includes/MagicWord.php

   1 <?php
   2 /**
   3  * File for magic words.
   4  *
   5  * See docs/magicword.txt.
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  * @ingroup Parser
  24  */
  25
  26 /**
  27  * This class encapsulates "magic words" such as "#redirect", __NOTOC__, etc.
  28  *
  29  * @par Usage:
  30  * @code
  31  *     if (MagicWord::get( 'redirect' )->match( $text ) ) {
  32  *       // some code
  33  *     }
  34  * @endcode
  35  *
  36  * Possible future improvements:
  37  *   * Simultaneous searching for a number of magic words
  38  *   * MagicWord::$mObjects in shared memory
  39  *
  40  * Please avoid reading the data out of one of these objects and then writing
  41  * special case code. If possible, add another match()-like function here.
  42  *
  43  * To add magic words in an extension, use $magicWords in a file listed in
  44  * $wgExtensionMessagesFiles[].
  45  *
  46  * @par Example:
  47  * @code
  48  * $magicWords = array();
  49  *
  50  * $magicWords['en'] = array(
  51  *      'magicwordkey' => array( 0, 'case_insensitive_magic_word' ),
  52  *      'magicwordkey2' => array( 1, 'CASE_sensitive_magic_word2' ),
  53  * );
  54  * @endcode
  55  *
  56  * For magic words which are also Parser variables, add a MagicWordwgVariableIDs
  57  * hook. Use string keys.
  58  *
  59  * @ingroup Parser
  60  */
  61 class MagicWord {
  62         /**#@-*/
  63
  64         /** @var int */
  65         public $mId;
  66
  67         /** @var array */
  68         public $mSynonyms;
  69
  70         /** @var bool */
  71         public $mCaseSensitive;
  72
  73         /** @var string */
  74         private $mRegex = '';
  75
  76         /** @var string */
  77         private $mRegexStart = '';
  78
  79         /** @var string */
  80         private $mRegexStartToEnd = '';
  81
  82         /** @var string */
  83         private $mBaseRegex = '';
  84
  85         /** @var string */
  86         private $mVariableRegex = '';
  87
  88         /** @var string */
  89         private $mVariableStartToEndRegex = '';
  90
  91         /** @var bool */
  92         private $mModified = false;
  93
  94         /** @var bool */
  95         private $mFound = false;
  96
  97         static public $mVariableIDsInitialised = false;
  98         static public $mVariableIDs = array(
  99                 '!',
 100                 'currentmonth',
 101                 'currentmonth1',
 102                 'currentmonthname',
 103                 'currentmonthnamegen',
 104                 'currentmonthabbrev',
 105                 'currentday',
 106                 'currentday2',
 107                 'currentdayname',
 108                 'currentyear',
 109                 'currenttime',
 110                 'currenthour',
 111                 'localmonth',
 112                 'localmonth1',
 113                 'localmonthname',
 114                 'localmonthnamegen',
 115                 'localmonthabbrev',
 116                 'localday',
 117                 'localday2',
 118                 'localdayname',
 119                 'localyear',
 120                 'localtime',
 121                 'localhour',
 122                 'numberofarticles',
 123                 'numberoffiles',
 124                 'numberofedits',
 125                 'articlepath',
 126                 'pageid',
 127                 'sitename',
 128                 'server',
 129                 'servername',
 130                 'scriptpath',
 131                 'stylepath',
 132                 'pagename',
 133                 'pagenamee',
 134                 'fullpagename',
 135                 'fullpagenamee',
 136                 'namespace',
 137                 'namespacee',
 138                 'namespacenumber',
 139                 'currentweek',
 140                 'currentdow',
 141                 'localweek',
 142                 'localdow',
 143                 'revisionid',
 144                 'revisionday',
 145                 'revisionday2',
 146                 'revisionmonth',
 147                 'revisionmonth1',
 148                 'revisionyear',
 149                 'revisiontimestamp',
 150                 'revisionuser',
 151                 'revisionsize',
 152                 'subpagename',
 153                 'subpagenamee',
 154                 'talkspace',
 155                 'talkspacee',
 156                 'subjectspace',
 157                 'subjectspacee',
 158                 'talkpagename',
 159                 'talkpagenamee',
 160                 'subjectpagename',
 161                 'subjectpagenamee',
 162                 'numberofusers',
 163                 'numberofactiveusers',
 164                 'numberofpages',
 165                 'currentversion',
 166                 'rootpagename',
 167                 'rootpagenamee',
 168                 'basepagename',
 169                 'basepagenamee',
 170                 'currenttimestamp',
 171                 'localtimestamp',
 172                 'directionmark',
 173                 'contentlanguage',
