1<?php 2 3namespace MediaWiki\Extensions\ParserFunctions; 4 5use DateTime; 6use DateTimeZone; 7use Exception; 8use ILanguageConverter; 9use Language; 10use MediaWiki\MediaWikiServices; 11use MWTimestamp; 12use Parser; 13use PPFrame; 14use PPNode; 15use Sanitizer; 16use StringUtils; 17use StubObject; 18use Title; 19 20/** 21 * Parser function handlers 22 * 23 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions 24 */ 25class ParserFunctions { 26 private static $mExprParser; 27 private static $mTimeCache = []; 28 private static $mTimeChars = 0; 29 30 /** ~10 seconds */ 31 private const MAX_TIME_CHARS = 6000; 32 33 /** 34 * Register ParserClearState hook. 35 * We defer this until needed to avoid the loading of the code of this file 36 * when no parser function is actually called. 37 */ 38 private static function registerClearHook() { 39 static $done = false; 40 if ( !$done ) { 41 global $wgHooks; 42 $wgHooks['ParserClearState'][] = function () { 43 self::$mTimeChars = 0; 44 }; 45 $done = true; 46 } 47 } 48 49 /** 50 * @return ExprParser 51 */ 52 private static function &getExprParser() { 53 if ( !isset( self::$mExprParser ) ) { 54 self::$mExprParser = new ExprParser; 55 } 56 return self::$mExprParser; 57 } 58 59 /** 60 * {{#expr: expression }} 61 * 62 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##expr 63 * 64 * @param Parser $parser 65 * @param string $expr 66 * @return string 67 */ 68 public static function expr( Parser $parser, $expr = '' ) { 69 try { 70 return self::getExprParser()->doExpression( $expr ); 71 } catch ( ExprError $e ) { 72 return '<strong class="error">' . htmlspecialchars( $e->getUserFriendlyMessage() ) . '</strong>'; 73 } 74 } 75 76 /** 77 * {{#ifexpr: expression | value if true | value if false }} 78 * 79 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##ifexpr 80 * 81 * @param Parser $parser 82 * @param PPFrame $frame 83 * @param array $args 84 * @return string 85 */ 86 public static function ifexpr( Parser $parser, PPFrame $frame, array $args ) { 87 $expr = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 88 $then = $args[1] ?? ''; 89 $else = $args[2] ?? ''; 90 91 try { 92 $result = self::getExprParser()->doExpression( $expr ); 93 if ( is_numeric( $result ) ) { 94 $result = (float)$result; 95 } 96 $result = $result ? $then : $else; 97 } catch ( ExprError $e ) { 98 return '<strong class="error">' . htmlspecialchars( $e->getUserFriendlyMessage() ) . '</strong>'; 99 } 100 101 if ( is_object( $result ) ) { 102 $result = trim( $frame->expand( $result ) ); 103 } 104 105 return $result; 106 } 107 108 /** 109 * {{#if: test string | value if test string is not empty | value if test string is empty }} 110 * 111 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##if 112 * 113 * @param Parser $parser 114 * @param PPFrame $frame 115 * @param array $args 116 * @return string 117 */ 118 public static function if( Parser $parser, PPFrame $frame, array $args ) { 119 $test = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 120 if ( $test !== '' ) { 121 return isset( $args[1] ) ? trim( $frame->expand( $args[1] ) ) : ''; 122 } else { 123 return isset( $args[2] ) ? trim( $frame->expand( $args[2] ) ) : ''; 124 } 125 } 126 127 /** 128 * {{#ifeq: string 1 | string 2 | value if identical | value if different }} 129 * 130 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##ifeq 131 * 132 * @param Parser $parser 133 * @param PPFrame $frame 134 * @param array $args 135 * @return string 136 */ 137 public static function ifeq( Parser $parser, PPFrame $frame, array $args ) { 138 $left = isset( $args[0] ) ? self::decodeTrimExpand( $args[0], $frame ) : ''; 139 $right = isset( $args[1] ) ? self::decodeTrimExpand( $args[1], $frame ) : ''; 140 141 // Strict compare is not possible here. 01 should equal 1 for example. 