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                         wfRunHooks( '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::$mDoubleUnderscoreArray ) ) {
 311                         wfRunHooks( 'GetDoubleUnderscoreIDs', array( &self::$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                 wfProfileIn( __METHOD__ );
 334                 $this->mId = $id;
 335                 $wgContLang->getMagic( $this );
 336                 if ( !$this->mSynonyms ) {
 337                         $this->mSynonyms = array( 'brionmademeputthishere' );
 338                         wfProfileOut( __METHOD__ );
 339                         throw new MWException( "Error: invalid magic word '$id'" );
 340                 }
 341                 wfProfileOut( __METHOD__ );
 342         }
 343
 344         /**
 345          * Preliminary initialisation
 346          * @private
 347          */
 348         function initRegex() {
 349                 // Sort the synonyms by length, descending, so that the longest synonym
 350                 // matches in precedence to the shortest
 351                 $synonyms = $this->mSynonyms;
 352                 usort( $synonyms, array( $this, 'compareStringLength' ) );
 353
 354                 $escSyn = array();
 355                 foreach ( $synonyms as $synonym ) {
 356                         // In case a magic word contains /, like that's going to happen;)
 357                         $escSyn[] = preg_quote( $synonym, '/' );
 358                 }
 359                 $this->mBaseRegex = implode( '|', $escSyn );
 360
 361                 $case = $this->mCaseSensitive ? '' : 'iu';
 362                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 363                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 364                 $this->mRegexStartToEnd = "/^(?:{$this->mBaseRegex})$/{$case}";
 365                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 366                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 367                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 368         }
 369
 370         /**
 371          * A comparison function that returns -1, 0 or 1 depending on whether the
 372          * first string is longer, the same length or shorter than the second
 373          * string.
 374          *
 375          * @param string $s1
 376          * @param string $s2
 377          *
 378          * @return int
 379          */
 380         function compareStringLength( $s1, $s2 ) {
 381                 $l1 = strlen( $s1 );
 382                 $l2 = strlen( $s2 );
 383                 if ( $l1 < $l2 ) {
 384                         return 1;
 385                 } elseif ( $l1 > $l2 ) {
 386                         return -1;
 387                 } else {
 388                         return 0;
 389                 }
 390         }
 391
 392         /**
 393          * Gets a regex representing matching the word
 394          *
 395          * @return string
 396          */
 397         function getRegex() {
 398                 if ( $this->mRegex == '' ) {
 399                         $this->initRegex();
 400                 }
 401                 return $this->mRegex;
 402         }
 403
 404         /**
 405          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 406          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 407          * the complete expression
 408          *
 409          * @return string
 410          */
 411         function getRegexCase() {
 412                 if ( $this->mRegex === '' ) {
 413                         $this->initRegex();
 414                 }
 415
 416                 return $this->mCaseSensitive ? '' : 'iu';
 417         }
 418
 419         /**
 420          * Gets a regex matching the word, if it is at the string start
 421          *
 422          * @return string
 423          */
 424         function getRegexStart() {
 425                 if ( $this->mRegex == '' ) {
 426                         $this->initRegex();
 427                 }
 428                 return $this->mRegexStart;
 429         }
 430
 431         /**
 432          * Gets a regex matching the word from start to end of a string
 433          *
 434          * @return string
 435          * @since 1.23
 436          */
 437         function getRegexStartToEnd() {
 438                 if ( $this->mRegexStartToEnd == '' ) {
 439                         $this->initRegex();
 440                 }
 441                 return $this->mRegexStartToEnd;
 442         }
 443
 444         /**
 445          * regex without the slashes and what not
 446          *
 447          * @return string
 448          */
 449         function getBaseRegex() {
 450                 if ( $this->mRegex == '' ) {
 451                         $this->initRegex();
 452                 }
 453                 return $this->mBaseRegex;
 454         }
 455
 456         /**
 457          * Returns true if the text contains the word
 458          *
 459          * @param string $text
 460          *
 461          * @return bool
 462          */
 463         function match( $text ) {
 464                 return (bool)preg_match( $this->getRegex(), $text );
 465         }
 466
 467         /**
 468          * Returns true if the text starts with the word
 469          *
 470          * @param string $text
 471          *
 472          * @return bool
 473          */
 474         function matchStart( $text ) {
 475                 return (bool)preg_match( $this->getRegexStart(), $text );
 476         }
 477
 478         /**
 479          * Returns true if the text matched the word
 480          *
 481          * @param string $text
 482          *
 483          * @return bool
 484          * @since 1.23
 485          */
 486         function matchStartToEnd( $text ) {
 487                 return (bool)preg_match( $this->getRegexStartToEnd(), $text );
 488         }
 489
 490         /**
 491          * Returns NULL if there's no match, the value of $1 otherwise
 492          * The return code is the matched string, if there's no variable
 493          * part in the regex and the matched variable part ($1) if there
 494          * is one.
