matecat /
xliff-parser
| Total Complexity | 51 |
| Total Lines | 482 |
| Duplicated Lines | 0 % |
| Changes | 9 | ||
| Bugs | 0 | Features | 0 |
Complex classes like AbstractXliffReplacer often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use AbstractXliffReplacer, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 8 | abstract class AbstractXliffReplacer { |
||
| 9 | protected $originalFP; |
||
| 10 | protected $outputFP; // output stream pointer |
||
| 11 | |||
| 12 | protected string $tuTagName; // <trans-unit> (forXliff v 1.*) or <unit> (forXliff v 2.*) |
||
| 13 | protected bool $inTU = false; // flag to check whether we are in a <trans-unit> |
||
| 14 | protected bool $inTarget = false; // flag to check whether we are in a <target>, to ignore everything |
||
| 15 | protected bool $isEmpty = false; // flag to check whether we are in an empty tag (<tag/>) |
||
| 16 | protected bool $targetWasWritten = false; // flag to check is <target> was written in the current unit |
||
| 17 | protected string $CDATABuffer = ""; // buffer for special tag |
||
| 18 | protected bool $bufferIsActive = false; // buffer for special tag |
||
| 19 | |||
| 20 | protected int $offset = 0; // offset for SAX pointer |
||
| 21 | |||
| 22 | protected string $currentBuffer; // the current piece of text it's been parsed |
||
| 23 | protected int $len; // length of the currentBuffer |
||
| 24 | protected array $segments; // array of translations |
||
| 25 | protected array $lastTransUnit = []; |
||
| 26 | protected int $segmentInUnitPosition = 0; |
||
| 27 | protected ?string $currentTransUnitId = null; // id of current <trans-unit> |
||
| 28 | protected ?string $currentTransUnitIsTranslatable = null; // 'translate' attribute of current <trans-unit> |
||
| 29 | protected bool $hasWrittenCounts = false; // check if <unit> already wrote segment counts (forXliff v 2.*) |
||
| 30 | protected string $targetLang; |
||
| 31 | protected bool $sourceInTarget = false; |
||
| 32 | |||
| 33 | protected array $nodesToBuffer; |
||
| 34 | |||
| 35 | protected array $transUnits; |
||
| 36 | |||
| 37 | /** @var int */ |
||
| 38 | protected int $xliffVersion; |
||
| 39 | |||
| 40 | /** |
||
| 41 | * @var XliffReplacerCallbackInterface|null |
||
| 42 | */ |
||
| 43 | protected ?XliffReplacerCallbackInterface $callback; |
||
| 44 | |||
| 45 | protected ?LoggerInterface $logger; |
||
| 46 | |||
| 47 | protected static $INTERNAL_TAG_PLACEHOLDER; |
||
| 48 | |||
| 49 | protected $counts = [ |
||
| 50 | 'raw_word_count' => 0, |
||
| 51 | 'eq_word_count' => 0, |
||
| 52 | ]; |
||
| 53 | |||
| 54 | /** |
||
| 55 | * AbstractXliffReplacer constructor. |
||
| 56 | * |
||
| 57 | * @param string $originalXliffPath |
||
| 58 | * @param int $xliffVersion |
||
| 59 | * @param array $segments |
||
| 60 | * @param array $transUnits |
||
| 61 | * @param string $trgLang |
||
| 62 | * @param string $outputFilePath |
||
| 63 | * @param bool $setSourceInTarget |
||
| 64 | * @param LoggerInterface|null $logger |
||
| 65 | * @param XliffReplacerCallbackInterface|null $callback |
||
| 66 | */ |
||
| 67 | public function __construct( |
||
| 89 | } |
||
| 90 | |||
| 91 | public function replaceTranslation() { |
||
| 92 | fwrite( $this->outputFP, '<?xml version="1.0" encoding="UTF-8"?>' ); |
||
| 93 | |||
| 94 | //create Sax parser |
||
| 95 | $xmlParser = $this->initSaxParser(); |
||
| 96 | |||
| 97 | while ( $this->currentBuffer = fread( $this->originalFP, 4096 ) ) { |
||
| 98 | /* |
||
| 99 | preprocess file |