 174                 'numberofadmins',
 175                 'cascadingsources',
 176         );
 177
 178         /* Array of caching hints for ParserCache */
 179         static public $mCacheTTLs = array(
 180                 'currentmonth' => 86400,
 181                 'currentmonth1' => 86400,
 182                 'currentmonthname' => 86400,
 183                 'currentmonthnamegen' => 86400,
 184                 'currentmonthabbrev' => 86400,
 185                 'currentday' => 3600,
 186                 'currentday2' => 3600,
 187                 'currentdayname' => 3600,
 188                 'currentyear' => 86400,
 189                 'currenttime' => 3600,
 190                 'currenthour' => 3600,
 191                 'localmonth' => 86400,
 192                 'localmonth1' => 86400,
 193                 'localmonthname' => 86400,
 194                 'localmonthnamegen' => 86400,
 195                 'localmonthabbrev' => 86400,
 196                 'localday' => 3600,
 197                 'localday2' => 3600,
 198                 'localdayname' => 3600,
 199                 'localyear' => 86400,
 200                 'localtime' => 3600,
 201                 'localhour' => 3600,
 202                 'numberofarticles' => 3600,
 203                 'numberoffiles' => 3600,
 204                 'numberofedits' => 3600,
 205                 'currentweek' => 3600,
 206                 'currentdow' => 3600,
 207                 'localweek' => 3600,
 208                 'localdow' => 3600,
 209                 'numberofusers' => 3600,
 210                 'numberofactiveusers' => 3600,
 211                 'numberofpages' => 3600,
 212                 'currentversion' => 86400,
 213                 'currenttimestamp' => 3600,
 214                 'localtimestamp' => 3600,
 215                 'pagesinnamespace' => 3600,
 216                 'numberofadmins' => 3600,
 217                 'numberingroup' => 3600,
 218         );
 219
 220         static public $mDoubleUnderscoreIDs = array(
 221                 'notoc',
 222                 'nogallery',
 223                 'forcetoc',
 224                 'toc',
 225                 'noeditsection',
 226                 'newsectionlink',
 227                 'nonewsectionlink',
 228                 'hiddencat',
 229                 'index',
 230                 'noindex',
 231                 'staticredirect',
 232                 'notitleconvert',
 233                 'nocontentconvert',
 234         );
 235
 236         static public $mSubstIDs = array(
 237                 'subst',
 238                 'safesubst',
 239         );
 240
 241         static public $mObjects = array();
 242         static public $mDoubleUnderscoreArray = null;
 243
 244         /**#@-*/
 245
 246         function __construct( $id = 0, $syn = array(), $cs = false ) {
 247                 $this->mId = $id;
 248                 $this->mSynonyms = (array)$syn;
 249                 $this->mCaseSensitive = $cs;
 250         }
 251
 252         /**
 253          * Factory: creates an object representing an ID
 254          *
 255          * @param int $id
 256          *
 257          * @return MagicWord
 258          */
 259         static function &get( $id ) {
 260                 if ( !isset( self::$mObjects[$id] ) ) {
 261                         $mw = new MagicWord();
 262                         $mw->load( $id );
 263                         self::$mObjects[$id] = $mw;
 264                 }
 265                 return self::$mObjects[$id];
 266         }
 267
 268         /**
 269          * Get an array of parser variable IDs
 270          *
 271          * @return array
 272          */
 273         static function getVariableIDs() {
 274                 if ( !self::$mVariableIDsInitialised ) {
 275                         # Get variable IDs
 276                         Hooks::run( 'MagicWordwgVariableIDs', array( &self::$mVariableIDs ) );
 277                         self::$mVariableIDsInitialised = true;
 278                 }
 279                 return self::$mVariableIDs;
 280         }
 281
 282         /**
 283          * Get an array of parser substitution modifier IDs
 284          * @return array
 285          */
 286         static function getSubstIDs() {
 287                 return self::$mSubstIDs;
 288         }
 289
 290         /**
 291          * Allow external reads of TTL array
 292          *
 293          * @param int $id
 294          * @return int
 295          */
 296         static function getCacheTTL( $id ) {
 297                 if ( array_key_exists( $id, self::$mCacheTTLs ) ) {
 298                         return self::$mCacheTTLs[$id];
 299                 } else {
 300                         return -1;
 301                 }
 302         }
 303
 304         /**
 305          * Get a MagicWordArray of double-underscore entities
 306          *
 307          * @return MagicWordArray
 308          */
 309         static function getDoubleUnderscoreArray() {
 310                 if ( is_null( self::$mDoubleUnderscoreArray ) ) {
 311                         Hooks::run( 'GetDoubleUnderscoreIDs', array( &self::$mDoubleUnderscoreIDs ) );
