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