Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
16.76% |
58 / 346 |
|
11.54% |
3 / 26 |
CRAP | |
0.00% |
0 / 1 |
| GettextFormat | |
16.81% |
58 / 345 |
|
11.54% |
3 / 26 |
5513.66 | |
0.00% |
0 / 1 |
| supportsFuzzy | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getFileExtensions | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| setOfflineMode | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| read | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| readFromVariable | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
12 | |||
| parseGettext | |
0.00% |
0 / 52 |
|
0.00% |
0 / 1 |
182 | |||
| parseGettextSection | |
0.00% |
0 / 44 |
|
0.00% |
0 / 1 |
240 | |||
| processGettextPluralMessage | |
0.00% |
0 / 10 |
|
0.00% |
0 / 1 |
20 | |||
| parseFlags | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
6 | |||
| expectKeyword | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| generateKeyFromItem | |
100.00% |
16 / 16 |
|
100.00% |
1 / 1 |
3 | |||
| processData | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
2 | |||
| handleWhitespace | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
20 | |||
| formatForWiki | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| parseHeaderTags | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
12 | |||
| writeReal | |
0.00% |
0 / 34 |
|
0.00% |
0 / 1 |
30 | |||
| doGettextHeader | |
0.00% |
0 / 45 |
|
0.00% |
0 / 1 |
42 | |||
| doAuthors | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
6 | |||
| formatMessageBlock | |
54.05% |
20 / 37 |
|
0.00% |
0 / 1 |
25.97 | |||
| formatTime | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| getPotTime | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
6 | |||
| formatDocumentation | |
22.22% |
2 / 9 |
|
0.00% |
0 / 1 |
11.53 | |||
| escape | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
1 | |||
| shouldOverwrite | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
1 | |||
| getExtraSchema | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
2 | |||
| isContentEqual | |
80.00% |
12 / 15 |
|
0.00% |
0 / 1 |
7.39 | |||
| 1 | <?php |
| 2 | declare( strict_types = 1 ); |
| 3 | |
| 4 | namespace MediaWiki\Extension\Translate\FileFormatSupport; |
| 5 | |
| 6 | use InvalidArgumentException; |
| 7 | use MediaWiki\Extension\Translate\LogNames; |
| 8 | use MediaWiki\Extension\Translate\MessageGroupConfiguration\MetaYamlSchemaExtender; |
| 9 | use MediaWiki\Extension\Translate\MessageLoading\Message; |
| 10 | use MediaWiki\Extension\Translate\MessageLoading\MessageCollection; |
| 11 | use MediaWiki\Extension\Translate\Services; |
| 12 | use MediaWiki\Extension\Translate\Utilities\GettextPlural; |
| 13 | use MediaWiki\Extension\Translate\Utilities\Utilities; |
| 14 | use MediaWiki\Language\LanguageCode; |
| 15 | use MediaWiki\Logger\LoggerFactory; |
| 16 | use MediaWiki\MediaWikiServices; |
| 17 | use MediaWiki\Specials\SpecialVersion; |
| 18 | use MediaWiki\Title\Title; |
| 19 | use RuntimeException; |
| 20 | |
| 21 | /** |
| 22 | * FileFormat class that implements support for gettext file format. |
| 23 | * |
| 24 | * @author Niklas Laxström |
| 25 | * @author Siebrand Mazeland |
| 26 | * @copyright Copyright © 2008-2010, Niklas Laxström, Siebrand Mazeland |
| 27 | * @license GPL-2.0-or-later |
| 28 | * @ingroup FileFormatSupport |
| 29 | */ |
| 30 | class GettextFormat extends SimpleFormat implements MetaYamlSchemaExtender { |