||
| 100 | */ |
||
| 101 | // obfuscate entities because sax automatically does html_entity_decode |
||
| 102 | $temporary_check_buffer = preg_replace( "/&(.*?);/", self::$INTERNAL_TAG_PLACEHOLDER . '$1' . self::$INTERNAL_TAG_PLACEHOLDER, $this->currentBuffer ); |
||
| 103 | |||
| 104 | //avoid cutting entities in half: |
||
| 105 | //the last fread could have truncated an entity (say, '<' in '&l'), thus invalidating the escaping |
||
| 106 | //***** and if there is an & that it is not an entity, this is an infinite loop !!!!! |
||
| 107 | // 9 is the max length of an entity. So, suppose that the & is at the end of buffer, |
||
| 108 | // add 9 Bytes and substitute the entities, if the & is present, and it is not at the end |
||
| 109 | //it can't be an entity, exit the loop |
||
| 110 | while ( true ) { |
||
| 111 | $_ampPos = strpos( $temporary_check_buffer, '&' ); |
||
| 112 | |||
| 113 | //check for real entity or escape it to safely exit from the loop!!! |
||
| 114 | if ( $_ampPos === false || strlen( substr( $temporary_check_buffer, $_ampPos ) ) > 9 ) { |
||
| 115 | break; |
||
| 116 | } |
||
| 117 | |||
| 118 | //if an entity is still present, fetch some more and repeat the escaping |
||
| 119 | $this->currentBuffer .= fread( $this->originalFP, 9 ); |
||
| 120 | $temporary_check_buffer = preg_replace( "/&(.*?);/", self::$INTERNAL_TAG_PLACEHOLDER . '$1' . self::$INTERNAL_TAG_PLACEHOLDER, $this->currentBuffer ); |
||
| 121 | } |
||
| 122 | |||
| 123 | //free stuff outside the loop |
||
| 124 | unset( $temporary_check_buffer ); |
||
| 125 | |||
| 126 | $this->currentBuffer = preg_replace( "/&(.*?);/", self::$INTERNAL_TAG_PLACEHOLDER . '$1' . self::$INTERNAL_TAG_PLACEHOLDER, $this->currentBuffer ); |
||
| 127 | $this->currentBuffer = str_replace( "&", self::$INTERNAL_TAG_PLACEHOLDER . 'amp' . self::$INTERNAL_TAG_PLACEHOLDER, $this->currentBuffer ); |
||
| 128 | |||
| 129 | //get length of chunk |
||
| 130 | $this->len = strlen( $this->currentBuffer ); |
||
| 131 | |||
| 132 | /* |
||
| 133 | * Get the accumulated this->offset in the document: |
||
| 134 | * as long as SAX pointer advances, we keep track of total bytes it has seen so far; |
||
| 135 | * this way, we can translate its global pointer in an address local to the current buffer of text to retrieve the last char of tag |
||
| 136 | */ |
||
| 137 | $this->offset += $this->len; |
||
| 138 | |||
| 139 | //parse chunk of text |
||
| 140 | $this->runParser( $xmlParser ); |
||
| 141 | |||
| 142 | } |
||
| 143 | |||
| 144 | // close Sax parser |
||
| 145 | $this->closeSaxParser( $xmlParser ); |
||
| 146 | |||
| 147 | } |
||
| 148 | |||
| 149 | /** |
||
| 150 | * @param $xmlParser |
||
| 151 | * |
||
| 152 | * @return void |
||
| 153 | */ |
||
| 154 | protected function runParser( $xmlParser ) { |
||
| 162 | ) ); |
||
| 163 | } |
||
| 164 | } |
||
| 165 | |||
| 166 | /** |
||
| 167 | * @param resource $parser |
||
| 168 | * |
||
| 169 | * @return string |
||
| 170 | */ |
||
| 171 | protected function getLastCharacter( $parser ): string { |
||
| 172 | |||
| 173 | //this logic helps detecting empty tags |
||
| 174 | //get current position of SAX pointer in all the stream of data is has read so far: |
||
| 175 | //it points at the end of current tag |
||
| 176 | $idx = xml_get_current_byte_index( $parser ); |
||
| 177 | |||
| 178 | //check whether the bounds of current tag are entirely in current buffer or the end of the current tag |
||