 495          *
 496          * @param string $text
 497          *
 498          * @return string
 499          */
 500         function matchVariableStartToEnd( $text ) {
 501                 $matches = array();
 502                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 503                 if ( $matchcount == 0 ) {
 504                         return null;
 505                 } else {
 506                         # multiple matched parts (variable match); some will be empty because of
 507                         # synonyms. The variable will be the second non-empty one so remove any
 508                         # blank elements and re-sort the indices.
 509                         # See also bug 6526
 510
 511                         $matches = array_values( array_filter( $matches ) );
 512
 513                         if ( count( $matches ) == 1 ) {
 514                                 return $matches[0];
 515                         } else {
 516                                 return $matches[1];
 517                         }
 518                 }
 519         }
 520
 521         /**
 522          * Returns true if the text matches the word, and alters the
 523          * input string, removing all instances of the word
 524          *
 525          * @param string $text
 526          *
 527          * @return bool
 528          */
 529         function matchAndRemove( &$text ) {
 530                 $this->mFound = false;
 531                 $text = preg_replace_callback(
 532                         $this->getRegex(),
 533                         array( &$this, 'pregRemoveAndRecord' ),
 534                         $text
 535                 );
 536
 537                 return $this->mFound;
 538         }
 539
 540         /**
 541          * @param string $text
 542          * @return bool
 543          */
 544         function matchStartAndRemove( &$text ) {
 545                 $this->mFound = false;
 546                 $text = preg_replace_callback(
 547                         $this->getRegexStart(),
 548                         array( &$this, 'pregRemoveAndRecord' ),
 549                         $text
 550                 );
 551
 552                 return $this->mFound;
 553         }
 554
 555         /**
 556          * Used in matchAndRemove()
 557          *
 558          * @return string
 559          */
 560         function pregRemoveAndRecord() {
 561                 $this->mFound = true;
 562                 return '';
 563         }
 564
 565         /**
 566          * Replaces the word with something else
 567          *
 568          * @param string $replacement
 569          * @param string $subject
 570          * @param int $limit
 571          *
 572          * @return string
 573          */
 574         function replace( $replacement, $subject, $limit = -1 ) {
 575                 $res = preg_replace(
 576                         $this->getRegex(),
 577                         StringUtils::escapeRegexReplacement( $replacement ),
 578                         $subject,
 579                         $limit
 580                 );
 581                 $this->mModified = $res !== $subject;
 582                 return $res;
 583         }
 584
 585         /**
 586          * Variable handling: {{SUBST:xxx}} style words
 587          * Calls back a function to determine what to replace xxx with
 588          * Input word must contain $1
 589          *
 590          * @param string $text
 591          * @param callable $callback
 592          *
 593          * @return string
 594          */
 595         function substituteCallback( $text, $callback ) {
 596                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 597                 $this->mModified = $res !== $text;
 598                 return $res;
 599         }
 600
 601         /**
 602          * Matches the word, where $1 is a wildcard
 603          *
 604          * @return string
 605          */
 606         function getVariableRegex() {
 607                 if ( $this->mVariableRegex == '' ) {
 608                         $this->initRegex();
 609                 }
 610                 return $this->mVariableRegex;
 611         }
 612
 613         /**
 614          * Matches the entire string, where $1 is a wildcard
 615          *
 616          * @return string
 617          */
 618         function getVariableStartToEndRegex() {
 619                 if ( $this->mVariableStartToEndRegex == '' ) {
 620                         $this->initRegex();
 621                 }
 622                 return $this->mVariableStartToEndRegex;
 623         }
 624
 625         /**
 626          * Accesses the synonym list directly
 627          *
 628          * @param int $i
 629          *
 630          * @return string
 631          */
 632         function getSynonym( $i ) {
 633                 return $this->mSynonyms[$i];
 634         }
 635
 636         /**
 637          * @return array
 638          */
 639         function getSynonyms() {
 640                 return $this->mSynonyms;
 641         }
 642
 643         /**
 644          * Returns true if the last call to replace() or substituteCallback()
 645          * returned a modified text, otherwise false.