 312                         self::$mDoubleUnderscoreArray = new MagicWordArray( self::$mDoubleUnderscoreIDs );
 313                 }
 314                 return self::$mDoubleUnderscoreArray;
 315         }
 316
 317         /**
 318          * Clear the self::$mObjects variable
 319          * For use in parser tests
 320          */
 321         public static function clearCache() {
 322                 self::$mObjects = array();
 323         }
 324
 325         /**
 326          * Initialises this object with an ID
 327          *
 328          * @param int $id
 329          * @throws MWException
 330          */
 331         function load( $id ) {
 332                 global $wgContLang;
 333                 $this->mId = $id;
 334                 $wgContLang->getMagic( $this );
 335                 if ( !$this->mSynonyms ) {
 336                         $this->mSynonyms = array( 'brionmademeputthishere' );
 337                         throw new MWException( "Error: invalid magic word '$id'" );
 338                 }
 339         }
 340
 341         /**
 342          * Preliminary initialisation
 343          * @private
 344          */
 345         function initRegex() {
 346                 // Sort the synonyms by length, descending, so that the longest synonym
 347                 // matches in precedence to the shortest
 348                 $synonyms = $this->mSynonyms;
 349                 usort( $synonyms, array( $this, 'compareStringLength' ) );
 350
 351                 $escSyn = array();
 352                 foreach ( $synonyms as $synonym ) {
 353                         // In case a magic word contains /, like that's going to happen;)
 354                         $escSyn[] = preg_quote( $synonym, '/' );
 355                 }
 356                 $this->mBaseRegex = implode( '|', $escSyn );
 357
 358                 $case = $this->mCaseSensitive ? '' : 'iu';
 359                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 360                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 361                 $this->mRegexStartToEnd = "/^(?:{$this->mBaseRegex})$/{$case}";
 362                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 363                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 364                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 365         }
 366
 367         /**
 368          * A comparison function that returns -1, 0 or 1 depending on whether the
 369          * first string is longer, the same length or shorter than the second
 370          * string.
 371          *
 372          * @param string $s1
 373          * @param string $s2
 374          *
 375          * @return int
 376          */
 377         function compareStringLength( $s1, $s2 ) {
 378                 $l1 = strlen( $s1 );
 379                 $l2 = strlen( $s2 );
 380                 if ( $l1 < $l2 ) {
 381                         return 1;
 382                 } elseif ( $l1 > $l2 ) {
 383                         return -1;
 384                 } else {
 385                         return 0;
 386                 }
 387         }
 388
 389         /**
 390          * Gets a regex representing matching the word
 391          *
 392          * @return string
 393          */
 394         function getRegex() {
 395                 if ( $this->mRegex == '' ) {
 396                         $this->initRegex();
 397                 }
 398                 return $this->mRegex;
 399         }
 400
 401         /**
 402          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 403          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 404          * the complete expression
 405          *
 406          * @return string
 407          */
 408         function getRegexCase() {
 409                 if ( $this->mRegex === '' ) {
 410                         $this->initRegex();
 411                 }
 412
 413                 return $this->mCaseSensitive ? '' : 'iu';
 414         }
 415
 416         /**
 417          * Gets a regex matching the word, if it is at the string start
 418          *
 419          * @return string
 420          */
 421         function getRegexStart() {
 422                 if ( $this->mRegex == '' ) {
 423                         $this->initRegex();
 424                 }
 425                 return $this->mRegexStart;
 426         }
 427
 428         /**
 429          * Gets a regex matching the word from start to end of a string
 430          *
 431          * @return string
 432          * @since 1.23
 433          */
 434         function getRegexStartToEnd() {
 435                 if ( $this->mRegexStartToEnd == '' ) {
 436                         $this->initRegex();
 437                 }
 438                 return $this->mRegexStartToEnd;
 439         }
 440
 441         /**
 442          * regex without the slashes and what not
 443          *
 444          * @return string
 445          */
 446         function getBaseRegex() {
 447                 if ( $this->mRegex == '' ) {
 448                         $this->initRegex();
