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