| 31 | private bool $allowPotMode = false; |
| 32 | private bool $offlineMode = false; |
| 33 | |
| 34 | public function supportsFuzzy(): string { |
| 35 | return 'yes'; |
| 36 | } |
| 37 | |
| 38 | public function getFileExtensions(): array { |
| 39 | return [ '.pot', '.po' ]; |
| 40 | } |
| 41 | |
| 42 | public function setOfflineMode( bool $value ): void { |
| 43 | $this->offlineMode = $value; |
| 44 | } |
| 45 | |
| 46 | /** @inheritDoc */ |
| 47 | public function read( $languageCode ) { |
| 48 | // This is somewhat hacky, but pot mode should only ever be used for the source language. |
| 49 | // See https://phabricator.wikimedia.org/T230361 |
| 50 | $this->allowPotMode = $this->getGroup()->getSourceLanguage() === $languageCode; |
| 51 | |
| 52 | try { |
| 53 | return parent::read( $languageCode ); |
| 54 | } finally { |
| 55 | $this->allowPotMode = false; |
| 56 | } |
| 57 | } |
| 58 | |
| 59 | public function readFromVariable( string $data ): array { |
| 60 | # Authors first |
| 61 | $matches = []; |
| 62 | preg_match_all( '/^#\s*Author:\s*(.*)$/m', $data, $matches ); |
| 63 | $authors = $matches[1]; |
| 64 | |
| 65 | # Then messages and everything else |
| 66 | $parsedData = $this->parseGettext( $data ); |
| 67 | $parsedData['AUTHORS'] = $authors; |
| 68 | |
| 69 | foreach ( $parsedData['MESSAGES'] as $key => $value ) { |
| 70 | if ( $value === '' ) { |
| 71 | unset( $parsedData['MESSAGES'][$key] ); |
| 72 | } |
| 73 | } |
| 74 | |
| 75 | return $parsedData; |
| 76 | } |
| 77 | |
| 78 | private function parseGettext( string $data ): array { |
| 79 | $mangler = $this->group->getMangler(); |
| 80 | $useCtxtAsKey = $this->extra['CtxtAsKey'] ?? false; |
| 81 | $keyAlgorithm = 'simple'; |
| 82 | if ( isset( $this->extra['keyAlgorithm'] ) ) { |
| 83 | $keyAlgorithm = $this->extra['keyAlgorithm']; |
| 84 | } |
| 85 | |
| 86 | $potmode = false; |
| 87 | |
| 88 | // Normalise newlines, to make processing easier |
| 89 | $data = str_replace( "\r\n", "\n", $data ); |
| 90 | |
| 91 | /* Delimit the file into sections, which are separated by two newlines. |
| 92 | * We are permissive and accept more than two. This parsing method isn't |
| 93 | * efficient wrt memory, but was easy to implement */ |
| 94 | $sections = preg_split( '/\n{2,}/', $data ); |
| 95 | |
| 96 | /* First one isn't an actual message. We'll handle it specially below */ |
| 97 | $headerSection = array_shift( $sections ); |
| 98 | /* Since this is the header section, we are only interested in the tags |
| 99 | * and msgid is empty. Somewhere we should extract the header comments |
| 100 | * too */ |
| 101 | $match = $this->expectKeyword( 'msgstr', $headerSection ); |
| 102 | if ( $match !== null ) { |
| 103 | $headerBlock = $this->formatForWiki( $match, 'trim' ); |
| 104 | $headers = $this->parseHeaderTags( $headerBlock ); |
| 105 | |
| 106 | // Check for pot-mode by checking if the header is fuzzy |
| 107 | $flags = $this->parseFlags( $headerSection ); |
| 108 | if ( in_array( 'fuzzy', $flags, true ) ) { |
| 109 | $potmode = $this->allowPotMode; |
| 110 | } |
| 111 | } else { |
| 112 | $message = "Gettext file header was not found:\n\n$data"; |
| 113 | throw new GettextParseException( $message ); |
| 114 | } |
| 115 | |
| 116 | $template = []; |
| 117 | $messages = []; |
| 118 | |
| 119 | // Extract some metadata from headers for easier use |
| 120 | $metadata = []; |
