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