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 $magicWords = array(
  16         'MAG_REDIRECT',
  17         'MAG_NOTOC',
  18         'MAG_START',
  19         'MAG_CURRENTMONTH',
  20         'MAG_CURRENTMONTHNAME',
  21         'MAG_CURRENTMONTHNAMEGEN',
  22         'MAG_CURRENTMONTHABBREV',
  23         'MAG_CURRENTDAY',
  24         'MAG_CURRENTDAY2',
  25         'MAG_CURRENTDAYNAME',
  26         'MAG_CURRENTYEAR',
  27         'MAG_CURRENTTIME',
  28         'MAG_NUMBEROFARTICLES',
  29         'MAG_SUBST',
  30         'MAG_MSG',
  31         'MAG_MSGNW',
  32         'MAG_NOEDITSECTION',
  33         'MAG_END',
  34         'MAG_IMG_THUMBNAIL',
  35         'MAG_IMG_RIGHT',
  36         'MAG_IMG_LEFT',
  37         'MAG_IMG_NONE',
  38         'MAG_IMG_WIDTH',
  39         'MAG_IMG_CENTER',
  40         'MAG_INT',
  41         'MAG_FORCETOC',
  42         'MAG_SITENAME',
  43         'MAG_NS',
  44         'MAG_LOCALURL',
  45         'MAG_LOCALURLE',
  46         'MAG_SERVER',
  47         'MAG_IMG_FRAMED',
  48         'MAG_PAGENAME',
  49         'MAG_PAGENAMEE',
  50         'MAG_NAMESPACE',
  51         'MAG_NAMESPACEE',
  52         'MAG_TOC',
  53         'MAG_GRAMMAR',
  54         'MAG_NOTITLECONVERT',
  55         'MAG_NOCONTENTCONVERT',
  56         'MAG_CURRENTWEEK',
  57         'MAG_CURRENTDOW',
  58         'MAG_REVISIONID',
  59         'MAG_SCRIPTPATH',
  60         'MAG_SERVERNAME',
  61         'MAG_NUMBEROFFILES',
  62         'MAG_IMG_MANUALTHUMB',
  63         'MAG_PLURAL',
  64         'MAG_FULLURL',
  65         'MAG_FULLURLE',
  66         'MAG_LCFIRST',
  67         'MAG_UCFIRST',
  68         'MAG_LC',
  69         'MAG_UC',
  70         'MAG_FULLPAGENAME',
  71         'MAG_FULLPAGENAMEE',
  72         'MAG_RAW',
  73 );
  74 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  75         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  76
  77 for ( $i = 0; $i < count( $magicWords ); ++$i )
  78         define( $magicWords[$i], $i );
  79
  80 $wgVariableIDs = array(
  81         MAG_CURRENTMONTH,
  82         MAG_CURRENTMONTHNAME,
  83         MAG_CURRENTMONTHNAMEGEN,
  84         MAG_CURRENTMONTHABBREV,
  85         MAG_CURRENTDAY,
  86         MAG_CURRENTDAY2,
  87         MAG_CURRENTDAYNAME,
  88         MAG_CURRENTYEAR,
  89         MAG_CURRENTTIME,
  90         MAG_NUMBEROFARTICLES,
  91         MAG_NUMBEROFFILES,
  92         MAG_SITENAME,
  93         MAG_SERVER,
  94         MAG_SERVERNAME,
  95         MAG_SCRIPTPATH,
  96         MAG_PAGENAME,
  97         MAG_PAGENAMEE,
  98         MAG_FULLPAGENAME,
  99         MAG_FULLPAGENAMEE,
 100         MAG_NAMESPACE,
 101         MAG_NAMESPACEE,
 102         MAG_CURRENTWEEK,
 103         MAG_CURRENTDOW,
 104         MAG_REVISIONID,
 105 );
 106 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 107         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 108
 109 /**
 110  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 111  * Usage:
 112  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 113  *
 114  * Possible future improvements:
 115  *   * Simultaneous searching for a number of magic words
 116  *   * $wgMagicWords in shared memory
 117  *
 118  * Please avoid reading the data out of one of these objects and then writing
 119  * special case code. If possible, add another match()-like function here.