 646          *
 647          * @return bool
 648          */
 649         function getWasModified() {
 650                 return $this->mModified;
 651         }
 652
 653         /**
 654          * $magicarr is an associative array of (magic word ID => replacement)
 655          * This method uses the php feature to do several replacements at the same time,
 656          * thereby gaining some efficiency. The result is placed in the out variable
 657          * $result. The return value is true if something was replaced.
 658          * @todo Should this be static? It doesn't seem to be used at all
 659          *
 660          * @param array $magicarr
 661          * @param string $subject
 662          * @param string $result
 663          *
 664          * @return bool
 665          */
 666         function replaceMultiple( $magicarr, $subject, &$result ) {
 667                 $search = array();
 668                 $replace = array();
 669                 foreach ( $magicarr as $id => $replacement ) {
 670                         $mw = MagicWord::get( $id );
 671                         $search[] = $mw->getRegex();
 672                         $replace[] = $replacement;
 673                 }
 674
 675                 $result = preg_replace( $search, $replace, $subject );
 676                 return $result !== $subject;
 677         }
 678
 679         /**
 680          * Adds all the synonyms of this MagicWord to an array, to allow quick
 681          * lookup in a list of magic words
 682          *
 683          * @param array $array
 684          * @param string $value
 685          */
 686         function addToArray( &$array, $value ) {
 687                 global $wgContLang;
 688                 foreach ( $this->mSynonyms as $syn ) {
 689                         $array[$wgContLang->lc( $syn )] = $value;
 690                 }
 691         }
 692
 693         /**
 694          * @return bool
 695          */
 696         function isCaseSensitive() {
 697                 return $this->mCaseSensitive;
 698         }
 699
 700         /**
 701          * @return int
 702          */
 703         function getId() {
 704                 return $this->mId;
 705         }
 706 }
 707
 708 /**
 709  * Class for handling an array of magic words
 710  * @ingroup Parser
 711  */
 712 class MagicWordArray {
 713         /** @var array */
 714         public $names = array();
 715
 716         /** @var array */
 717         private $hash;
 718
 719         private $baseRegex;
 720
 721         private $regex;
 722
 723         /** @todo Unused? */
 724         private $matches;
 725
 726         /**
 727          * @param array $names
 728          */
 729         function __construct( $names = array() ) {
 730                 $this->names = $names;
 731         }
 732
 733         /**
 734          * Add a magic word by name
 735          *
 736          * @param string $name
 737          */
 738         public function add( $name ) {
 739                 $this->names[] = $name;
 740                 $this->hash = $this->baseRegex = $this->regex = null;
 741         }
 742
 743         /**
 744          * Add a number of magic words by name
 745          *
 746          * @param array $names
 747          */
 748         public function addArray( $names ) {
 749                 $this->names = array_merge( $this->names, array_values( $names ) );
 750                 $this->hash = $this->baseRegex = $this->regex = null;
 751         }
 752
 753         /**
 754          * Get a 2-d hashtable for this array
 755          * @return array
 756          */
 757         function getHash() {
 758                 if ( is_null( $this->hash ) ) {
 759                         global $wgContLang;
 760                         $this->hash = array( 0 => array(), 1 => array() );
 761                         foreach ( $this->names as $name ) {
 762                                 $magic = MagicWord::get( $name );
 763                                 $case = intval( $magic->isCaseSensitive() );
 764                                 foreach ( $magic->getSynonyms() as $syn ) {
 765                                         if ( !$case ) {
 766                                                 $syn = $wgContLang->lc( $syn );
 767                                         }
 768                                         $this->hash[$case][$syn] = $name;