142 /** @noinspection TypeUnsafeComparisonInspection */ 143 if ( $left == $right ) { 144 return isset( $args[2] ) ? trim( $frame->expand( $args[2] ) ) : ''; 145 } else { 146 return isset( $args[3] ) ? trim( $frame->expand( $args[3] ) ) : ''; 147 } 148 } 149 150 /** 151 * {{#iferror: test string | value if error | value if no error }} 152 * 153 * Error is when the input string contains an HTML object with class="error", as 154 * generated by other parser functions such as #expr, #time and #rel2abs. 155 * 156 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##iferror 157 * 158 * @param Parser $parser 159 * @param PPFrame $frame 160 * @param array $args 161 * @return string 162 */ 163 public static function iferror( Parser $parser, PPFrame $frame, array $args ) { 164 $test = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 165 $then = $args[1] ?? false; 166 $else = $args[2] ?? false; 167 168 if ( preg_match( 169 '/<(?:strong|span|p|div)\s(?:[^\s>]*\s+)*?class="(?:[^"\s>]*\s+)*?error(?:\s[^">]*)?"/', 170 $test ) 171 ) { 172 $result = $then; 173 } elseif ( $else === false ) { 174 $result = $test; 175 } else { 176 $result = $else; 177 } 178 if ( $result === false ) { 179 return ''; 180 } 181 182 return trim( $frame->expand( $result ) ); 183 } 184 185 /** 186 * {{#switch: comparison string 187 * | case = result 188 * | case = result 189 * | ... 190 * | default result 191 * }} 192 * 193 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##switch 194 * 195 * @param Parser $parser 196 * @param PPFrame $frame 197 * @param array $args 198 * @return string 199 */ 200 public static function switch( Parser $parser, PPFrame $frame, array $args ) { 201 if ( count( $args ) === 0 ) { 202 return ''; 203 } 204 $primary = self::decodeTrimExpand( array_shift( $args ), $frame ); 205 $found = $defaultFound = false; 206 $default = null; 207 $lastItemHadNoEquals = false; 208 $lastItem = ''; 209 $mwDefault = $parser->getMagicWordFactory()->get( 'default' ); 210 foreach ( $args as $arg ) { 211 $bits = $arg->splitArg(); 212 $nameNode = $bits['name']; 213 $index = $bits['index']; 214 $valueNode = $bits['value']; 215 216 if ( $index === '' ) { 217 # Found "=" 218 $lastItemHadNoEquals = false; 219 if ( $found ) { 220 # Multiple input match 221 return trim( $frame->expand( $valueNode ) ); 222 } else { 223 $test = self::decodeTrimExpand( $nameNode, $frame ); 224 /** @noinspection TypeUnsafeComparisonInspection */ 225 if ( $test == $primary ) { 226 # Found a match, return now 227 return trim( $frame->expand( $valueNode ) ); 228 } elseif ( $defaultFound || $mwDefault->matchStartToEnd( $test ) ) { 229 $default = $valueNode; 230 $defaultFound = false; 231 } # else wrong case, continue 232 } 233 } else { 234 # Multiple input, single output 235 # If the value matches, set a flag and continue 236 $lastItemHadNoEquals = true; 237 // $lastItem is an "out" variable 238 $decodedTest = self::decodeTrimExpand( $valueNode, $frame, $lastItem ); 239 /** @noinspection TypeUnsafeComparisonInspection */ 240 if ( $decodedTest == $primary ) { 241 $found = true; 242 } elseif ( $mwDefault->matchStartToEnd( $decodedTest ) ) { 243 $defaultFound = true; 244 } 245 } 246 } 247 # Default case 248 # Check if the last item had no = sign, thus specifying the default case 249 if ( $lastItemHadNoEquals ) { 250 return $lastItem; 251 } elseif ( $default !