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 );
  85 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  86         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  87
  88 for ( $i = 0; $i < count( $magicWords ); ++$i )
  89         define( $magicWords[$i], $i );
  90
  91 $wgVariableIDs = array(
  92         MAG_CURRENTMONTH,
  93         MAG_CURRENTMONTHNAME,
  94         MAG_CURRENTMONTHNAMEGEN,
  95         MAG_CURRENTMONTHABBREV,
  96         MAG_CURRENTDAY,
  97         MAG_CURRENTDAY2,
  98         MAG_CURRENTDAYNAME,
  99         MAG_CURRENTYEAR,
 100         MAG_CURRENTTIME,
 101         MAG_NUMBEROFARTICLES,
 102         MAG_NUMBEROFFILES,
 103         MAG_SITENAME,
 104         MAG_SERVER,
 105         MAG_SERVERNAME,
 106         MAG_SCRIPTPATH,
 107         MAG_PAGENAME,
 108         MAG_PAGENAMEE,
 109         MAG_FULLPAGENAME,
 110         MAG_FULLPAGENAMEE,
 111         MAG_NAMESPACE,
 112         MAG_NAMESPACEE,
 113         MAG_CURRENTWEEK,
 114         MAG_CURRENTDOW,
 115         MAG_REVISIONID,
 116         MAG_SUBPAGENAME,
 117         MAG_SUBPAGENAMEE,
 118         MAG_DISPLAYTITLE,
 119         MAG_TALKSPACE,
 120         MAG_TALKSPACEE,
 121         MAG_SUBJECTSPACE,
 122         MAG_SUBJECTSPACEE,
 123         MAG_TALKPAGENAME,
 124         MAG_TALKPAGENAMEE,
 125         MAG_SUBJECTPAGENAME,
 126         MAG_SUBJECTPAGENAMEE,
 127 );
 128 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 129         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 130
 131 /**
 132  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 133  * Usage:
 134  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 135  *
 136  * Possible future improvements:
 137  *   * Simultaneous searching for a number of magic words
 138  *   * $wgMagicWords in shared memory
 139  *
 140  * Please avoid reading the data out of one of these objects and then writing
 141  * special case code. If possible, add another match()-like function here.
 142  *
 143  * @package MediaWiki
 144  */
 145 class MagicWord {
 146         /**#@+
 147          * @access private
 148          */
 149         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 150         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 151         var $mModified;
 152         /**#@-*/
 153
 154         function MagicWord($id = 0, $syn = '', $cs = false) {
 155                 $this->mId = $id;
 156                 $this->mSynonyms = (array)$syn;
 157                 $this->mCaseSensitive = $cs;
 158                 $this->mRegex = '';
 159                 $this->mRegexStart = '';
 160                 $this->mVariableRegex = '';
 161                 $this->mVariableStartToEndRegex = '';
 162                 $this->mModified = false;
 163         }
 164
 165         /**
 166          * Factory: creates an object representing an ID
 167          * @static
 168          */
 169         function &get( $id ) {
 170                 global $wgMagicWords;
 171
 172                 if ( !is_array( $wgMagicWords ) ) {
 173                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 174                 }
 175                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 176                         $mw = new MagicWord();
 177                         $mw->load( $id );
 178                         $wgMagicWords[$id] = $mw;
 179                 }
 180                 return $wgMagicWords[$id];
 181         }
 182
 183         # Initialises this object with an ID
 184         function load( $id ) {
 185                 global $wgContLang;
 186                 $this->mId = $id;
 187                 $wgContLang->getMagic( $this );
 188         }
 189
 190         /**
 191          * Preliminary initialisation
 192          * @access private
 193          */
 194         function initRegex() {
 195                 #$variableClass = Title::legalChars();
 196                 # This was used for matching "$1" variables, but different uses of the feature will have
 197                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 198                 # not here. - IMSoP
 199
 200                 $escSyn = array();
 201                 foreach ( $this->mSynonyms as $synonym )
 202                         // In case a magic word contains /, like that's going to happen;)
 203                         $escSyn[] = preg_quote( $synonym, '/' );
 204                 $this->mBaseRegex = implode( '|', $escSyn );
 205
 206                 $case = $this->mCaseSensitive ? '' : 'i';
 207                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 208                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 209                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 210                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 211                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 212         }
 213
 214         /**
 215          * Gets a regex representing matching the word
 216          */
 217         function getRegex() {
 218                 if ($this->mRegex == '' ) {
 219                         $this->initRegex();
 220                 }
 221                 return $this->mRegex;
 222         }
 223
 224         /**
 225          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 226          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 227          * the complete expression
 228          */
 229         function getRegexCase() {
 230                 if ( $this->mRegex === '' )
 231                         $this->initRegex();
 232
 233                 return $this->mCaseSensitive ? '' : 'i';
 234         }
 235
 236         /**
 237          * Gets a regex matching the word, if it is at the string start
 238          */
 239         function getRegexStart() {
 240                 if ($this->mRegex == '' ) {
 241                         $this->initRegex();
 242                 }
 243                 return $this->mRegexStart;
 244         }
 245
 246         /**
 247          * regex without the slashes and what not
 248          */
 249         function getBaseRegex() {
 250                 if ($this->mRegex == '') {
 251                         $this->initRegex();
 252                 }
 253                 return $this->mBaseRegex;
 254         }
 255
 256         /**
 257          * Returns true if the text contains the word
 258          * @return bool
 259          */
 260         function match( $text ) {
 261                 return preg_match( $this->getRegex(), $text );
 262         }
 263
 264         /**
 265          * Returns true if the text starts with the word
 266          * @return bool
 267          */
 268         function matchStart( $text ) {
 269                 return preg_match( $this->getRegexStart(), $text );
 270         }
 271
 272         /**
 273          * Returns NULL if there's no match, the value of $1 otherwise
 274          * The return code is the matched string, if there's no variable
 275          * part in the regex and the matched variable part ($1) if there
 276          * is one.