 769                                 }
 770                         }
 771                 }
 772                 return $this->hash;
 773         }
 774
 775         /**
 776          * Get the base regex
 777          * @return array
 778          */
 779         function getBaseRegex() {
 780                 if ( is_null( $this->baseRegex ) ) {
 781                         $this->baseRegex = array( 0 => '', 1 => '' );
 782                         foreach ( $this->names as $name ) {
 783                                 $magic = MagicWord::get( $name );
 784                                 $case = intval( $magic->isCaseSensitive() );
 785                                 foreach ( $magic->getSynonyms() as $i => $syn ) {
 786                                         // Group name must start with a non-digit in PCRE 8.34+
 787                                         $it = strtr( $i, '0123456789', 'abcdefghij' );
 788                                         $group = "(?P<{$it}_{$name}>" . preg_quote( $syn, '/' ) . ')';
 789                                         if ( $this->baseRegex[$case] === '' ) {
 790                                                 $this->baseRegex[$case] = $group;
 791                                         } else {
 792                                                 $this->baseRegex[$case] .= '|' . $group;
 793                                         }
 794                                 }
 795                         }
 796                 }
 797                 return $this->baseRegex;
 798         }
 799
 800         /**
 801          * Get an unanchored regex that does not match parameters
 802          * @return array
 803          */
 804         function getRegex() {
 805                 if ( is_null( $this->regex ) ) {
 806                         $base = $this->getBaseRegex();
 807                         $this->regex = array( '', '' );
 808                         if ( $this->baseRegex[0] !== '' ) {
 809                                 $this->regex[0] = "/{$base[0]}/iuS";
 810                         }
 811                         if ( $this->baseRegex[1] !== '' ) {
 812                                 $this->regex[1] = "/{$base[1]}/S";
 813                         }
 814                 }
 815                 return $this->regex;
 816         }
 817
 818         /**
 819          * Get a regex for matching variables with parameters
 820          *
 821          * @return string
 822          */
 823         function getVariableRegex() {
 824                 return str_replace( "\\$1", "(.*?)", $this->getRegex() );
 825         }
 826
 827         /**
 828          * Get a regex anchored to the start of the string that does not match parameters
 829          *
 830          * @return array
 831          */
 832         function getRegexStart() {
 833                 $base = $this->getBaseRegex();
 834                 $newRegex = array( '', '' );
 835                 if ( $base[0] !== '' ) {
 836                         $newRegex[0] = "/^(?:{$base[0]})/iuS";
 837                 }
 838                 if ( $base[1] !== '' ) {
 839                         $newRegex[1] = "/^(?:{$base[1]})/S";
 840                 }
 841                 return $newRegex;
 842         }
 843
 844         /**
 845          * Get an anchored regex for matching variables with parameters
 846          *
 847          * @return array
 848          */
 849         function getVariableStartToEndRegex() {
 850                 $base = $this->getBaseRegex();
 851                 $newRegex = array( '', '' );
 852                 if ( $base[0] !== '' ) {
 853                         $newRegex[0] = str_replace( "\\$1", "(.*?)", "/^(?:{$base[0]})$/iuS" );
 854                 }
 855                 if ( $base[1] !== '' ) {
 856                         $newRegex[1] = str_replace( "\\$1", "(.*?)", "/^(?:{$base[1]})$/S" );
 857                 }
 858                 return $newRegex;
 859         }
 860
 861         /**
 862          * @since 1.20
 863          * @return array
 864          */
 865         public function getNames() {
 866                 return $this->names;
 867         }
 868
 869         /**
 870          * Parse a match array from preg_match
 871          * Returns array(magic word ID, parameter value)
 872          * If there is no parameter value, that element will be false.