| 121 | if ( isset( $headers['X-Language-Code'] ) ) { |
| 122 | $metadata['code'] = $headers['X-Language-Code']; |
| 123 | } |
| 124 | |
| 125 | if ( isset( $headers['X-Message-Group'] ) ) { |
| 126 | $metadata['group'] = $headers['X-Message-Group']; |
| 127 | } |
| 128 | |
| 129 | /* At this stage we are only interested how many plurals forms we should |
| 130 | * be expecting when parsing the rest of this file. */ |
| 131 | $pluralCount = null; |
| 132 | if ( $potmode ) { |
| 133 | $pluralCount = 2; |
| 134 | } elseif ( isset( $headers['Plural-Forms'] ) ) { |
| 135 | $pluralCount = $metadata['plural'] = GettextPlural::getPluralCount( $headers['Plural-Forms'] ); |
| 136 | } |
| 137 | |
| 138 | $metadata['plural'] = $pluralCount; |
| 139 | |
| 140 | // Then parse the messages |
| 141 | foreach ( $sections as $section ) { |
| 142 | $item = $this->parseGettextSection( $section, $pluralCount ); |
| 143 | if ( $item === null ) { |
| 144 | continue; |
| 145 | } |
| 146 | |
| 147 | if ( $useCtxtAsKey ) { |
| 148 | if ( !isset( $item['ctxt'] ) ) { |
| 149 | error_log( "ctxt missing for: $section" ); |
| 150 | continue; |
| 151 | } |
| 152 | $key = $item['ctxt']; |
| 153 | } else { |
| 154 | $key = $this->generateKeyFromItem( $item, $keyAlgorithm ); |
| 155 | } |
| 156 | |
| 157 | $key = $mangler->mangle( $key ); |
| 158 | $messages[$key] = $potmode ? $item['id'] : $item['str']; |
| 159 | $template[$key] = $item; |
| 160 | } |
| 161 | |
| 162 | return [ |
| 163 | 'MESSAGES' => $messages, |
| 164 | 'EXTRA' => [ |
| 165 | 'TEMPLATE' => $template, |
| 166 | 'METADATA' => $metadata, |
| 167 | 'HEADERS' => $headers, |
| 168 | ], |
| 169 | ]; |
| 170 | } |
| 171 | |
| 172 | private function parseGettextSection( string $section, ?int $pluralCount ): ?array { |
| 173 | if ( trim( $section ) === '' ) { |
| 174 | return null; |
| 175 | } |
| 176 | |
| 177 | /* These inactive sections are of no interest to us. Multiline mode |
| 178 | * is needed because there may be flags or other annoying stuff |
| 179 | * before the commented out sections. |
| 180 | */ |
| 181 | if ( preg_match( '/^#~/m', $section ) ) { |
| 182 | return null; |
| 183 | } |
| 184 | |
| 185 | $item = [ |
| 186 | 'ctxt' => false, |
| 187 | 'id' => '', |
| 188 | 'str' => '', |
| 189 | 'flags' => [], |
| 190 | 'comments' => [], |
| 191 | ]; |
| 192 | |
| 193 | $match = $this->expectKeyword( 'msgid', $section ); |
| 194 | if ( $match !== null ) { |
| 195 | $item['id'] = $this->formatForWiki( $match ); |
| 196 | } else { |
| 197 | throw new RuntimeException( "Unable to parse msgid:\n\n$section" ); |
| 198 | } |
| 199 | |
| 200 | $match = $this->expectKeyword( 'msgctxt', $section ); |
| 201 | if ( $match !== null ) { |
| 202 | $item['ctxt'] = $this->formatForWiki( $match ); |
| 203 | } |
| 204 | |
| 205 | $pluralMessage = false; |
| 206 | $match = $this->expectKeyword( 'msgid_plural', $section ); |
| 207 | if ( $match !== null ) { |
| 208 | $pluralMessage = true; |
| 209 | $plural = $this->formatForWiki( $match ); |
| 210 | $item['id'] = GettextPlural::flatten( [ $item['id'], $plural ] ); |
| 211 | } |
| 212 | |
| 213 | if ( $pluralMessage ) { |
| 214 | $pluralMessageText = $this->processGettextPluralMessage( $pluralCount, $section ); |
| 215 | |
| 216 | // Keep the translation empty if no form has translation |
| 217 | if ( $pluralMessageText !== '' ) { |
| 218 | $item['str'] = $pluralMessageText; |