| 179 | //is outside current buffer (in the latter case, it's in next buffer to be read by the while loop); |
||
| 180 | //this check is necessary because we may have truncated a tag in half with current read, |
||
| 181 | //and the other half may be encountered in the next buffer it will be passed |
||
| 182 | return $this->currentBuffer[ $idx - $this->offset ] ?? $this->currentBuffer[ $this->len - 1 ]; |
||
| 183 | |||
| 184 | } |
||
| 185 | |||
| 186 | /** |
||
| 187 | * @return string |
||
| 188 | */ |
||
| 189 | private function getInternalTagPlaceholder(): string { |
||
| 190 | return "§" . |
||
| 191 | substr( |
||
| 192 | str_replace( |
||
| 193 | [ '+', '/' ], |
||
| 194 | '', |
||
| 195 | base64_encode( openssl_random_pseudo_bytes( 10, $_crypto_strong ) ) |
||
| 196 | ), |
||
| 197 | 0, |
||
| 198 | 4 |
||
| 199 | ); |
||
| 200 | } |
||
| 201 | |||
| 202 | private function createOutputFileIfDoesNotExist( string $outputFilePath ) { |
||
| 203 | // create output file |
||
| 204 | if ( !file_exists( $outputFilePath ) ) { |
||
| 205 | touch( $outputFilePath ); |
||
| 206 | } |
||
| 207 | } |
||
| 208 | |||
| 209 | /** |
||
| 210 | * @param string $originalXliffPath |
||
| 211 | * @param string $outputFilePath |
||
| 212 | */ |
||
| 213 | private function setFileDescriptors( string $originalXliffPath, string $outputFilePath ) { |
||
| 214 | $this->outputFP = fopen( $outputFilePath, 'w+' ); |
||
| 215 | |||
| 216 | $streamArgs = null; |
||
| 217 | |||
| 218 | if ( !( $this->originalFP = fopen( $originalXliffPath, "r", false, stream_context_create( $streamArgs ) ) ) ) { |
||
| 219 | throw new RuntimeException( "could not open XML input" ); |
||
| 220 | } |
||
| 221 | } |
||
| 222 | |||
| 223 | /** |
||
| 224 | * AbstractXliffReplacer destructor. |
||
| 225 | */ |
||
| 226 | public function __destruct() { |
||
| 227 | //this stream can be closed outside the class |
||
| 228 | //to permit multiple concurrent downloads, so suppress warnings |
||
| 229 | if ( is_resource( $this->originalFP ) ) { |
||
| 230 | fclose( $this->originalFP ); |
||
| 231 | } |
||
| 232 | |||
| 233 | if ( is_resource( $this->outputFP ) ) { |
||
| 234 | fclose( $this->outputFP ); |
||
| 235 | } |
||
| 236 | |||
| 237 | } |
||
| 238 | |||
| 239 | /** |
||
| 240 | * Init Sax parser |
||
| 241 | * |
||
| 242 | * @return resource |
||
| 243 | */ |
||
| 244 | protected function initSaxParser() { |
||
| 245 | $xmlSaxParser = xml_parser_create( 'UTF-8' ); |
||
| 246 | xml_set_object( $xmlSaxParser, $this ); |
||
| 247 | xml_parser_set_option( $xmlSaxParser, XML_OPTION_CASE_FOLDING, false ); |
||
| 248 | xml_set_element_handler( $xmlSaxParser, 'tagOpen', 'tagClose' ); |
||
| 249 | xml_set_character_data_handler( $xmlSaxParser, 'characterData' ); |
||
| 250 | |||
| 251 | return $xmlSaxParser; |
||
|
Bug
Best Practice
introduced
by
|
|||
| 252 | } |
||
| 253 | |||
| 254 | /** |
||
| 255 | * @param resource $xmlSaxParser |
||
| 256 | */ |
||
| 257 | protected function closeSaxParser( $xmlSaxParser ) { |
||
| 258 | xml_parser_free( $xmlSaxParser ); |
||
| 259 | } |
||
| 260 | |||
| 261 | /** |
||
| 262 | * @param resource $parser |
||
| 263 | * @param string $name |
||
| 264 | * @param array $attr |
||
| 265 | * |
||
| 266 | * @return mixed |
||
| 267 | */ |
||
| 268 | abstract protected function tagOpen( $parser, string $name, array $attr ); |
||
| 269 | |||
| 270 | /** |
||
| 271 | * @param resource $parser |
||
| 272 | * @param string $name |
||
| 273 | * |
||
| 274 | * @return mixed |
||
| 275 | */ |
||