 449                 }
 450                 return $this->mBaseRegex;
 451         }
 452
 453         /**
 454          * Returns true if the text contains the word
 455          *
 456          * @param string $text
 457          *
 458          * @return bool
 459          */
 460         function match( $text ) {
 461                 return (bool)preg_match( $this->getRegex(), $text );
 462         }
 463
 464         /**
 465          * Returns true if the text starts with the word
 466          *
 467          * @param string $text
 468          *
 469          * @return bool
 470          */
 471         function matchStart( $text ) {
 472                 return (bool)preg_match( $this->getRegexStart(), $text );
 473         }
 474
 475         /**
 476          * Returns true if the text matched the word
 477          *
 478          * @param string $text
 479          *
 480          * @return bool
 481          * @since 1.23
 482          */
 483         function matchStartToEnd( $text ) {
 484                 return (bool)preg_match( $this->getRegexStartToEnd(), $text );
 485         }
 486
 487         /**
 488          * Returns NULL if there's no match, the value of $1 otherwise
 489          * The return code is the matched string, if there's no variable
 490          * part in the regex and the matched variable part ($1) if there
 491          * is one.
 492          *
 493          * @param string $text
 494          *
 495          * @return string
 496          */
 497         function matchVariableStartToEnd( $text ) {
 498                 $matches = array();
 499                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 500                 if ( $matchcount == 0 ) {
 501                         return null;
 502                 } else {
 503                         # multiple matched parts (variable match); some will be empty because of
 504                         # synonyms. The variable will be the second non-empty one so remove any
 505                         # blank elements and re-sort the indices.
 506                         # See also bug 6526
 507
 508                         $matches = array_values( array_filter( $matches ) );
 509
 510                         if ( count( $matches ) == 1 ) {
 511                                 return $matches[0];
 512                         } else {
 513                                 return $matches[1];
 514                         }
 515                 }
 516         }
 517
 518         /**
 519          * Returns true if the text matches the word, and alters the
 520          * input string, removing all instances of the word
 521          *
 522          * @param string $text
 523          *
 524          * @return bool
 525          */
 526         function matchAndRemove( &$text ) {
 527                 $this->mFound = false;
 528                 $text = preg_replace_callback(
 529                         $this->getRegex(),
 530                         array( &$this, 'pregRemoveAndRecord' ),
 531                         $text
 532                 );
 533
 534                 return $this->mFound;
 535         }
 536
 537         /**
 538          * @param string $text
 539          * @return bool
 540          */
 541         function matchStartAndRemove( &$text ) {
 542                 $this->mFound = false;
 543                 $text = preg_replace_callback(
 544                         $this->getRegexStart(),
 545                         array( &$this, 'pregRemoveAndRecord' ),
 546                         $text
 547                 );
 548
 549                 return $this->mFound;
 550         }
 551
 552         /**
 553          * Used in matchAndRemove()
 554          *
 555          * @return string
 556          */
 557         function pregRemoveAndRecord() {
 558                 $this->mFound = true;
 559                 return '';
 560         }
 561
 562         /**
 563          * Replaces the word with something else
 564          *
 565          * @param string $replacement
 566          * @param string $subject
 567          * @param int $limit
 568          *
 569          * @return string
 570          */
 571         function replace( $replacement, $subject, $limit = -1 ) {
 572                 $res = preg_replace(
 573                         $this->getRegex(),
 574                         StringUtils::escapeRegexReplacement( $replacement ),
 575                         $subject,
 576                         $limit
 577                 );
 578                 $this->mModified = $res !== $subject;
 579                 return $res;
 580         }
 581
 582         /**
 583          * Variable handling: {{SUBST:xxx}} style words
 584          * Calls back a function to determine what to replace xxx with
 585          * Input word must contain $1
 586          *
 587          * @param string $text
 588          * @param callable $callback
 589          *
 590          * @return string
 591          */
 592         function substituteCallback( $text, $callback ) {
