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