 277          */
 278         function matchVariableStartToEnd( $text ) {
 279                 $matches = array();
 280                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 281                 if ( $matchcount == 0 ) {
 282                         return NULL;
 283                 } elseif ( count($matches) == 1 ) {
 284                         return $matches[0];
 285                 } else {
 286                         # multiple matched parts (variable match); some will be empty because of synonyms
 287                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 288                         $matches = array_values(array_filter($matches));
 289                         return $matches[1];
 290                 }
 291         }
 292
 293
 294         /**
 295          * Returns true if the text matches the word, and alters the
 296          * input string, removing all instances of the word
 297          */
 298         function matchAndRemove( &$text ) {
 299                 global $wgMagicFound;
 300                 $wgMagicFound = false;
 301                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 302                 return $wgMagicFound;
 303         }
 304
 305         function matchStartAndRemove( &$text ) {
 306                 global $wgMagicFound;
 307                 $wgMagicFound = false;
 308                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 309                 return $wgMagicFound;
 310         }
 311
 312
 313         /**
 314          * Replaces the word with something else
 315          */
 316         function replace( $replacement, $subject ) {
 317                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject );
 318                 $this->mModified = !($res === $subject);
 319                 return $res;
 320         }
 321
 322         /**
 323          * Variable handling: {{SUBST:xxx}} style words
 324          * Calls back a function to determine what to replace xxx with
 325          * Input word must contain $1
 326          */
 327         function substituteCallback( $text, $callback ) {
 328                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 329                 $this->mModified = !($res === $text);
 330                 return $res;
 331         }
 332
 333         /**
 334          * Matches the word, where $1 is a wildcard
 335          */
 336         function getVariableRegex()     {
 337                 if ( $this->mVariableRegex == '' ) {
 338                         $this->initRegex();
 339                 }
 340                 return $this->mVariableRegex;
 341         }
 342
 343         /**
 344          * Matches the entire string, where $1 is a wildcard
 345          */
 346         function getVariableStartToEndRegex() {
 347                 if ( $this->mVariableStartToEndRegex == '' ) {
 348                         $this->initRegex();
 349                 }
 350                 return $this->mVariableStartToEndRegex;
 351         }
 352
 353         /**
 354          * Accesses the synonym list directly
 355          */
 356         function getSynonym( $i ) {
 357                 return $this->mSynonyms[$i];
 358         }
 359
 360         /**
 361          * Returns true if the last call to replace() or substituteCallback()
 362          * returned a modified text, otherwise false.
 363          */
 364         function getWasModified(){
 365                 return $this->mModified;
 366         }
 367
 368         /**
 369          * $magicarr is an associative array of (magic word ID => replacement)
 370          * This method uses the php feature to do several replacements at the same time,
 371          * thereby gaining some efficiency. The result is placed in the out variable
 372          * $result. The return value is true if something was replaced.
 373          * @static
 374          **/
 375         function replaceMultiple( $magicarr, $subject, &$result ){
 376                 $search = array();
 377                 $replace = array();
 378                 foreach( $magicarr as $id => $replacement ){
 379                         $mw = MagicWord::get( $id );
 380                         $search[] = $mw->getRegex();
 381                         $replace[] = $replacement;
 382                 }
 383
 384                 $result = preg_replace( $search, $replace, $subject );
 385                 return !($result === $subject);
 386         }
 387
 388         /**
 389          * Adds all the synonyms of this MagicWord to an array, to allow quick
 390          * lookup in a list of magic words
 391          */
 392         function addToArray( &$array, $value ) {
 393                 foreach ( $this->mSynonyms as $syn ) {
 394                         $array[$syn] = $value;
 395                 }
 396         }
 397 }
 398
 399 /**
 400  * Used in matchAndRemove()
 401  * @access private
 402  **/
 403 function pregRemoveAndRecord( $match ) {
 404         global $wgMagicFound;
 405         $wgMagicFound = true;
 406         return '';
 407 }
 408
 409 ?>