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