== null ) { 252 return trim( $frame->expand( $default ) ); 253 } else { 254 return ''; 255 } 256 } 257 258 /** 259 * {{#rel2abs: path }} or {{#rel2abs: path | base path }} 260 * 261 * Returns the absolute path to a subpage, relative to the current article 262 * title. Treats titles as slash-separated paths. 263 * 264 * Following subpage link syntax instead of standard path syntax, an 265 * initial slash is treated as a relative path, and vice versa. 266 * 267 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##rel2abs 268 * 269 * @param Parser $parser 270 * @param string $to 271 * @param string $from 272 * 273 * @return string 274 */ 275 public static function rel2abs( Parser $parser, $to = '', $from = '' ) { 276 $from = trim( $from ); 277 if ( $from === '' ) { 278 $from = $parser->getTitle()->getPrefixedText(); 279 } 280 281 $to = rtrim( $to, ' /' ); 282 283 // if we have an empty path, or just one containing a dot 284 if ( $to === '' || $to === '.' ) { 285 return $from; 286 } 287 288 // if the path isn't relative 289 if ( substr( $to, 0, 1 ) !== '/' && 290 substr( $to, 0, 2 ) !== './' && 291 substr( $to, 0, 3 ) !== '../' && 292 $to !== '..' 293 ) { 294 $from = ''; 295 } 296 // Make a long path, containing both, enclose it in /.../ 297 $fullPath = '/' . $from . '/' . $to . '/'; 298 299 // remove redundant current path dots 300 $fullPath = preg_replace( '!/(\./)+!', '/', $fullPath ); 301 302 // remove double slashes 303 $fullPath = preg_replace( '!/{2,}!', '/', $fullPath ); 304 305 // remove the enclosing slashes now 306 $fullPath = trim( $fullPath, '/' ); 307 $exploded = explode( '/', $fullPath ); 308 $newExploded = []; 309 310 foreach ( $exploded as $current ) { 311 if ( $current === '..' ) { // removing one level 312 if ( !count( $newExploded ) ) { 313 // attempted to access a node above root node 314 $msg = wfMessage( 'pfunc_rel2abs_invalid_depth', $fullPath ) 315 ->inContentLanguage()->escaped(); 316 return '<strong class="error">' . $msg . '</strong>'; 317 } 318 // remove last level from the stack 319 array_pop( $newExploded ); 320 } else { 321 // add the current level to the stack 322 $newExploded[] = $current; 323 } 324 } 325 326 // we can now join it again 327 return implode( '/', $newExploded ); 328 } 329 330 /** 331 * @param Parser $parser 332 * @param PPFrame $frame 333 * @param string $titletext 334 * @param string $then 335 * @param string $else 336 * 337 * @return string 338 */ 339 private static function ifexistInternal( 340 Parser $parser, PPFrame $frame, $titletext = '', $then = '', $else = '' 341 ) { 342 $title = Title::newFromText( $titletext ); 343 self::getLanguageConverter( $parser->getContentLanguage() ) 344 ->findVariantLink( $titletext, $title, true ); 345 if ( $title ) { 346 if ( $title->getNamespace() === NS_MEDIA ) { 347 /* If namespace is specified as NS_MEDIA, then we want to 348 * check the physical file, not the "description" page. 349 */ 350 if ( !$parser->incrementExpensiveFunctionCount() ) { 351 return $else; 352 } 353 $file = MediaWikiServices::getInstance()->getRepoGroup()->findFile( $title ); 354 if ( !$file ) { 355 return $else; 356 } 357 $parser->getOutput()->addImage( 358 $file->getName(), $file->getTimestamp(), $file->getSha1() ); 359 return $file->exists() ? $then : $else; 360 } elseif ( $title->isSpecialPage() ) { 361 /* Don't bother with the count for special pages, 362 * since their existence can be checked without 363 * accessing the database. 