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