| 219 | } |
| 220 | } else { |
| 221 | $match = $this->expectKeyword( 'msgstr', $section ); |
| 222 | if ( $match !== null ) { |
| 223 | $item['str'] = $this->formatForWiki( $match ); |
| 224 | } else { |
| 225 | throw new RuntimeException( "Unable to parse msgstr:\n\n$section" ); |
| 226 | } |
| 227 | } |
| 228 | |
| 229 | // Parse flags |
| 230 | $flags = $this->parseFlags( $section ); |
| 231 | foreach ( $flags as $key => $flag ) { |
| 232 | if ( $flag === 'fuzzy' ) { |
| 233 | $item['str'] = TRANSLATE_FUZZY . $item['str']; |
| 234 | unset( $flags[$key] ); |
| 235 | } |
| 236 | } |
| 237 | $item['flags'] = $flags; |
| 238 | |
| 239 | // Rest of the comments |
| 240 | $matches = []; |
| 241 | if ( preg_match_all( '/^#(.?) (.*)$/m', $section, $matches, PREG_SET_ORDER ) ) { |
| 242 | foreach ( $matches as $match ) { |
| 243 | if ( $match[1] !== ',' && !str_starts_with( $match[1], '[Wiki]' ) ) { |
| 244 | $item['comments'][$match[1]][] = $match[2]; |
| 245 | } |
| 246 | } |
| 247 | } |
| 248 | |
| 249 | return $item; |
| 250 | } |
| 251 | |
| 252 | private function processGettextPluralMessage( ?int $pluralCount, string $section ): string { |
| 253 | $actualForms = []; |
| 254 | |
| 255 | for ( $i = 0; $i < $pluralCount; $i++ ) { |
| 256 | $match = $this->expectKeyword( "msgstr\\[$i\\]", $section ); |
| 257 | |
| 258 | if ( $match !== null ) { |
| 259 | $actualForms[] = $this->formatForWiki( $match ); |
| 260 | } else { |
| 261 | $actualForms[] = ''; |
| 262 | error_log( "Plural $i not found, expecting total of $pluralCount for $section" ); |
| 263 | } |
| 264 | } |
| 265 | |
| 266 | if ( array_sum( array_map( 'strlen', $actualForms ) ) > 0 ) { |
| 267 | return GettextPlural::flatten( $actualForms ); |
| 268 | } else { |
| 269 | return ''; |
| 270 | } |
| 271 | } |
| 272 | |
| 273 | private function parseFlags( string $section ): array { |
| 274 | $matches = []; |
| 275 | if ( preg_match( '/^#,(.*)$/mu', $section, $matches ) ) { |
| 276 | return array_map( 'trim', explode( ',', $matches[1] ) ); |
| 277 | } else { |
| 278 | return []; |
| 279 | } |
| 280 | } |
| 281 | |
| 282 | private function expectKeyword( string $name, string $section ): ?string { |
| 283 | /* Catches the multiline textblock that comes after keywords msgid, |
| 284 | * msgstr, msgid_plural, msgctxt. |
| 285 | */ |
| 286 | $poformat = '".*"\n?(^".*"$\n?)*'; |
| 287 | |
| 288 | $matches = []; |
| 289 | if ( preg_match( "/^$name\s($poformat)/mx", $section, $matches ) ) { |
| 290 | return $matches[1]; |
| 291 | } else { |
| 292 | return null; |
| 293 | } |
| 294 | } |
| 295 | |
| 296 | /** |
| 297 | * Generates unique key for each message. Changing this WILL BREAK ALL |
| 298 | * existing pages! |
| 299 | * @param array $item As returned by parseGettextSection |
| 300 | * @param string $algorithm Algorithm used to generate message keys: simple or legacy |
| 301 | */ |
| 302 | public function generateKeyFromItem( array $item, string $algorithm = 'simple' ): string { |
| 303 | $lang = MediaWikiServices::getInstance()->getLanguageFactory()->getLanguage( 'en' ); |
| 304 | |
| 305 | if ( $item['ctxt'] === '' ) { |
| 306 | /* Messages with msgctxt as empty string should be different |
| 307 | * from messages without any msgctxt. To avoid BC break make |
| 308 | * the empty ctxt a special case */ |
| 309 | $hash = sha1( $item['id'] . 'MSGEMPTYCTXT' ); |
| 310 | } else { |