364 */ 365 return MediaWikiServices::getInstance()->getSpecialPageFactory() 366 ->exists( $title->getDBkey() ) ? $then : $else; 367 } elseif ( $title->isExternal() ) { 368 /* Can't check the existence of pages on other sites, 369 * so just return $else. Makes a sort of sense, since 370 * they don't exist _locally_. 371 */ 372 return $else; 373 } else { 374 $pdbk = $title->getPrefixedDBkey(); 375 $lc = MediaWikiServices::getInstance()->getLinkCache(); 376 $id = $lc->getGoodLinkID( $pdbk ); 377 if ( $id !== 0 ) { 378 $parser->getOutput()->addLink( $title, $id ); 379 return $then; 380 } elseif ( $lc->isBadLink( $pdbk ) ) { 381 $parser->getOutput()->addLink( $title, 0 ); 382 return $else; 383 } 384 if ( !$parser->incrementExpensiveFunctionCount() ) { 385 return $else; 386 } 387 $id = $title->getArticleID(); 388 $parser->getOutput()->addLink( $title, $id ); 389 390 // bug 70495: don't just check whether the ID != 0 391 if ( $title->exists() ) { 392 return $then; 393 } 394 } 395 } 396 return $else; 397 } 398 399 /** 400 * {{#ifexist: page title | value if exists | value if doesn't exist }} 401 * 402 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##ifexist 403 * 404 * @param Parser $parser 405 * @param PPFrame $frame 406 * @param array $args 407 * @return string 408 */ 409 public static function ifexist( Parser $parser, PPFrame $frame, array $args ) { 410 $title = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 411 $then = $args[1] ?? null; 412 $else = $args[2] ?? null; 413 414 $result = self::ifexistInternal( $parser, $frame, $title, $then, $else ); 415 if ( $result === null ) { 416 return ''; 417 } else { 418 return trim( $frame->expand( $result ) ); 419 } 420 } 421 422 /** 423 * Used by time() and localTime() 424 * 425 * @param Parser $parser 426 * @param PPFrame|null $frame 427 * @param string $format 428 * @param string $date 429 * @param string $language 430 * @param string|bool $local 431 * @return string 432 */ 433 private static function timeCommon( 434 Parser $parser, PPFrame $frame = null, $format = '', $date = '', $language = '', $local = false 435 ) { 436 global $wgLocaltimezone; 437 self::registerClearHook(); 438 if ( $date === '' ) { 439 $cacheKey = $parser->getOptions()->getTimestamp(); 440 $timestamp = new MWTimestamp( $cacheKey ); 441 $date = $timestamp->getTimestamp( TS_ISO_8601 ); 442 $useTTL = true; 443 } else { 444 $cacheKey = $date; 445 $useTTL = false; 446 } 447 if ( isset( self::$mTimeCache[$format][$cacheKey][$language][$local] ) ) { 448 $cachedVal = self::$mTimeCache[$format][$cacheKey][$language][$local]; 449 if ( $useTTL && $cachedVal[1] !== null && $frame ) { 450 $frame->setTTL( $cachedVal[1] ); 451 } 452 return $cachedVal[0]; 453 } 454 455 # compute the timestamp string $ts 456 # PHP >= 5.2 can handle dates before 1970 or after 2038 using the DateTime object 457 458 $invalidTime = false; 459 460 # the DateTime constructor must be used because it throws exceptions 461 # when errors occur, whereas date_create appears to just output a warning 462 # that can't really be detected from within the code 463 try { 464 465 # Default input timezone is UTC. 466 $utc = new DateTimeZone( 'UTC' ); 467 468 # Correct for DateTime interpreting 'XXXX' as XX:XX o'clock 469 if ( preg_match( '/^[0-9]{4}$/', $date ) ) { 470 $date = '00:00 ' . $date; 471 } 472 473 # Parse date 474 # UTC is a default input timezone. 475 $dateObject = new DateTime( $date, $utc ); 476 477 # Set output timezone. 