| 276 | abstract protected function tagClose( $parser, string $name ); |
||
| 277 | |||
| 278 | /** |
||
| 279 | * @param resource $parser |
||
| 280 | * @param string $data |
||
| 281 | * |
||
| 282 | * @return mixed |
||
| 283 | */ |
||
| 284 | protected function characterData( $parser, string $data ): void { |
||
| 290 | } |
||
| 291 | } |
||
| 292 | |||
| 293 | /** |
||
| 294 | * postprocess escaped data and write to disk |
||
| 295 | * |
||
| 296 | * @param resource $fp |
||
| 297 | * @param string $data |
||
| 298 | * @param bool $treatAsCDATA |
||
| 299 | */ |
||
| 300 | protected function postProcAndFlush( $fp, string $data, bool $treatAsCDATA = false ) { |
||
| 301 | //postprocess string |
||
| 302 | $data = preg_replace( "/" . self::$INTERNAL_TAG_PLACEHOLDER . '(.*?)' . self::$INTERNAL_TAG_PLACEHOLDER . "/", '&$1;', $data ); |
||
| 303 | $data = str_replace( ' ', ' ', $data ); |
||
| 304 | if ( !$treatAsCDATA ) { |
||
| 305 | //unix2dos |
||
| 306 | $data = str_replace( "\r\n", "\r", $data ); |
||
| 307 | $data = str_replace( "\n", "\r", $data ); |
||
| 308 | $data = str_replace( "\r", "\r\n", $data ); |
||
| 309 | } |
||
| 310 | |||
| 311 | //flush to disk |
||
| 312 | fwrite( $fp, $data ); |
||
| 313 | } |
||
| 314 | |||
| 315 | /** |
||
| 316 | * @param string $name |
||
| 317 | * @param array $attr |
||
| 318 | * |
||
| 319 | * @return void |
||
| 320 | */ |
||
| 321 | protected function handleOpenUnit( string $name, array $attr ) { |
||
| 322 | |||
| 323 | // check if we are entering into a <trans-unit> (xliff v1.*) or <unit> (xliff v2.*) |
||
| 324 | if ( $this->tuTagName === $name ) { |
||
| 325 | $this->inTU = true; |
||
| 326 | |||
| 327 | // get id |
||
| 328 | // trim to first 100 characters because this is the limit on Matecat's DB |
||
| 329 | $this->currentTransUnitId = substr( $attr[ 'id' ], 0, 100 ); |
||
| 330 | |||
| 331 | // `translate` attribute can be only yes or no |
||
| 332 | // current 'translate' attribute of the current trans-unit |
||
| 333 | $this->currentTransUnitIsTranslatable = empty( $attr[ 'translate' ] ) ? 'yes' : $attr[ 'translate' ]; |
||
| 334 | |||
| 335 | $this->setLastTransUnitSegments(); |
||
| 336 | |||
| 337 | } |
||
| 338 | } |
||
| 339 | |||
| 340 | /** |
||
| 341 | * @param string $name |
||
| 342 | * @param array $attr |
||
| 343 | * @param string $tag |
||
| 344 | * |
||
| 345 | * @return string |
||
| 346 | */ |
||
| 347 | protected function handleOpenXliffTag( string $name, array $attr, string $tag ): string { |
||
| 348 | |||
| 349 | // Add MateCat specific namespace. |
||
| 350 | // Add trgLang |
||
| 351 | if ( $name === 'xliff' ) { |
||
| 352 | if ( !array_key_exists( 'xmlns:mtc', $attr ) ) { |
||
| 353 | $tag .= ' xmlns:mtc="https://www.matecat.com" '; |
||
| 354 | } |
||
| 355 | $tag = preg_replace( '/trgLang="(.*?)"/', 'trgLang="' . $this->targetLang . '"', $tag ); |
||
| 356 | } |
||
| 357 | |||
| 358 | return $tag; |
||
| 359 | |||
| 360 | } |
||
| 361 | |||
| 362 | /** |
||
| 363 | * @param string $name |
||
| 364 | * |
||
| 365 | * @return void |
||
| 366 | */ |
||
| 367 | protected function checkSetInTarget( string $name ) { |
||
| 368 | // check if we are entering into a <target> |
||
| 369 | if ( 'target' === $name ) { |
||
| 370 | if ( $this->currentTransUnitIsTranslatable === 'no' ) { |
||
| 371 | $this->inTarget = false; |
||
| 372 | } else { |
||
| 373 | $this->inTarget = true; |
||
| 374 | } |
||
| 375 | } |
||
| 376 | } |
||
| 377 | |||
| 378 | /** |
||
| 379 | * @param string $name |
