined $language;
|
|---|
| 472 | $encoding = $self->use_default_encoding($filename) unless defined $encoding;
|
|---|
| 473 |
|
|---|
| 474 | $best_language = $language;
|
|---|
| 475 | $best_encoding = $encoding;
|
|---|
| 476 | }
|
|---|
| 477 | else { # if (scalar @$results > 3) {
|
|---|
| 478 | if ($unicode_format) { # in case the first had a BOM
|
|---|
| 479 | $best_encoding=$unicode_format;
|
|---|
| 480 | }
|
|---|
| 481 | else {
|
|---|
| 482 | # Find the most frequent encoding in the textcat results returned
|
|---|
| 483 | # Returns "" if there's no encoding more frequent than another
|
|---|
| 484 | $best_encoding = $self->{'textcat'}->most_frequent_encoding($results);
|
|---|
| 485 | }
|
|---|
| 486 |
|
|---|
| 487 | if ($best_encoding eq "") { # encoding still not set, use defaults
|
|---|
| 488 | $best_language = $self->use_default_language($filename);
|
|---|
| 489 | $best_encoding = $self->use_default_encoding($filename);
|
|---|
| 490 | }
|
|---|
| 491 | elsif (!$self->{'extract_language'}) { # know encoding but don't need to discover language
|
|---|
| 492 | $best_language = $self->use_default_language($filename);
|
|---|
| 493 | }
|
|---|
| 494 | else { # textcat again using the most frequent encoding or the $unicode_format set above
|
|---|
| 495 | $results = $self->{'textcat'}->classify_contents_for_encoding(\$text, $filename, $best_encoding);
|
|---|
| 496 | my $language;
|
|---|
| 497 | ($language) = $results->[0] =~ m/^([^-]*)(?:-(.*))?$/ if (scalar @$results > 0);
|
|---|
| 498 | if (!defined $language || scalar @$results > 3) {
|
|---|
| 499 | # if no result or too many results, use default language for the encoding previously found
|
|---|
| 500 | $best_language = $self->use_default_language($filename);
|
|---|
| 501 | }
|
|---|
| 502 | else { # fewer than 3 results, use the language of the first result
|
|---|
| 503 | $best_language = $language;
|
|---|
| 504 | }
|
|---|
| 505 | }
|
|---|
| 506 | }
|
|---|
| 507 | }
|
|---|
| 508 | }
|
|---|
| 509 |
|
|---|
| 510 | if($best_encoding eq "" || $best_language eq "") {
|
|---|
| 511 | print STDERR "****Shouldn't happen: encoding and/or language still not set. Using defaults.\n";
|
|---|
| 512 | $best_encoding = $self->use_default_encoding($filename) if $best_encoding eq "";
|
|---|
| 513 | $best_language = $self->use_default_language($filename) if $best_language eq "";
|
|---|
| 514 | }
|
|---|
| 515 | # print STDERR "****Content language: $best_language; Encoding: $best_encoding.\n";
|
|---|
| 516 |
|
|---|
| 517 |
|
|---|
| 518 | if ($best_encoding =~ /^iso_8859/ && &unicode::check_is_utf8($text)) {
|
|---|
| 519 | # the text is valid utf8, so assume that's the real encoding
|
|---|
| 520 | # (since textcat is based on probabilities)
|
|---|
| 521 | $best_encoding = 'utf8';
|
|---|
| 522 | }
|
|---|
| 523 |
|
|---|
| 524 | # check for equivalents where textcat doesn't have some encodings...
|
|---|
| 525 | # eg MS versions of standard encodings
|
|---|
| 526 | if ($best_encoding =~ /^iso_8859_(\d+)/) {
|
|---|
| 527 | my $iso = $1; # which variant of the iso standard?
|
|---|
| 528 | # iso-8859 sets don't use chars 0x80-0x9f, windows codepages do
|
|---|
| 529 | if ($text =~ /[\x80-\x9f]/) {
|
|---|
| 530 | # Western Europe
|
|---|
| 531 | if ($iso == 1 or $iso == 15) { $best_encoding = 'windows_1252' }
|
|---|
| 532 | elsif ($iso == 2) {$best_encoding = 'windows_1250'} # Central Europe
|
|---|
| 533 | elsif ($iso == 5) {$best_encoding = 'windows_1251'} # Cyrillic
|
|---|
| 534 | elsif ($iso == 6) {$best_encoding = 'windows_1256'} # Arabic
|
|---|
| 535 | elsif ($iso == 7) {$best_encoding = 'windows_1253'} # Greek
|
|---|
| 536 | elsif ($iso == 8) {$best_encoding = 'windows_1255'} # Hebrew
|
|---|
| 537 | elsif ($iso == 9) {$best_encoding = 'windows_1254'} # Turkish
|
|---|
| 538 | }
|
|---|
| 539 | }
|
|---|
| 540 |
|
|---|
| 541 | if ($best_encoding !~ /^(ascii|utf8|unicode)$/ &&
|
|---|
| 542 | !defined $encodings::encodings->{$best_encoding}) {
|
|---|
| 543 | if ($self->{'verbosity'}) {
|
|---|
| 544 | gsprintf($outhandle, "ReadTextFile: {ReadTextFile.unsupported_encoding}\n",
|
|---|
| 545 | $filename, $best_encoding, $self->{'default_encoding'});
|
|---|
| 546 | }
|
|---|
| 547 | $best_encoding = $self->{'default_encoding'};
|
|---|
| 548 | }
|
|---|
| 549 |
|
|---|
| 550 | return ($best_language, $best_encoding);
|
|---|
| 551 | }
|
|---|
| 552 |
|
|---|
| 553 |
|
|---|
| 554 | sub use_default_language {
|
|---|
| 555 | my $self = shift (@_);
|
|---|
| 556 | my ($filename) = @_;
|
|---|
| 557 |
|
|---|
| 558 | if ($self->{'verbosity'}>2) {
|
|---|
| 559 | gsprintf($self->{'outhandle'},
|
|---|
| 560 | "ReadTextFile: {ReadTextFile.could_not_extract_language}\n",
|
|---|
| 561 | $filename, $self->{'default_language'});
|
|---|
| 562 | }
|
|---|
| 563 | return $self->{'default_language'};
|
|---|
| 564 | }
|
|---|
| 565 |
|
|---|
| 566 | sub use_default_encoding {
|
|---|
| 567 | my $self = shift (@_);
|
|---|
| 568 | my ($filename) = @_;
|
|---|
| 569 |
|
|---|
| 570 | if ($self->{'verbosity'}>2) {
|
|---|
| 571 | gsprintf($self->{'outhandle'},
|
|---|
| 572 | "ReadTextFile: {ReadTextFile.could_not_extract_encoding}\n",
|
|---|
| 573 | $filename, $self->{'default_encoding'});
|
|---|
| 574 | }
|
|---|
| 575 | return $self->{'default_encoding'};
|
|---|
| 576 | }
|
|---|
| 577 |
|
|---|
| 578 | # Overridden by exploding plugins (eg. ISISPlug)
|
|---|
| 579 | sub clean_up_after_exploding
|
|---|
| 580 | {
|
|---|
| 581 | my $self = shift(@_);
|
|---|
| 582 | }
|
|---|
| 583 |
|
|---|
| 584 |
|
|---|
| 585 | 1;
|
|---|