478 if ( $local ) { 479 if ( isset( $wgLocaltimezone ) ) { 480 $tz = new DateTimeZone( $wgLocaltimezone ); 481 } else { 482 $tz = new DateTimeZone( date_default_timezone_get() ); 483 } 484 } else { 485 $tz = $utc; 486 } 487 $dateObject->setTimezone( $tz ); 488 # Generate timestamp 489 $ts = $dateObject->format( 'YmdHis' ); 490 491 } catch ( Exception $ex ) { 492 $invalidTime = true; 493 } 494 495 $ttl = null; 496 # format the timestamp and return the result 497 if ( $invalidTime ) { 498 $result = '<strong class="error">' . 499 wfMessage( 'pfunc_time_error' )->inContentLanguage()->escaped() . 500 '</strong>'; 501 } else { 502 self::$mTimeChars += strlen( $format ); 503 if ( self::$mTimeChars > self::MAX_TIME_CHARS ) { 504 return '<strong class="error">' . 505 wfMessage( 'pfunc_time_too_long' )->inContentLanguage()->escaped() . 506 '</strong>'; 507 } else { 508 if ( $ts < 0 ) { // Language can't deal with BC years 509 return '<strong class="error">' . 510 wfMessage( 'pfunc_time_too_small' )->inContentLanguage()->escaped() . 511 '</strong>'; 512 } elseif ( $ts < 100000000000000 ) { // Language can't deal with years after 9999 513 if ( $language !== '' && Language::isValidBuiltInCode( $language ) ) { 514 // use whatever language is passed as a parameter 515 $langObject = Language::factory( $language ); 516 } else { 517 // use wiki's content language 518 $langObject = $parser->getFunctionLang(); 519 // $ttl is passed by reference, which doesn't work right on stub objects 520 StubObject::unstub( $langObject ); 521 } 522 $result = $langObject->sprintfDate( $format, $ts, $tz, $ttl ); 523 } else { 524 return '<strong class="error">' . 525 wfMessage( 'pfunc_time_too_big' )->inContentLanguage()->escaped() . 526 '</strong>'; 527 } 528 } 529 } 530 self::$mTimeCache[$format][$cacheKey][$language][$local] = [ $result, $ttl ]; 531 if ( $useTTL && $ttl !== null && $frame ) { 532 $frame->setTTL( $ttl ); 533 } 534 return $result; 535 } 536 537 /** 538 * {{#time: format string }} 539 * {{#time: format string | date/time object }} 540 * {{#time: format string | date/time object | language code }} 541 * 542 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##time 543 * 544 * @param Parser $parser 545 * @param PPFrame $frame 546 * @param array $args 547 * @return string 548 */ 549 public static function time( Parser $parser, PPFrame $frame, array $args ) { 550 $format = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 551 $date = isset( $args[1] ) ? trim( $frame->expand( $args[1] ) ) : ''; 552 $language = isset( $args[2] ) ? trim( $frame->expand( $args[2] ) ) : ''; 553 $local = isset( $args[3] ) && trim( $frame->expand( $args[3] ) ); 554 return self::timeCommon( $parser, $frame, $format, $date, $language, $local ); 555 } 556 557 /** 558 * {{#timel: ... }} 559 * 560 * Identical to {{#time: ... }}, except that it uses the local time of the wiki 561 * (as set in $wgLocaltimezone) when no date is given. 562 * 563 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##timel 564 * 565 * @param Parser $parser 566 * @param PPFrame $frame 567 * @param array $args 568 * @return string 569 */ 570 public static function localTime( Parser $parser, PPFrame $frame, array $args ) { 571 $format = isset( $args[0] ) ? trim( $frame->expand( $args[0] ) ) : ''; 572 $date = isset( $args[1] ) ? trim( $frame->expand( $args[1] ) ) : ''; 573 $language = isset( $args[2] ) ? trim( $frame->expand( $args[2] ) ) : ''; 574 return self::timeCommon( $parser, $frame, $format, $date, $language, true ); 575 } 576 577 /** 578 * Obtain a specified number of slash-separated parts of a title, 579 * e.g. {{#titleparts:Hello/World|1}} => "Hello" 580 * 581 * @link https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##titleparts 582 * 583 * @param Parser $parser Parent parser 584 * @param string $title Title to split 585 * @param int $parts Number of parts to keep 586 * @param int $offset Offset starting at 1 587 * @return string 588 */ 589 public static function titleparts( Parser $parser, $title = '', $parts = 0, $offset = 0 ) { 590 $parts = (int)$parts; 591 $offset = (int)$offset; 592 $ntitle = Title::newFromText( $title ); 593 if ( $ntitle instanceof Title ) { 594 $bits = explode( '/', $ntitle->getPrefixedText(), 25 ); 595 if ( count( $bits ) <= 0 ) { 596 return $ntitle->getPrefixedText(); 597 } else { 598 if ( $offset > 0 ) { 599 --$offset; 600 } 601 if ( $parts === 0 ) { 602 return implode( '/', array_slice( $bits, $offset ) ); 603 } else { 604 return implode( '/', array_slice( $bits, $offset, $parts ) ); 605 } 606 } 607 } else { 608 return $title; 609 } 610 } 611 612 /** 613 * Verifies parameter is less than max string length. 614 * 615 * @param string $text 616 * @return bool 617 */ 618 private static function checkLength( $text ) { 619 global $wgPFStringLengthLimit; 620 return ( mb_strlen( $text ) < $wgPFStringLengthLimit ); 621 } 622 623 /** 624 * Generates error message. Called when string is too long. 625 * @return string 626 */ 627 private static function tooLongError() { 628 global $wgPFStringLengthLimit; 629 $msg = wfMessage( 'pfunc_string_too_long' )->numParams( $wgPFStringLengthLimit ); 630 return '<strong class="error">' . $msg->inContentLanguage()->escaped() . '</strong>'; 631 } 632 633 /** 634 * {{#len:string}} 635 * 636 * Reports number of characters in string. 637 * 638 * @param Parser $parser 639 * @param string $inStr 640 * @return int 641 */ 642 public static function runLen( Parser $parser, $inStr = '' ) { 643 $inStr = $parser->killMarkers( (string)$inStr ); 644 return mb_strlen( $inStr ); 645 } 646 647 /** 648 * {{#pos: string | needle | offset}} 649 * 650 * Finds first occurrence of "needle" in "string" starting at "offset". 651 * 652 * Note: If the needle is an empty string, single space is used instead. 653 * Note: If the needle is not found, empty string is returned. 654 * @param Parser $parser 655 * @param string $inStr 656 * @param int|string $inNeedle 657 * @param int $inOffset 658 * @return int|string 659 */ 660 public static function runPos( Parser $parser, $inStr = '', $inNeedle = '', $inOffset = 0 ) { 661 $inStr = $parser->killMarkers( (string)$inStr ); 662 $inNeedle = $parser->killMarkers( (string)$inNeedle ); 663 664 if ( !self::checkLength( $inStr ) || 665 !self::checkLength( $inNeedle ) ) { 666 return self::tooLongError(); 667 } 668 669 if ( $inNeedle === '' ) { 670 $inNeedle = ' '; 671 } 672 673 $pos = mb_strpos( $inStr, $inNeedle, min( (int)$inOffset, mb_strlen( $inStr ) ) ); 674 if ( $pos === false ) { 675 $pos = ''; 676 } 677 678 return $pos; 679 } 680 681 /** 682 * {{#rpos: string | needle}} 683 * 684 * Finds last occurrence of "needle" in "string". 685 * 686 * Note: If the needle is an empty string, single space is used instead. 687 * Note: If the needle is not found, -1 is returned. 