 593                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 594                 $this->mModified = $res !== $text;
 595                 return $res;
 596         }
 597
 598         /**
 599          * Matches the word, where $1 is a wildcard
 600          *
 601          * @return string
 602          */
 603         function getVariableRegex() {
 604                 if ( $this->mVariableRegex == '' ) {
 605                         $this->initRegex();
 606                 }
 607                 return $this->mVariableRegex;
 608         }
 609
 610         /**
 611          * Matches the entire string, where $1 is a wildcard
 612          *
 613          * @return string
 614          */
 615         function getVariableStartToEndRegex() {
 616                 if ( $this->mVariableStartToEndRegex == '' ) {
 617                         $this->initRegex();
 618                 }
 619                 return $this->mVariableStartToEndRegex;
 620         }
 621
 622         /**
 623          * Accesses the synonym list directly
 624          *
 625          * @param int $i
 626          *
 627          * @return string
 628          */
 629         function getSynonym( $i ) {
 630                 return $this->mSynonyms[$i];
 631         }
 632
 633         /**
 634          * @return array
 635          */
 636         function getSynonyms() {
 637                 return $this->mSynonyms;
 638         }
 639
 640         /**
 641          * Returns true if the last call to replace() or substituteCallback()
 642          * returned a modified text, otherwise false.
 643          *
 644          * @return bool
 645          */
 646         function getWasModified() {
 647                 return $this->mModified;
 648         }
 649
 650         /**
 651          * $magicarr is an associative array of (magic word ID => replacement)
 652          * This method uses the php feature to do several replacements at the same time,
 653          * thereby gaining some efficiency. The result is placed in the out variable
 654          * $result. The return value is true if something was replaced.
 655          * @deprecated since 1.25, unused
 656          *
 657          * @param array $magicarr
 658          * @param string $subject
 659          * @param string $result
 660          *
 661          * @return bool
 662          */
 663         function replaceMultiple( $magicarr, $subject, &$result ) {
 664                 wfDeprecated( __METHOD__, '1.25' );
 665                 $search = array();
 666                 $replace = array();
 667                 foreach ( $magicarr as $id => $replacement ) {
 668                         $mw = MagicWord::get( $id );
 669                         $search[] = $mw->getRegex();
 670                         $replace[] = $replacement;
 671                 }
 672
 673                 $result = preg_replace( $search, $replace, $subject );
 674                 return $result !== $subject;
 675         }
 676
 677         /**
 678          * Adds all the synonyms of this MagicWord to an array, to allow quick
 679          * lookup in a list of magic words
 680          *
 681          * @param array $array
 682          * @param string $value
 683          */
 684         function addToArray( &$array, $value ) {
 685                 global $wgContLang;
 686                 foreach ( $this->mSynonyms as $syn ) {
 687                         $array[$wgContLang->lc( $syn )] = $value;
 688                 }
 689         }
 690
 691         /**
 692          * @return bool
 693          */
 694         function isCaseSensitive() {
 695                 return $this->mCaseSensitive;
 696         }
 697
 698         /**
 699          * @return int
 700          */
 701         function getId() {
 702                 return $this->mId;
 703         }
 704 }
 705
 706 /**
 707  * Class for handling an array of magic words
 708  * @ingroup Parser
 709  */
 710 class MagicWordArray {
 711         /** @var array */
 712         public $names = array();
 713
 714         /** @var array */
 715         private $hash;
 716
 717         private $baseRegex;
 718
 719         private $regex;
 720
 721         /**
 722          * @param array $names
 723          */
 724         function __construct( $names = array() ) {
 725                 $this->names = $names;
 726         }
 727
 728         /**
 729          * Add a magic word by name
 730          *
 731          * @param string $name
 732          */
 733         public function add( $name ) {
 734                 $this->names[] = $name;
 735                 $this->hash = $this->baseRegex = $this->regex = null;
 736         }
 737
 738         /**
 739          * Add a number of magic words by name
 740          *
 741          * @param array $names
 742          */
 743         public function addArray( $names ) {
 744                 $this->names = array_merge( $this->names, array_values( $names ) );
 745                 $this->hash = $this->baseRegex = $this->regex = null;
 746         }
 747
 748         /**
 749          * Get a 2-d hashtable for this array
 750          * @return array
 751          */
 752         function getHash() {