 120  *
 121  * @package MediaWiki
 122  */
 123 class MagicWord {
 124         /**#@+
 125          * @access private
 126          */
 127         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 128         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 129         var $mModified;
 130         /**#@-*/
 131
 132         function MagicWord($id = 0, $syn = '', $cs = false) {
 133                 $this->mId = $id;
 134                 $this->mSynonyms = (array)$syn;
 135                 $this->mCaseSensitive = $cs;
 136                 $this->mRegex = '';
 137                 $this->mRegexStart = '';
 138                 $this->mVariableRegex = '';
 139                 $this->mVariableStartToEndRegex = '';
 140                 $this->mModified = false;
 141         }
 142
 143         /**
 144          * Factory: creates an object representing an ID
 145          * @static
 146          */
 147         function &get( $id ) {
 148                 global $wgMagicWords;
 149
 150                 if ( !is_array( $wgMagicWords ) ) {
 151                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 152                 }
 153                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 154                         $mw = new MagicWord();
 155                         $mw->load( $id );
 156                         $wgMagicWords[$id] = $mw;
 157                 }
 158                 return $wgMagicWords[$id];
 159         }
 160
 161         # Initialises this object with an ID
 162         function load( $id ) {
 163                 global $wgContLang;
 164                 $this->mId = $id;
 165                 $wgContLang->getMagic( $this );
 166         }
 167
 168         /**
 169          * Preliminary initialisation
 170          * @access private
 171          */
 172         function initRegex() {
 173                 #$variableClass = Title::legalChars();
 174                 # This was used for matching "$1" variables, but different uses of the feature will have
 175                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 176                 # not here. - IMSoP
 177
 178                 $escSyn = array();
 179                 foreach ( $this->mSynonyms as $synonym )
 180                         // In case a magic word contains /, like that's going to happen;)
 181                         $escSyn[] = preg_quote( $synonym, '/' );
 182                 $this->mBaseRegex = implode( '|', $escSyn );
 183
 184                 $case = $this->mCaseSensitive ? '' : 'i';
 185                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 186                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 187                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 188                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 189                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 190         }
 191
 192         /**
 193          * Gets a regex representing matching the word
 194          */
 195         function getRegex() {
 196                 if ($this->mRegex == '' ) {
 197                         $this->initRegex();
 198                 }
 199                 return $this->mRegex;
 200         }
 201
 202         /**
 203          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 204          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 205          * the complete expression
 206          */
 207         function getRegexCase() {
 208                 if ( $this->mRegex === '' )
 209                         $this->initRegex();
 210
 211                 return $this->mCaseSensitive ? '' : 'i';
 212         }
 213
 214         /**
 215          * Gets a regex matching the word, if it is at the string start
 216          */
 217         function getRegexStart() {
 218                 if ($this->mRegex == '' ) {
 219                         $this->initRegex();
 220                 }
 221                 return $this->mRegexStart;
 222         }
 223
 224         /**
 225          * regex without the slashes and what not
 226          */
 227         function getBaseRegex() {
 228                 if ($this->mRegex == '') {
 229                         $this->initRegex();
 230                 }
 231                 return $this->mBaseRegex;
 232         }
 233
 234         /**
 235          * Returns true if the text contains the word
 236          * @return bool
 237          */
 238         function match( $text ) {
 239                 return preg_match( $this->getRegex(), $text );
 240         }
 241
 242         /**
 243          * Returns true if the text starts with the word
 244          * @return bool
 245          */
 246         function matchStart( $text ) {
 247                 return preg_match( $this->getRegexStart(), $text );
 248         }
 249
 250         /**
 251          * Returns NULL if there's no match, the value of $1 otherwise
 252          * The return code is the matched string, if there's no variable
 253          * part in the regex and the matched variable part ($1) if there
 254          * is one.
