includes/MagicWord.php

   1 <?php
   2 /**
   3  * File for magic words
   4  * @package MediaWiki
   5  * @subpackage Parser
   6  */
   7
   8 /**
   9  * private
  10  */
  11 $wgMagicFound = false;
  12
  13 /** Actual keyword to be used is set in Language.php */
  14
  15 $magicWords = array(
  16         'MAG_REDIRECT',
  17         'MAG_NOTOC',
  18         'MAG_START',
  19         'MAG_CURRENTMONTH',
  20         'MAG_CURRENTMONTHNAME',
  21         'MAG_CURRENTMONTHNAMEGEN',
  22         'MAG_CURRENTMONTHABBREV',
  23         'MAG_CURRENTDAY',
  24         'MAG_CURRENTDAY2',
  25         'MAG_CURRENTDAYNAME',
  26         'MAG_CURRENTYEAR',
  27         'MAG_CURRENTTIME',
  28         'MAG_NUMBEROFARTICLES',
  29         'MAG_SUBST',
  30         'MAG_MSG',
  31         'MAG_MSGNW',
  32         'MAG_NOEDITSECTION',
  33         'MAG_END',
  34         'MAG_IMG_THUMBNAIL',
  35         'MAG_IMG_RIGHT',
  36         'MAG_IMG_LEFT',
  37         'MAG_IMG_NONE',
  38         'MAG_IMG_WIDTH',
  39         'MAG_IMG_CENTER',
  40         'MAG_INT',
  41         'MAG_FORCETOC',
  42         'MAG_SITENAME',
  43         'MAG_NS',
  44         'MAG_LOCALURL',
  45         'MAG_LOCALURLE',
  46         'MAG_SERVER',
  47         'MAG_IMG_FRAMED',
  48         'MAG_PAGENAME',
  49         'MAG_PAGENAMEE',
  50         'MAG_NAMESPACE',
  51         'MAG_NAMESPACEE',
  52         'MAG_TOC',
  53         'MAG_GRAMMAR',
  54         'MAG_NOTITLECONVERT',
  55         'MAG_NOCONTENTCONVERT',
  56         'MAG_CURRENTWEEK',
  57         'MAG_CURRENTDOW',
  58         'MAG_REVISIONID',
  59         'MAG_SCRIPTPATH',
  60         'MAG_SERVERNAME',
  61         'MAG_NUMBEROFFILES',
  62         'MAG_IMG_MANUALTHUMB',
  63         'MAG_PLURAL',
  64         'MAG_FULLURL',
  65         'MAG_FULLURLE',
  66         'MAG_LCFIRST',
  67         'MAG_UCFIRST',
  68         'MAG_LC',
  69         'MAG_UC',
  70         'MAG_FULLPAGENAME',
  71         'MAG_FULLPAGENAMEE',
  72         'MAG_RAW',
  73         'MAG_SUBPAGENAME',
  74         'MAG_SUBPAGENAMEE',
  75         'MAG_DISPLAYTITLE',
  76         'MAG_TALKSPACE',
  77         'MAG_TALKSPACEE',
  78         'MAG_SUBJECTSPACE',
  79         'MAG_SUBJECTSPACEE',
  80         'MAG_TALKPAGENAME',
  81         'MAG_TALKPAGENAMEE',
  82         'MAG_SUBJECTPAGENAME',
  83         'MAG_SUBJECTPAGENAMEE',
  84         'MAG_NUMBEROFUSERS',
  85         'MAG_RAWSUFFIX',
  86         'MAG_NEWSECTIONLINK',
  87         'MAG_NUMBEROFPAGES',
  88         'MAG_CURRENTVERSION',
  89         'MAG_BASEPAGENAME',
  90         'MAG_BASEPAGENAMEE',
  91         'MAG_URLENCODE',
  92         'MAG_CURRENTTIMESTAMP',
  93         'MAG_DIRECTIONMARK',
  94         'MAG_LANGUAGE',
  95         'MAG_CONTENTLANGUAGE',
  96         'MAG_PAGESINNAMESPACE',
  97         'MAG_NOGALLERY',
  98         'MAG_NUMBEROFADMINS',
  99         'MAG_FORMATNUM',
 100 );
 101 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 102         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
 103
 104 for ( $i = 0; $i < count( $magicWords ); ++$i )
 105         define( $magicWords[$i], $i );
 106
 107 $wgVariableIDs = array(
 108         MAG_CURRENTMONTH,
 109         MAG_CURRENTMONTHNAME,
 110         MAG_CURRENTMONTHNAMEGEN,
 111         MAG_CURRENTMONTHABBREV,
 112         MAG_CURRENTDAY,
 113         MAG_CURRENTDAY2,
 114         MAG_CURRENTDAYNAME,
 115         MAG_CURRENTYEAR,
 116         MAG_CURRENTTIME,
 117         MAG_NUMBEROFARTICLES,
 118         MAG_NUMBEROFFILES,
 119         MAG_SITENAME,
 120         MAG_SERVER,
 121         MAG_SERVERNAME,
 122         MAG_SCRIPTPATH,
 123         MAG_PAGENAME,
 124         MAG_PAGENAMEE,
 125         MAG_FULLPAGENAME,
 126         MAG_FULLPAGENAMEE,
 127         MAG_NAMESPACE,
 128         MAG_NAMESPACEE,
 129         MAG_CURRENTWEEK,
 130         MAG_CURRENTDOW,
 131         MAG_REVISIONID,
 132         MAG_SUBPAGENAME,
 133         MAG_SUBPAGENAMEE,
 134         MAG_DISPLAYTITLE,
 135         MAG_TALKSPACE,
 136         MAG_TALKSPACEE,
 137         MAG_SUBJECTSPACE,
 138         MAG_SUBJECTSPACEE,
 139         MAG_TALKPAGENAME,
 140         MAG_TALKPAGENAMEE,
 141         MAG_SUBJECTPAGENAME,
 142         MAG_SUBJECTPAGENAMEE,
 143         MAG_NUMBEROFUSERS,
 144         MAG_RAWSUFFIX,
 145         MAG_NEWSECTIONLINK,
 146         MAG_NUMBEROFPAGES,
 147         MAG_CURRENTVERSION,
 148         MAG_BASEPAGENAME,
 149         MAG_BASEPAGENAMEE,
 150         MAG_URLENCODE,
 151         MAG_CURRENTTIMESTAMP,
 152         MAG_DIRECTIONMARK,
 153         MAG_LANGUAGE,
 154         MAG_CONTENTLANGUAGE,
 155         MAG_PAGESINNAMESPACE,
 156         MAG_NUMBEROFADMINS,
 157 );
 158 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 159         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 160
 161 /**
 162  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 163  * Usage:
 164  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 165  *
 166  * Possible future improvements:
 167  *   * Simultaneous searching for a number of magic words
 168  *   * $wgMagicWords in shared memory
 169  *
 170  * Please avoid reading the data out of one of these objects and then writing