| 311 | $hash = sha1( $item['ctxt'] . $item['id'] ); |
| 312 | } |
| 313 | |
| 314 | if ( $algorithm === 'simple' ) { |
| 315 | $hash = substr( $hash, 0, 6 ); |
| 316 | $snippet = $lang->truncateForDatabase( $item['id'], 30, '' ); |
| 317 | $snippet = str_replace( ' ', '_', trim( $snippet ) ); |
| 318 | } else { // legacy |
| 319 | $legalChars = Title::legalChars(); |
| 320 | $snippet = $item['id']; |
| 321 | $snippet = preg_replace( "/[^$legalChars]/", ' ', $snippet ); |
| 322 | $snippet = preg_replace( '/[:&%\/_]/', ' ', $snippet ); |
| 323 | $snippet = preg_replace( '/ {2,}/', ' ', $snippet ); |
| 324 | $snippet = $lang->truncateForDatabase( $snippet, 30, '' ); |
| 325 | $snippet = str_replace( ' ', '_', trim( $snippet ) ); |
| 326 | } |
| 327 | |
| 328 | return "$hash-$snippet"; |
| 329 | } |
| 330 | |
| 331 | /** |
| 332 | * This method processes the gettext text block format. |
| 333 | */ |
| 334 | private function processData( string $data ): string { |
| 335 | $quotePattern = '/(^"|"$\n?)/m'; |
| 336 | $data = preg_replace( $quotePattern, '', $data ); |
| 337 | return stripcslashes( $data ); |
| 338 | } |
| 339 | |
| 340 | /** |
| 341 | * This method handles the whitespace at the end of the data. |
| 342 | * @throws InvalidArgumentException |
| 343 | */ |
| 344 | private function handleWhitespace( string $data, string $whitespace ): string { |
| 345 | if ( preg_match( '/\s$/', $data ) ) { |
| 346 | if ( $whitespace === 'mark' ) { |
| 347 | $data .= '\\'; |
| 348 | } elseif ( $whitespace === 'trim' ) { |
| 349 | $data = rtrim( $data ); |
| 350 | } else { |
| 351 | // This condition will never happen as long as $whitespace is 'mark' or 'trim' |
| 352 | throw new InvalidArgumentException( "Unknown action for whitespace: $whitespace" ); |
| 353 | } |
| 354 | } |
| 355 | |
| 356 | return $data; |
| 357 | } |
| 358 | |
| 359 | /** |
| 360 | * This parses the Gettext text block format. Since trailing whitespace is |
| 361 | * not allowed in MediaWiki pages, the default action is to append |
| 362 | * \-character at the end of the message. You can also choose to ignore it |
| 363 | * and use the trim action instead. |
| 364 | */ |
| 365 | private function formatForWiki( string $data, string $whitespace = 'mark' ): string { |
| 366 | $data = $this->processData( $data ); |
| 367 | return $this->handleWhitespace( $data, $whitespace ); |
| 368 | } |
| 369 | |
| 370 | private function parseHeaderTags( string $headers ): array { |
| 371 | $tags = []; |
| 372 | foreach ( explode( "\n", $headers ) as $line ) { |
| 373 | if ( !str_contains( $line, ':' ) ) { |
| 374 | error_log( __METHOD__ . ": $line" ); |
| 375 | } |
| 376 | [ $key, $value ] = explode( ':', $line, 2 ); |
| 377 | $tags[trim( $key )] = trim( $value ); |
| 378 | } |
| 379 | |
| 380 | return $tags; |
| 381 | } |
| 382 | |
| 383 | protected function writeReal( MessageCollection $collection ): string { |
| 384 | // FIXME: this should be the source language |
| 385 | $pot = $this->read( 'en' ) ?? []; |
| 386 | $code = $collection->code; |
| 387 | $template = $this->read( $code ) ?? []; |
| 388 | $output = $this->doGettextHeader( $collection, $template['EXTRA'] ?? [] ); |
| 389 | |
| 390 | $pluralRule = GettextPlural::getPluralRule( $code ); |
| 391 | if ( !$pluralRule ) { |
| 392 | $pluralRule = GettextPlural::getPluralRule( 'en' ); |
| 393 | LoggerFactory::getInstance( LogNames::MAIN )->warning( |