688 * @param Parser $parser 689 * @param string $inStr 690 * @param int|string $inNeedle 691 * @return int|string 692 */ 693 public static function runRPos( Parser $parser, $inStr = '', $inNeedle = '' ) { 694 $inStr = $parser->killMarkers( (string)$inStr ); 695 $inNeedle = $parser->killMarkers( (string)$inNeedle ); 696 697 if ( !self::checkLength( $inStr ) || 698 !self::checkLength( $inNeedle ) ) { 699 return self::tooLongError(); 700 } 701 702 if ( $inNeedle === '' ) { 703 $inNeedle = ' '; 704 } 705 706 $pos = mb_strrpos( $inStr, $inNeedle ); 707 if ( $pos === false ) { 708 $pos = -1; 709 } 710 711 return $pos; 712 } 713 714 /** 715 * {{#sub: string | start | length }} 716 * 717 * Returns substring of "string" starting at "start" and having 718 * "length" characters. 719 * 720 * Note: If length is zero, the rest of the input is returned. 721 * Note: A negative value for "start" operates from the end of the 722 * "string". 723 * Note: A negative value for "length" returns a string reduced in 724 * length by that amount. 725 * 726 * @param Parser $parser 727 * @param string $inStr 728 * @param int $inStart 729 * @param int $inLength 730 * @return string 731 */ 732 public static function runSub( Parser $parser, $inStr = '', $inStart = 0, $inLength = 0 ) { 733 $inStr = $parser->killMarkers( (string)$inStr ); 734 735 if ( !self::checkLength( $inStr ) ) { 736 return self::tooLongError(); 737 } 738 739 if ( (int)$inLength === 0 ) { 740 $result = mb_substr( $inStr, (int)$inStart ); 741 } else { 742 $result = mb_substr( $inStr, (int)$inStart, (int)$inLength ); 743 } 744 745 return $result; 746 } 747 748 /** 749 * {{#count: string | substr }} 750 * 751 * Returns number of occurrences of "substr" in "string". 752 * 753 * Note: If "substr" is empty, a single space is used. 754 * 755 * @param Parser $parser 756 * @param string $inStr 757 * @param string $inSubStr 758 * @return int|string 759 */ 760 public static function runCount( Parser $parser, $inStr = '', $inSubStr = '' ) { 761 $inStr = $parser->killMarkers( (string)$inStr ); 762 $inSubStr = $parser->killMarkers( (string)$inSubStr ); 763 764 if ( !self::checkLength( $inStr ) || 765 !self::checkLength( $inSubStr ) ) { 766 return self::tooLongError(); 767 } 768 769 if ( $inSubStr === '' ) { 770 $inSubStr = ' '; 771 } 772 773 $result = mb_substr_count( $inStr, $inSubStr ); 774 775 return $result; 776 } 777 778 /** 779 * {{#replace:string | from | to | limit }} 780 * 781 * Replaces each occurrence of "from" in "string" with "to". 782 * At most "limit" replacements are performed. 783 * 784 * Note: Armored against replacements that would generate huge strings. 785 * Note: If "from" is an empty string, single space is used instead. 786 * 787 * @param Parser $parser 788 * @param string $inStr 789 * @param string $inReplaceFrom 790 * @param string $inReplaceTo 791 * @param int $inLimit 792 * @return mixed|string 793 */ 794 public static function runReplace( Parser $parser, $inStr = '', 795 $inReplaceFrom = '', $inReplaceTo = '', $inLimit = -1 ) { 796 global $wgPFStringLengthLimit; 797 798 $inStr = $parser->killMarkers( (string)$inStr ); 799 $inReplaceFrom = $parser->killMarkers( (string)$inReplaceFrom ); 800 $inReplaceTo = $parser->killMarkers( (string)$inReplaceTo ); 801 802 if ( !self::checkLength( $inStr ) || 803 !self::checkLength( $inReplaceFrom ) || 804 !self::checkLength( $inReplaceTo ) ) { 805 return self::tooLongError(); 806 } 807 808 if ( $inReplaceFrom === '' ) { 809 $inReplaceFrom = ' '; 810 } 811 812 // Precompute limit to avoid generating enormous string: 813 $diff = mb_strlen( $inReplaceTo ) - mb_strlen( $inReplaceFrom ); 814 if ( $diff > 0 ) { 815 $limit = ( ( $wgPFStringLengthLimit - mb_strlen( $inStr ) ) / $diff ) + 1; 816 } else { 817 $limit = -1; 818 } 819 820 $inLimit = (int)$inLimit; 821 if ( $inLimit >= 0 ) { 822 if ( $limit > $inLimit || $limit == -1 ) { 823 $limit = $inLimit; 824 } 825 } 826 827 // Use regex to allow limit and handle UTF-8 correctly. 