 171  * special case code. If possible, add another match()-like function here.
 172  *
 173  * @package MediaWiki
 174  */
 175 class MagicWord {
 176         /**#@+
 177          * @private
 178          */
 179         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 180         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 181         var $mModified;
 182         /**#@-*/
 183
 184         function MagicWord($id = 0, $syn = '', $cs = false) {
 185                 $this->mId = $id;
 186                 $this->mSynonyms = (array)$syn;
 187                 $this->mCaseSensitive = $cs;
 188                 $this->mRegex = '';
 189                 $this->mRegexStart = '';
 190                 $this->mVariableRegex = '';
 191                 $this->mVariableStartToEndRegex = '';
 192                 $this->mModified = false;
 193         }
 194
 195         /**
 196          * Factory: creates an object representing an ID
 197          * @static
 198          */
 199         static function &get( $id ) {
 200                 global $wgMagicWords;
 201
 202                 if ( !is_array( $wgMagicWords ) ) {
 203                         throw new MWException( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 204                 }
 205                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 206                         $mw = new MagicWord();
 207                         $mw->load( $id );
 208                         $wgMagicWords[$id] = $mw;
 209                 }
 210                 return $wgMagicWords[$id];
 211         }
 212
 213         # Initialises this object with an ID
 214         function load( $id ) {
 215                 global $wgContLang;
 216                 $this->mId = $id;
 217                 $wgContLang->getMagic( $this );
 218         }
 219
 220         /**
 221          * Preliminary initialisation
 222          * @private
 223          */
 224         function initRegex() {
 225                 #$variableClass = Title::legalChars();
 226                 # This was used for matching "$1" variables, but different uses of the feature will have
 227                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 228                 # not here. - IMSoP
 229
 230                 $escSyn = array();
 231                 foreach ( $this->mSynonyms as $synonym )
 232                         // In case a magic word contains /, like that's going to happen;)
 233                         $escSyn[] = preg_quote( $synonym, '/' );
 234                 $this->mBaseRegex = implode( '|', $escSyn );
 235
 236                 $case = $this->mCaseSensitive ? '' : 'i';
 237                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 238                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 239                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 240                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 241                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 242         }
 243
 244         /**
 245          * Gets a regex representing matching the word
 246          */
 247         function getRegex() {
 248                 if ($this->mRegex == '' ) {
 249                         $this->initRegex();
 250                 }
 251                 return $this->mRegex;
 252         }
 253
 254         /**
 255          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 256          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 257          * the complete expression
 258          */
 259         function getRegexCase() {
 260                 if ( $this->mRegex === '' )
 261                         $this->initRegex();
 262
 263                 return $this->mCaseSensitive ? '' : 'i';
 264         }
 265
 266         /**
 267          * Gets a regex matching the word, if it is at the string start
 268          */
 269         function getRegexStart() {
 270                 if ($this->mRegex == '' ) {
 271                         $this->initRegex();
 272                 }
 273                 return $this->mRegexStart;
 274         }
 275
 276         /**
 277          * regex without the slashes and what not
 278          */
 279         function getBaseRegex() {
 280                 if ($this->mRegex == '') {
 281                         $this->initRegex();
 282                 }
 283                 return $this->mBaseRegex;
 284         }
 285
 286         /**
 287          * Returns true if the text contains the word
 288          * @return bool
 289          */
 290         function match( $text ) {
 291                 return preg_match( $this->getRegex(), $text );
 292         }
 293
 294         /**
 295          * Returns true if the text starts with the word
 296          * @return bool
 297          */
 298         function matchStart( $text ) {
 299                 return preg_match( $this->getRegexStart(), $text );
 300         }
 301
 302         /**
 303          * Returns NULL if there's no match, the value of $1 otherwise
 304          * The return code is the matched string, if there's no variable
 305          * part in the regex and the matched variable part ($1) if there
 306          * is one.
