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