 255          */
 256         function matchVariableStartToEnd( $text ) {
 257                 $matches = array();
 258                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 259                 if ( $matchcount == 0 ) {
 260                         return NULL;
 261                 } elseif ( count($matches) == 1 ) {
 262                         return $matches[0];
 263                 } else {
 264                         # multiple matched parts (variable match); some will be empty because of synonyms
 265                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 266                         $matches = array_values(array_filter($matches));
 267                         return $matches[1];
 268                 }
 269         }
 270
 271
 272         /**
 273          * Returns true if the text matches the word, and alters the
 274          * input string, removing all instances of the word
 275          */
 276         function matchAndRemove( &$text ) {
 277                 global $wgMagicFound;
 278                 $wgMagicFound = false;
 279                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 280                 return $wgMagicFound;
 281         }
 282
 283         function matchStartAndRemove( &$text ) {
 284                 global $wgMagicFound;
 285                 $wgMagicFound = false;
 286                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 287                 return $wgMagicFound;
 288         }
 289
 290
 291         /**
 292          * Replaces the word with something else
 293          */
 294         function replace( $replacement, $subject ) {
 295                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject );
 296                 $this->mModified = !($res === $subject);
 297                 return $res;
 298         }
 299
 300         /**
 301          * Variable handling: {{SUBST:xxx}} style words
 302          * Calls back a function to determine what to replace xxx with
 303          * Input word must contain $1
 304          */
 305         function substituteCallback( $text, $callback ) {
 306                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 307                 $this->mModified = !($res === $text);
 308                 return $res;
 309         }
 310
 311         /**
 312          * Matches the word, where $1 is a wildcard
 313          */
 314         function getVariableRegex()     {
 315                 if ( $this->mVariableRegex == '' ) {
 316                         $this->initRegex();
 317                 }
 318                 return $this->mVariableRegex;
 319         }
 320
 321         /**
 322          * Matches the entire string, where $1 is a wildcard
 323          */
 324         function getVariableStartToEndRegex() {
 325                 if ( $this->mVariableStartToEndRegex == '' ) {
 326                         $this->initRegex();
 327                 }
 328                 return $this->mVariableStartToEndRegex;
 329         }
 330
 331         /**
 332          * Accesses the synonym list directly
 333          */
 334         function getSynonym( $i ) {
 335                 return $this->mSynonyms[$i];
 336         }
 337
 338         /**
 339          * Returns true if the last call to replace() or substituteCallback()
 340          * returned a modified text, otherwise false.
 341          */
 342         function getWasModified(){
 343                 return $this->mModified;
 344         }
 345
 346         /**
 347          * $magicarr is an associative array of (magic word ID => replacement)
 348          * This method uses the php feature to do several replacements at the same time,
 349          * thereby gaining some efficiency. The result is placed in the out variable
 350          * $result. The return value is true if something was replaced.
 351          * @static
 352          **/
 353         function replaceMultiple( $magicarr, $subject, &$result ){
 354                 $search = array();
 355                 $replace = array();
 356                 foreach( $magicarr as $id => $replacement ){
 357                         $mw = MagicWord::get( $id );
 358                         $search[] = $mw->getRegex();
 359                         $replace[] = $replacement;
 360                 }
 361
 362                 $result = preg_replace( $search, $replace, $subject );
 363                 return !($result === $subject);
 364         }
 365
 366         /**
 367          * Adds all the synonyms of this MagicWord to an array, to allow quick
 368          * lookup in a list of magic words
 369          */
 370         function addToArray( &$array, $value ) {
 371                 foreach ( $this->mSynonyms as $syn ) {
 372                         $array[$syn] = $value;
 373                 }
 374         }
 375 }
 376
 377 /**
 378  * Used in matchAndRemove()
 379  * @access private
 380  **/
 381 function pregRemoveAndRecord( $match ) {
 382         global $wgMagicFound;
 383         $wgMagicFound = true;
 384         return '';
 385 }
 386
 387 ?>