 753                 if ( is_null( $this->hash ) ) {
 754                         global $wgContLang;
 755                         $this->hash = array( 0 => array(), 1 => array() );
 756                         foreach ( $this->names as $name ) {
 757                                 $magic = MagicWord::get( $name );
 758                                 $case = intval( $magic->isCaseSensitive() );
 759                                 foreach ( $magic->getSynonyms() as $syn ) {
 760                                         if ( !$case ) {
 761                                                 $syn = $wgContLang->lc( $syn );
 762                                         }
 763                                         $this->hash[$case][$syn] = $name;
 764                                 }
 765                         }
 766                 }
 767                 return $this->hash;
 768         }
 769
 770         /**
 771          * Get the base regex
 772          * @return array
 773          */
 774         function getBaseRegex() {
 775                 if ( is_null( $this->baseRegex ) ) {
 776                         $this->baseRegex = array( 0 => '', 1 => '' );
 777                         foreach ( $this->names as $name ) {
 778                                 $magic = MagicWord::get( $name );
 779                                 $case = intval( $magic->isCaseSensitive() );
 780                                 foreach ( $magic->getSynonyms() as $i => $syn ) {
 781                                         // Group name must start with a non-digit in PCRE 8.34+
 782                                         $it = strtr( $i, '0123456789', 'abcdefghij' );
 783                                         $group = "(?P<{$it}_{$name}>" . preg_quote( $syn, '/' ) . ')';
 784                                         if ( $this->baseRegex[$case] === '' ) {
 785                                                 $this->baseRegex[$case] = $group;
 786                                         } else {
 787                                                 $this->baseRegex[$case] .= '|' . $group;
 788                                         }
 789                                 }
 790                         }
 791                 }
 792                 return $this->baseRegex;
 793         }
 794
 795         /**
 796          * Get an unanchored regex that does not match parameters
 797          * @return array
 798          */
 799         function getRegex() {
 800                 if ( is_null( $this->regex ) ) {
 801                         $base = $this->getBaseRegex();
 802                         $this->regex = array( '', '' );
 803                         if ( $this->baseRegex[0] !== '' ) {
 804                                 $this->regex[0] = "/{$base[0]}/iuS";
 805                         }
 806                         if ( $this->baseRegex[1] !== '' ) {
 807                                 $this->regex[1] = "/{$base[1]}/S";
 808                         }
 809                 }
 810                 return $this->regex;
 811         }
 812
 813         /**
 814          * Get a regex for matching variables with parameters
 815          *
 816          * @return string
 817          */
 818         function getVariableRegex() {
 819                 return str_replace( "\\$1", "(.*?)", $this->getRegex() );
 820         }
 821
 822         /**
 823          * Get a regex anchored to the start of the string that does not match parameters
 824          *
 825          * @return array
 826          */
 827         function getRegexStart() {
 828                 $base = $this->getBaseRegex();
 829                 $newRegex = array( '', '' );
 830                 if ( $base[0] !== '' ) {
 831                         $newRegex[0] = "/^(?:{$base[0]})/iuS";
 832                 }
 833                 if ( $base[1] !== '' ) {
 834                         $newRegex[1] = "/^(?:{$base[1]})/S";
 835                 }
 836                 return $newRegex;
 837         }
 838
 839         /**
 840          * Get an anchored regex for matching variables with parameters
 841          *
 842          * @return array
 843          */
 844         function getVariableStartToEndRegex() {
 845                 $base = $this->getBaseRegex();
 846                 $newRegex = array( '', '' );
 847                 if ( $base[0] !== '' ) {
 848                         $newRegex[0] = str_replace( "\\$1", "(.*?)", "/^(?:{$base[0]})$/iuS" );
 849                 }
 850                 if ( $base[1] !== '' ) {
 851                         $newRegex[1] = str_replace( "\\$1", "(.*?)", "/^(?:{$base[1]})$/S" );
 852                 }
 853                 return $newRegex;
 854         }
 855
 856         /**
 857          * @since 1.20
 858          * @return array
 859          */
 860         public function getNames() {
 861                 return $this->names;
 862         }
 863
 864         /**
 865          * Parse a match array from preg_match
 866          * Returns array(magic word ID, parameter value)
 867          * If there is no parameter value, that element will be false.
