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