||
| 380 | * |
||
| 381 | * @return void |
||
| 382 | */ |
||
| 383 | protected function setInBuffer( string $name ) { |
||
| 384 | if ( in_array( $name, $this->nodesToBuffer ) ) { |
||
| 385 | $this->bufferIsActive = true; |
||
| 386 | } |
||
| 387 | } |
||
| 388 | |||
| 389 | /** |
||
| 390 | * @param array $seg |
||
| 391 | */ |
||
| 392 | protected function updateSegmentCounts( array $seg = [] ) { |
||
| 404 | } |
||
| 405 | |||
| 406 | protected function resetCounts() { |
||
| 407 | $this->counts[ 'segments_count_array' ] = []; |
||
| 408 | $this->counts[ 'raw_word_count' ] = 0; |
||
| 409 | $this->counts[ 'eq_word_count' ] = 0; |
||
| 410 | } |
||
| 411 | |||
| 412 | /** |
||
| 413 | * @param resource $parser |
||
| 414 | * @param string $tag |
||
| 415 | * |
||
| 416 | * @return void |
||
| 417 | */ |
||
| 418 | protected function checkForSelfClosedTagAndFlush( $parser, string $tag ) { |
||
| 419 | |||
| 420 | $lastChar = $this->getLastCharacter( $parser ); |
||
| 421 | |||
| 422 | //trim last space |
||
| 423 | $tag = rtrim( $tag ); |
||
| 424 | |||
| 425 | //detect empty tag |
||
| 426 | $this->isEmpty = $lastChar == '/'; |
||
| 427 | if ( $this->isEmpty ) { |
||
| 428 | $tag .= $lastChar; |
||
| 429 | } |
||
| 430 | |||
| 431 | //add tag ending |
||
| 432 | $tag .= ">"; |
||
| 433 | |||
| 434 | //set a Buffer for the segSource Source tag |
||
| 435 | if ( $this->bufferIsActive ) { // we are opening a critical CDATA section |
||
| 436 | //these are NOT source/seg-source/value empty tags, THERE IS A CONTENT, write it in buffer |
||
| 437 | $this->CDATABuffer .= $tag; |
||
| 438 | } else { |
||
| 439 | $this->postProcAndFlush( $this->outputFP, $tag ); |
||
| 440 | } |
||
| 441 | |||
| 442 | } |
||
| 443 | |||
| 444 | /** |
||
| 445 | * A trans-unit can contain a list of segments because of mrk tags |
||
| 446 | * Copy the segment's list for this trans-unit in a different structure |
||
| 447 | * |
||
| 448 | * @return void |
||
| 449 | */ |
||
| 450 | protected function setLastTransUnitSegments() { |
||
| 451 | |||
| 452 | /* |
||
| 453 | * At the end of every cycle the segment grouping information is lost: unset( 'matecat|' . $this->currentId ) |
||
| 454 | * |
||
| 455 | * We need to take the info about the last segment parsed |
||
| 456 | * ( normally more than 1 db row because of mrk tags ) |
||
| 457 | * |
||
| 458 | * So, copy the current segment data group into another structure to keep the last segment |
||
| 459 | * for the next tagOpen ( possible sdl:seg-defs ) |
||
| 460 | * |
||
| 461 | */ |
||
| 462 | $this->lastTransUnit = []; |
||
| 463 | |||
| 464 | if ( !isset( $this->transUnits[ $this->currentTransUnitId ] ) ) { |
||
| 465 | return; |
||
| 466 | } |
||
| 467 | |||
| 468 | $listOfSegmentsIds = $this->transUnits[ $this->currentTransUnitId ]; |
||
| 469 | $last_value = null; |
||
| 470 | $segmentsCount = count( $listOfSegmentsIds ); |
||
| 471 | for ( $i = 0; $i < $segmentsCount; $i++ ) { |
||
| 472 | $id = $listOfSegmentsIds[ $i ]; |
||
| 473 | if ( isset( $this->segments[ $id ] ) && ( $i == 0 || $last_value + 1 == $listOfSegmentsIds[ $i ] ) ) { |
||
| 474 | $last_value = $listOfSegmentsIds[ $i ]; |
||
| 475 | $this->lastTransUnit[] = $this->segments[ $id ]; |
||
| 476 | } |
||
| 477 | } |
||
| 478 | |||
| 479 | } |
||
| 480 | |||
| 481 | /** |
||
| 482 | * @return array |
||
| 483 | */ |
||
| 484 | protected function getCurrentSegment(): array { |
||
| 490 | } |
||
| 491 | |||
| 492 | } |
Loading history...