| 394 | "T235180: Missing Gettext plural rule for '{languagecode}'", |
| 395 | [ 'languagecode' => $code ] |
| 396 | ); |
| 397 | } |
| 398 | $pluralCount = GettextPlural::getPluralCount( $pluralRule ); |
| 399 | |
| 400 | $documentationLanguageCode = MediaWikiServices::getInstance() |
| 401 | ->getMainConfig() |
| 402 | ->get( 'TranslateDocumentationLanguageCode' ); |
| 403 | $documentationCollection = null; |
| 404 | if ( is_string( $documentationLanguageCode ) ) { |
| 405 | $documentationCollection = clone $collection; |
| 406 | $documentationCollection->resetForNewLanguage( $documentationLanguageCode ); |
| 407 | $documentationCollection->loadTranslations(); |
| 408 | } |
| 409 | |
| 410 | /** @var Message $m */ |
| 411 | foreach ( $collection as $key => $m ) { |
| 412 | $transTemplate = $template['EXTRA']['TEMPLATE'][$key] ?? []; |
| 413 | $potTemplate = $pot['EXTRA']['TEMPLATE'][$key] ?? []; |
| 414 | $documentation = isset( $documentationCollection[$key] ) ? |
| 415 | $documentationCollection[$key]->translation() : null; |
| 416 | |
| 417 | $output .= $this->formatMessageBlock( |
| 418 | $key, |
| 419 | $m, |
| 420 | $transTemplate, |
| 421 | $potTemplate, |
| 422 | $pluralCount, |
| 423 | $documentation |
| 424 | ); |
| 425 | } |
| 426 | |
| 427 | return $output; |
| 428 | } |
| 429 | |
| 430 | private function doGettextHeader( MessageCollection $collection, array $template ): string { |
| 431 | global $wgSitename; |
| 432 | |
| 433 | $code = $collection->code; |
| 434 | $name = Utilities::getLanguageName( $code ); |
| 435 | $native = Utilities::getLanguageName( $code, $code ); |
| 436 | $authors = $this->doAuthors( $collection ); |
| 437 | if ( isset( $this->extra['header'] ) ) { |
| 438 | $extra = "# --\n" . $this->extra['header']; |
| 439 | } else { |
| 440 | $extra = ''; |
| 441 | } |
| 442 | |
| 443 | $group = $this->getGroup(); |
| 444 | $output = |
| 445 | <<<EOT |
| 446 | # Translation of {$group->getLabel()} to $name ($native) |
| 447 | # Exported from $wgSitename |
| 448 | # |
| 449 | $authors$extra |
| 450 | EOT; |
| 451 | |
| 452 | // Make sure there is no empty line before msgid |
| 453 | $output = trim( $output ) . "\n"; |
| 454 | |
| 455 | $specs = $template['HEADERS'] ?? []; |
| 456 | |
| 457 | $timestamp = wfTimestampNow(); |
| 458 | $specs['PO-Revision-Date'] = $this->formatTime( $timestamp ); |
| 459 | if ( $this->offlineMode ) { |
| 460 | $specs['POT-Creation-Date'] = $this->formatTime( $timestamp ); |
| 461 | } else { |
| 462 | $specs['X-POT-Import-Date'] = $this->formatTime( wfTimestamp( TS_MW, $this->getPotTime() ) ); |
| 463 | } |
| 464 | $specs['Content-Type'] = 'text/plain; charset=UTF-8'; |
| 465 | $specs['Content-Transfer-Encoding'] = '8bit'; |
| 466 | |
| 467 | $specs['Language'] = LanguageCode::bcp47( $this->group->mapCode( $code ) ); |
| 468 | |
| 469 | Services::getInstance()->getHookRunner()->onTranslate_GettextFormat_headerFields( |
| 470 | $specs, |
| 471 | $this->group, |
| 472 | $code |
| 473 | ); |
| 474 | |
| 475 | $specs['X-Generator'] = 'MediaWiki ' |
| 476 | . SpecialVersion::getVersion() |
| 477 | . '; Translate ' |
| 478 | . Utilities::getVersion(); |
| 479 | |
| 480 | if ( $this->offlineMode ) { |
| 481 | $specs['X-Language-Code'] = $code; |
| 482 | $specs['X-Message-Group'] = $group->getId(); |
| 483 | } |
| 484 | |
| 485 | $specs['Plural-Forms'] = GettextPlural::getPluralRule( $code ) |