828 $inReplaceFrom = preg_quote( $inReplaceFrom, '/' ); 829 $inReplaceTo = StringUtils::escapeRegexReplacement( $inReplaceTo ); 830 831 $result = preg_replace( '/' . $inReplaceFrom . '/u', 832 $inReplaceTo, $inStr, $limit ); 833 834 if ( !self::checkLength( $result ) ) { 835 return self::tooLongError(); 836 } 837 838 return $result; 839 } 840 841 /** 842 * {{#explode:string | delimiter | position | limit}} 843 * 844 * Breaks "string" into chunks separated by "delimiter" and returns the 845 * chunk identified by "position". 846 * 847 * Note: Negative position can be used to specify tokens from the end. 848 * Note: If the divider is an empty string, single space is used instead. 849 * Note: Empty string is returned if there are not enough exploded chunks. 850 * 851 * @param Parser $parser 852 * @param string $inStr 853 * @param string $inDiv 854 * @param int $inPos 855 * @param int|null $inLim 856 * @return string 857 */ 858 public static function runExplode( 859 Parser $parser, $inStr = '', $inDiv = '', $inPos = 0, $inLim = null 860 ) { 861 $inStr = $parser->killMarkers( (string)$inStr ); 862 $inDiv = $parser->killMarkers( (string)$inDiv ); 863 864 if ( $inDiv === '' ) { 865 $inDiv = ' '; 866 } 867 868 if ( !self::checkLength( $inStr ) || 869 !self::checkLength( $inDiv ) ) { 870 return self::tooLongError(); 871 } 872 873 $inDiv = preg_quote( $inDiv, '/' ); 874 875 $matches = preg_split( '/' . $inDiv . '/u', $inStr, $inLim ); 876 877 if ( $inPos >= 0 && isset( $matches[$inPos] ) ) { 878 $result = $matches[$inPos]; 879 } elseif ( $inPos < 0 && isset( $matches[count( $matches ) + $inPos] ) ) { 880 $result = $matches[count( $matches ) + $inPos]; 881 } else { 882 $result = ''; 883 } 884 885 return $result; 886 } 887 888 /** 889 * {{#urldecode:string}} 890 * 891 * Decodes URL-encoded (like%20that) strings. 892 * 893 * @param Parser $parser 894 * @param string $inStr 895 * @return string 896 */ 897 public static function runUrlDecode( Parser $parser, $inStr = '' ) { 898 $inStr = $parser->killMarkers( (string)$inStr ); 899 if ( !self::checkLength( $inStr ) ) { 900 return self::tooLongError(); 901 } 902 903 return urldecode( $inStr ); 904 } 905 906 /** 907 * Take a PPNode (-ish thing), expand it, remove entities, and trim. 908 * 909 * For use when doing string comparisions, where user expects entities 910 * to be equal for what they stand for (e.g. comparisions with {{PAGENAME}}) 911 * 912 * @param PPNode|string $obj Thing to expand 913 * @param PPFrame $frame 914 * @param string|null &$trimExpanded Expanded and trimmed version of PPNode, 915 * but with char refs intact 916 * @return string The trimmed, expanded and entity reference decoded version of the PPNode 917 */ 918 private static function decodeTrimExpand( $obj, PPFrame $frame, &$trimExpanded = null ) { 919 $expanded = $frame->expand( $obj ); 920 $trimExpanded = trim( $expanded ); 921 return trim( Sanitizer::decodeCharReferences( $expanded ) ); 922 } 923 924 /** 925 * @since 1.35 926 * @param Language $language 927 * @return ILanguageConverter 928 */ 929 private static function getLanguageConverter( Language $language ) : ILanguageConverter { 930 return MediaWikiServices::getInstance() 931 ->getLanguageConverterFactory() 932 ->getLanguageConverter( $language ); 933 } 934} 935