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