| 486 | ?: GettextPlural::getPluralRule( 'en' ); |
| 487 | |
| 488 | $output .= 'msgid ""' . "\n"; |
| 489 | $output .= 'msgstr ""' . "\n"; |
| 490 | $output .= '""' . "\n"; |
| 491 | |
| 492 | foreach ( $specs as $k => $v ) { |
| 493 | $output .= $this->escape( "$k: $v\n" ) . "\n"; |
| 494 | } |
| 495 | |
| 496 | $output .= "\n"; |
| 497 | |
| 498 | return $output; |
| 499 | } |
| 500 | |
| 501 | private function doAuthors( MessageCollection $collection ): string { |
| 502 | $output = ''; |
| 503 | $authors = $collection->getAuthors(); |
| 504 | $authors = $this->filterAuthors( $authors, $collection->code ); |
| 505 | |
| 506 | foreach ( $authors as $author ) { |
| 507 | $output .= "# Author: $author\n"; |
| 508 | } |
| 509 | |
| 510 | return $output; |
| 511 | } |
| 512 | |
| 513 | private function formatMessageBlock( |
| 514 | string $key, |
| 515 | Message $message, |
| 516 | array $trans, |
| 517 | array $pot, |
| 518 | int $pluralCount, |
| 519 | ?string $documentation |
| 520 | ): string { |
| 521 | $header = $this->formatDocumentation( $documentation ); |
| 522 | $content = ''; |
| 523 | |
| 524 | $comments = $pot['comments'] ?? $trans['comments'] ?? []; |
| 525 | foreach ( $comments as $type => $typecomments ) { |
| 526 | foreach ( $typecomments as $comment ) { |
| 527 | $header .= "#$type $comment\n"; |
| 528 | } |
| 529 | } |
| 530 | |
| 531 | $flags = $pot['flags'] ?? $trans['flags'] ?? []; |
| 532 | $flags = array_merge( $message->getTags(), $flags ); |
| 533 | |
| 534 | if ( $this->offlineMode ) { |
| 535 | $content .= 'msgctxt ' . $this->escape( $key ) . "\n"; |
| 536 | } else { |
| 537 | $ctxt = $pot['ctxt'] ?? $trans['ctxt'] ?? false; |
| 538 | if ( $ctxt !== false ) { |
| 539 | $content .= 'msgctxt ' . $this->escape( $ctxt ) . "\n"; |
| 540 | } |
| 541 | } |
| 542 | |
| 543 | $msgid = $message->definition(); |
| 544 | $msgstr = $message->translation() ?? ''; |
| 545 | if ( strpos( $msgstr, TRANSLATE_FUZZY ) !== false ) { |
| 546 | $msgstr = str_replace( TRANSLATE_FUZZY, '', $msgstr ); |
| 547 | // Might be fuzzy infile |
| 548 | $flags[] = 'fuzzy'; |
| 549 | } |
| 550 | |
| 551 | if ( GettextPlural::hasPlural( $msgid ) ) { |
| 552 | $forms = GettextPlural::unflatten( $msgid, 2 ); |
| 553 | $content .= 'msgid ' . $this->escape( $forms[0] ) . "\n"; |
| 554 | $content .= 'msgid_plural ' . $this->escape( $forms[1] ) . "\n"; |
| 555 | |
| 556 | try { |
| 557 | $forms = GettextPlural::unflatten( $msgstr, $pluralCount ); |
| 558 | foreach ( $forms as $index => $form ) { |
| 559 | $content .= "msgstr[$index] " . $this->escape( $form ) . "\n"; |
| 560 | } |
| 561 | } catch ( GettextPluralException $e ) { |
| 562 | $flags[] = 'invalid-plural'; |
| 563 | for ( $i = 0; $i < $pluralCount; $i++ ) { |
| 564 | $content .= "msgstr[$i] \"\"\n"; |
| 565 | } |
| 566 | } |
| 567 | } else { |
| 568 | $content .= 'msgid ' . $this->escape( $msgid ) . "\n"; |
| 569 | $content .= 'msgstr ' . $this->escape( $msgstr ) . "\n"; |
| 570 | } |
| 571 | |
| 572 | if ( $flags ) { |
| 573 | sort( $flags ); |
| 574 | $header .= '#, ' . implode( ', ', array_unique( $flags ) ) . "\n"; |
| 575 | } |
| 576 | |
| 577 | $output = $header ?: "#\n"; |
| 578 | $output .= $content . "\n"; |
| 579 | |
| 580 | return $output; |
| 581 | } |
| 582 | |
| 583 | private function formatTime( string $time ): string { |
| 584 | $lang = MediaWikiServices::getInstance()->getLanguageFactory()->getLanguage( 'en' ); |
