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