 307          */
 308         function matchVariableStartToEnd( $text ) {
 309                 $matches = array();
 310                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 311                 if ( $matchcount == 0 ) {
 312                         return NULL;
 313                 } elseif ( count($matches) == 1 ) {
 314                         return $matches[0];
 315                 } else {
 316                         # multiple matched parts (variable match); some will be empty because of
 317                         # synonyms. The variable will be the second non-empty one so remove any
 318                         # blank elements and re-sort the indices.
 319                         $matches = array_values(array_filter($matches));
 320                         return $matches[1];
 321                 }
 322         }
 323
 324
 325         /**
 326          * Returns true if the text matches the word, and alters the
 327          * input string, removing all instances of the word
 328          */
 329         function matchAndRemove( &$text ) {
 330                 global $wgMagicFound;
 331                 $wgMagicFound = false;
 332                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 333                 return $wgMagicFound;
 334         }
 335
 336         function matchStartAndRemove( &$text ) {
 337                 global $wgMagicFound;
 338                 $wgMagicFound = false;
 339                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 340                 return $wgMagicFound;
 341         }
 342
 343
 344         /**
 345          * Replaces the word with something else
 346          */
 347         function replace( $replacement, $subject, $limit=-1 ) {
 348                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject, $limit );
 349                 $this->mModified = !($res === $subject);
 350                 return $res;
 351         }
 352
 353         /**
 354          * Variable handling: {{SUBST:xxx}} style words
 355          * Calls back a function to determine what to replace xxx with
 356          * Input word must contain $1
 357          */
 358         function substituteCallback( $text, $callback ) {
 359                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 360                 $this->mModified = !($res === $text);
 361                 return $res;
 362         }
 363
 364         /**
 365          * Matches the word, where $1 is a wildcard
 366          */
 367         function getVariableRegex()     {
 368                 if ( $this->mVariableRegex == '' ) {
 369                         $this->initRegex();
 370                 }
 371                 return $this->mVariableRegex;
 372         }
 373
 374         /**
 375          * Matches the entire string, where $1 is a wildcard
 376          */
 377         function getVariableStartToEndRegex() {
 378                 if ( $this->mVariableStartToEndRegex == '' ) {
 379                         $this->initRegex();
 380                 }
 381                 return $this->mVariableStartToEndRegex;
 382         }
 383
 384         /**
 385          * Accesses the synonym list directly
 386          */
 387         function getSynonym( $i ) {
 388                 return $this->mSynonyms[$i];
 389         }
 390
 391         function getSynonyms() {
 392                 return $this->mSynonyms;
 393         }
 394
 395         /**
 396          * Returns true if the last call to replace() or substituteCallback()
 397          * returned a modified text, otherwise false.
 398          */
 399         function getWasModified(){
 400                 return $this->mModified;
 401         }
 402
 403         /**
 404          * $magicarr is an associative array of (magic word ID => replacement)
 405          * This method uses the php feature to do several replacements at the same time,
 406          * thereby gaining some efficiency. The result is placed in the out variable
 407          * $result. The return value is true if something was replaced.
 408          * @static
 409          **/
 410         function replaceMultiple( $magicarr, $subject, &$result ){
 411                 $search = array();
 412                 $replace = array();
 413                 foreach( $magicarr as $id => $replacement ){
 414                         $mw = MagicWord::get( $id );
 415                         $search[] = $mw->getRegex();
 416                         $replace[] = $replacement;
 417                 }
 418
 419                 $result = preg_replace( $search, $replace, $subject );
 420                 return !($result === $subject);
 421         }
 422
 423         /**
 424          * Adds all the synonyms of this MagicWord to an array, to allow quick
 425          * lookup in a list of magic words
 426          */
 427         function addToArray( &$array, $value ) {
 428                 foreach ( $this->mSynonyms as $syn ) {
 429                         $array[$syn] = $value;
 430                 }
 431         }
 432
 433         function isCaseSensitive() {
 434                 return $this->mCaseSensitive;
 435         }
 436 }
 437
 438 /**
 439  * Used in matchAndRemove()
 440  * @private
 441  **/
 442 function pregRemoveAndRecord( $match ) {
 443         global $wgMagicFound;
 444         $wgMagicFound = true;
 445         return '';
 446 }
 447
 448 ?>