 873          *
 874          * @param array $m
 875          *
 876          * @throws MWException
 877          * @return array
 878          */
 879         function parseMatch( $m ) {
 880                 reset( $m );
 881                 while ( list( $key, $value ) = each( $m ) ) {
 882                         if ( $key === 0 || $value === '' ) {
 883                                 continue;
 884                         }
 885                         $parts = explode( '_', $key, 2 );
 886                         if ( count( $parts ) != 2 ) {
 887                                 // This shouldn't happen
 888                                 // continue;
 889                                 throw new MWException( __METHOD__ . ': bad parameter name' );
 890                         }
 891                         list( /* $synIndex */, $magicName ) = $parts;
 892                         $paramValue = next( $m );
 893                         return array( $magicName, $paramValue );
 894                 }
 895                 // This shouldn't happen either
 896                 throw new MWException( __METHOD__ . ': parameter not found' );
 897         }
 898
 899         /**
 900          * Match some text, with parameter capture
 901          * Returns an array with the magic word name in the first element and the
 902          * parameter in the second element.
 903          * Both elements are false if there was no match.
 904          *
 905          * @param string $text
 906          *
 907          * @return array
 908          */
 909         public function matchVariableStartToEnd( $text ) {
 910                 $regexes = $this->getVariableStartToEndRegex();
 911                 foreach ( $regexes as $regex ) {
 912                         if ( $regex !== '' ) {
 913                                 $m = array();
 914                                 if ( preg_match( $regex, $text, $m ) ) {
 915                                         return $this->parseMatch( $m );
 916                                 }
 917                         }
 918                 }
 919                 return array( false, false );
 920         }
 921
 922         /**
 923          * Match some text, without parameter capture
 924          * Returns the magic word name, or false if there was no capture
 925          *
 926          * @param string $text
 927          *
 928          * @return string|bool False on failure
 929          */
 930         public function matchStartToEnd( $text ) {
 931                 $hash = $this->getHash();
 932                 if ( isset( $hash[1][$text] ) ) {
 933                         return $hash[1][$text];
 934                 }
 935                 global $wgContLang;
 936                 $lc = $wgContLang->lc( $text );
 937                 if ( isset( $hash[0][$lc] ) ) {
 938                         return $hash[0][$lc];
 939                 }
 940                 return false;
 941         }
 942
 943         /**
 944          * Returns an associative array, ID => param value, for all items that match
 945          * Removes the matched items from the input string (passed by reference)
 946          *
 947          * @param string $text
 948          *
 949          * @return array
 950          */
 951         public function matchAndRemove( &$text ) {
 952                 $found = array();
 953                 $regexes = $this->getRegex();
 954                 foreach ( $regexes as $regex ) {
 955                         if ( $regex === '' ) {
 956                                 continue;
 957                         }
 958                         preg_match_all( $regex, $text, $matches, PREG_SET_ORDER );
 959                         foreach ( $matches as $m ) {
 960                                 list( $name, $param ) = $this->parseMatch( $m );
 961                                 $found[$name] = $param;
 962                         }
 963                         $text = preg_replace( $regex, '', $text );
 964                 }
 965                 return $found;
 966         }
 967
 968         /**
 969          * Return the ID of the magic word at the start of $text, and remove
 970          * the prefix from $text.
 971          * Return false if no match found and $text is not modified.
 972          * Does not match parameters.
 973          *
 974          * @param string $text
 975          *
 976          * @return int|bool False on failure
 977          */
 978         public function matchStartAndRemove( &$text ) {
 979                 $regexes = $this->getRegexStart();
 980                 foreach ( $regexes as $regex ) {
 981                         if ( $regex === '' ) {
 982                                 continue;
 983                         }
 984                         if ( preg_match( $regex, $text, $m ) ) {
 985                                 list( $id, ) = $this->parseMatch( $m );
 986                                 if ( strlen( $m[0] ) >= strlen( $text ) ) {
 987                                         $text = '';
 988                                 } else {
 989                                         $text = substr( $text, strlen( $m[0] ) );
 990                                 }
 991                                 return $id;
 992                         }
 993                 }
 994                 return false;
 995         }
 996 }