| 585 | |
| 586 | return $lang->sprintfDate( 'xnY-xnm-xnd xnH:xni:xns+0000', $time ); |
| 587 | } |
| 588 | |
| 589 | private function getPotTime(): string { |
| 590 | $cache = $this->group->getMessageGroupCache( $this->group->getSourceLanguage() ); |
| 591 | |
| 592 | return $cache->exists() ? $cache->getTimestamp() : wfTimestampNow(); |
| 593 | } |
| 594 | |
| 595 | private function formatDocumentation( ?string $documentation ): string { |
| 596 | if ( !is_string( $documentation ) ) { |
| 597 | return ''; |
| 598 | } |
| 599 | |
| 600 | if ( !$this->offlineMode ) { |
| 601 | return ''; |
| 602 | } |
| 603 | |
| 604 | $lines = explode( "\n", $documentation ); |
| 605 | $out = ''; |
| 606 | foreach ( $lines as $line ) { |
| 607 | $out .= "#. [Wiki] $line\n"; |
| 608 | } |
| 609 | |
| 610 | return $out; |
| 611 | } |
| 612 | |
| 613 | private function escape( string $line ): string { |
| 614 | // There may be \ as a last character, for keeping trailing whitespace |
| 615 | $line = preg_replace( '/(\s)\\\\$/', '\1', $line ); |
| 616 | $line = addcslashes( $line, '\\"' ); |
| 617 | $line = str_replace( "\n", '\n', $line ); |
| 618 | return '"' . $line . '"'; |
| 619 | } |
| 620 | |
| 621 | public function shouldOverwrite( string $a, string $b ): bool { |
| 622 | $regex = '/^"(.+)-Date: \d\d\d\d-\d\d-\d\d \d\d:\d\d:\d\d\+\d\d\d\d\\\\n"$/m'; |
| 623 | |
| 624 | $a = preg_replace( $regex, '', $a ); |
| 625 | $b = preg_replace( $regex, '', $b ); |
| 626 | |
| 627 | return $a !== $b; |
| 628 | } |
| 629 | |
| 630 | public static function getExtraSchema(): array { |
| 631 | return [ |
| 632 | 'root' => [ |
| 633 | '_type' => 'array', |
| 634 | '_children' => [ |
| 635 | 'FILES' => [ |
| 636 | '_type' => 'array', |
| 637 | '_children' => [ |
| 638 | 'header' => [ |
| 639 | '_type' => 'text', |
| 640 | ], |
| 641 | 'keyAlgorithm' => [ |
| 642 | '_type' => 'enum', |
| 643 | '_values' => [ 'simple', 'legacy' ], |
| 644 | ], |
| 645 | 'CtxtAsKey' => [ |
| 646 | '_type' => 'boolean', |
| 647 | ], |
| 648 | ] |
| 649 | ] |
| 650 | ] |
| 651 | ] |
| 652 | ]; |
| 653 | } |
| 654 | |
| 655 | public function isContentEqual( ?string $a, ?string $b ): bool { |
| 656 | if ( $a === $b ) { |
| 657 | return true; |
| 658 | } |
| 659 | |
| 660 | if ( $a === null || $b === null ) { |
| 661 | return false; |
| 662 | } |
| 663 | |
| 664 | try { |
| 665 | $parsedA = GettextPlural::parsePluralForms( $a ); |
| 666 | $parsedB = GettextPlural::parsePluralForms( $b ); |
| 667 | |
| 668 | // if they have the different number of plural forms, just fail |
| 669 | if ( count( $parsedA[1] ) !== count( $parsedB[1] ) ) { |
| 670 | return false; |
| 671 | } |
| 672 | |
| 673 | } catch ( GettextPluralException $e ) { |
| 674 | // Something failed, invalid syntax? |
| 675 | return false; |
| 676 | } |
| 677 | |
| 678 | $expectedPluralCount = count( $parsedA[1] ); |
| 679 | |
| 680 | // GettextPlural::unflatten() will return an empty array when $expectedPluralCount is 0 |
| 681 | // So if they do not have translations and are different strings, they are not equal |
| 682 | if ( $expectedPluralCount === 0 ) { |
| 683 | return false; |
| 684 | } |
| 685 | |
| 686 | return GettextPlural::unflatten( $a, $expectedPluralCount ) |
| 687 | === GettextPlural::unflatten( $b, $expectedPluralCount ); |
| 688 | } |
| 689 | } |
| 690 | |
| 691 | class_alias( GettextFormat::class, 'GettextFFS' ); |