 868          *
 869          * @param array $m
 870          *
 871          * @throws MWException
 872          * @return array
 873          */
 874         function parseMatch( $m ) {
 875                 reset( $m );
 876                 while ( list( $key, $value ) = each( $m ) ) {
 877                         if ( $key === 0 || $value === '' ) {
 878                                 continue;
 879                         }
 880                         $parts = explode( '_', $key, 2 );
 881                         if ( count( $parts ) != 2 ) {
 882                                 // This shouldn't happen
 883                                 // continue;
 884                                 throw new MWException( __METHOD__ . ': bad parameter name' );
 885                         }
 886                         list( /* $synIndex */, $magicName ) = $parts;
 887                         $paramValue = next( $m );
 888                         return array( $magicName, $paramValue );
 889                 }
 890                 // This shouldn't happen either
 891                 throw new MWException( __METHOD__ . ': parameter not found' );
 892         }
 893
 894         /**
 895          * Match some text, with parameter capture
 896          * Returns an array with the magic word name in the first element and the
 897          * parameter in the second element.
 898          * Both elements are false if there was no match.
 899          *
 900          * @param string $text
 901          *
 902          * @return array
 903          */
 904         public function matchVariableStartToEnd( $text ) {
 905                 $regexes = $this->getVariableStartToEndRegex();
 906                 foreach ( $regexes as $regex ) {
 907                         if ( $regex !== '' ) {
 908                                 $m = array();
 909                                 if ( preg_match( $regex, $text, $m ) ) {
 910                                         return $this->parseMatch( $m );
 911                                 }
 912                         }
 913                 }
 914                 return array( false, false );
 915         }
 916
 917         /**
 918          * Match some text, without parameter capture
 919          * Returns the magic word name, or false if there was no capture
 920          *
 921          * @param string $text
 922          *
 923          * @return string|bool False on failure
 924          */
 925         public function matchStartToEnd( $text ) {
 926                 $hash = $this->getHash();
 927                 if ( isset( $hash[1][$text] ) ) {
 928                         return $hash[1][$text];
 929                 }
 930                 global $wgContLang;
 931                 $lc = $wgContLang->lc( $text );
 932                 if ( isset( $hash[0][$lc] ) ) {
 933                         return $hash[0][$lc];
 934                 }
 935                 return false;
 936         }
 937
 938         /**
 939          * Returns an associative array, ID => param value, for all items that match
 940          * Removes the matched items from the input string (passed by reference)
 941          *
 942          * @param string $text
 943          *
 944          * @return array
 945          */
 946         public function matchAndRemove( &$text ) {
 947                 $found = array();
 948                 $regexes = $this->getRegex();
 949                 foreach ( $regexes as $regex ) {
 950                         if ( $regex === '' ) {
 951                                 continue;
 952                         }
 953                         $matches = array();
 954                         if ( preg_match_all( $regex, $text, $matches, PREG_SET_ORDER ) ) {
 955                                 foreach ( $matches as $m ) {
 956                                         list( $name, $param ) = $this->parseMatch( $m );
 957                                         $found[$name] = $param;
 958                                 }
 959                         }
 960                         $text = preg_replace( $regex, '', $text );
 961                 }
 962                 return $found;
 963         }
 964
 965         /**
 966          * Return the ID of the magic word at the start of $text, and remove
 967          * the prefix from $text.
 968          * Return false if no match found and $text is not modified.
 969          * Does not match parameters.
 970          *
 971          * @param string $text
 972          *
 973          * @return int|bool False on failure
 974          */
 975         public function matchStartAndRemove( &$text ) {
 976                 $regexes = $this->getRegexStart();
 977                 foreach ( $regexes as $regex ) {
 978                         if ( $regex === '' ) {
 979                                 continue;
 980                         }
 981                         if ( preg_match( $regex, $text, $m ) ) {
 982                                 list( $id, ) = $this->parseMatch( $m );
 983                                 if ( strlen( $m[0] ) >= strlen( $text ) ) {
 984                                         $text = '';
 985                                 } else {
 986                                         $text = substr( $text, strlen( $m[0] ) );
 987                                 }
 988                                 return $id;
 989                         }
 990                 }
 991                 return false;
 992         }
 993 }