includes/MagicWord.php

   1 <?php
   2 /**
   3  * File for magic words
   4  * @package MediaWiki
   5  * @subpackage Parser
   6  */
   7
   8 /**
   9  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  10  * Usage:
  11  *     if (MagicWord::get( 'redirect' )->match( $text ) )
  12  *
  13  * Possible future improvements:
  14  *   * Simultaneous searching for a number of magic words
  15  *   * MagicWord::$mObjects in shared memory
  16  *
  17  * Please avoid reading the data out of one of these objects and then writing
  18  * special case code. If possible, add another match()-like function here.
  19  *
  20  * To add magic words in an extension, use the LanguageGetMagic hook. For
  21  * magic words which are also Parser variables, add a MagicWordwgVariableIDs
  22  * hook. Use string keys.
  23  *
  24  * @package MediaWiki
  25  */
  26 class MagicWord {
  27         /**#@+
  28          * @private
  29          */
  30         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  31         var $mRegexStart, $mBaseRegex, $mVariableRegex;
  32         var $mModified, $mFound;
  33
  34         static public $mVariableIDsInitialised = false;
  35         static public $mVariableIDs = array(
  36                 'currentmonth',
  37                 'currentmonthname',
  38                 'currentmonthnamegen',
  39                 'currentmonthabbrev',
  40                 'currentday',
  41                 'currentday2',
  42                 'currentdayname',
  43                 'currentyear',
  44                 'currenttime',
  45                 'currenthour',
  46                 'localmonth',
  47                 'localmonthname',
  48                 'localmonthnamegen',
  49                 'localmonthabbrev',
  50                 'localday',
  51                 'localday2',
  52                 'localdayname',
  53                 'localyear',
  54                 'localtime',
  55                 'localhour',
  56                 'numberofarticles',
  57                 'numberoffiles',
  58                 'sitename',
  59                 'server',
  60                 'servername',
  61                 'scriptpath',
  62                 'pagename',
  63                 'pagenamee',
  64                 'fullpagename',
  65                 'fullpagenamee',
  66                 'namespace',
  67                 'namespacee',
  68                 'currentweek',
  69                 'currentdow',
  70                 'localweek',
  71                 'localdow',
  72                 'revisionid',
  73                 'subpagename',
  74                 'subpagenamee',
  75                 'displaytitle',
  76                 'talkspace',
  77                 'talkspacee',
  78                 'subjectspace',
  79                 'subjectspacee',
  80                 'talkpagename',
  81                 'talkpagenamee',
  82                 'subjectpagename',
  83                 'subjectpagenamee',
  84                 'numberofusers',
  85                 'rawsuffix',
  86                 'newsectionlink',
  87                 'numberofpages',
  88                 'currentversion',
  89                 'basepagename',
  90                 'basepagenamee',
  91                 'urlencode',
  92                 'currenttimestamp',
  93                 'localtimestamp',
  94                 'directionmark',
  95                 'language',
  96                 'contentlanguage',
  97                 'pagesinnamespace',
  98                 'numberofadmins',
  99         );
 100
 101         static public $mObjects = array();
 102
 103         /**#@-*/
 104
 105         function MagicWord($id = 0, $syn = '', $cs = false) {
 106                 $this->mId = $id;
 107                 $this->mSynonyms = (array)$syn;
 108                 $this->mCaseSensitive = $cs;
 109                 $this->mRegex = '';
 110                 $this->mRegexStart = '';
 111                 $this->mVariableRegex = '';
 112                 $this->mVariableStartToEndRegex = '';
 113                 $this->mModified = false;
 114         }
 115
 116         /**
 117          * Factory: creates an object representing an ID
 118          * @static
 119          */
 120         static function &get( $id ) {
 121                 if (!array_key_exists( $id, self::$mObjects ) ) {
 122                         $mw = new MagicWord();
 123                         $mw->load( $id );
 124                         self::$mObjects[$id] = $mw;
 125                 }
 126                 return self::$mObjects[$id];
 127         }
 128
 129         /**
 130          * Get an array of parser variable IDs
 131          */
 132         static function getVariableIDs() {
 133                 if ( !self::$mVariableIDsInitialised ) {
 134                         # Deprecated constant definition hook, available for extensions that need it
 135                         $magicWords = array();
 136                         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
 137                         foreach ( $magicWords as $word ) {
 138                                 define( $word, $word );
 139                         }
 140
 141                         # Get variable IDs
 142                         wfRunHooks( 'MagicWordwgVariableIDs', array( &self::$mVariableIDs ) );
 143                         self::$mVariableIDsInitialised = true;
 144                 }
 145                 return self::$mVariableIDs;
 146         }
 147
 148         # Initialises this object with an ID
 149         function load( $id ) {
 150                 global $wgContLang;
 151                 $this->mId = $id;
 152                 $wgContLang->getMagic( $this );
 153         }
 154
 155         /**
 156          * Preliminary initialisation
 157          * @private
 158          */
 159         function initRegex() {
 160                 #$variableClass = Title::legalChars();
 161                 # This was used for matching "$1" variables, but different uses of the feature will have
 162                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 163                 # not here. - IMSoP
 164
 165                 $escSyn = array();
 166                 foreach ( $this->mSynonyms as $synonym )
 167                         // In case a magic word contains /, like that's going to happen;)
 168                         $escSyn[] = preg_quote( $synonym, '/' );
 169                 $this->mBaseRegex = implode( '|', $escSyn );
 170
 171                 $case = $this->mCaseSensitive ? '' : 'i';
 172                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 173                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 174                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 175                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 176                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 177         }
 178
 179         /**
 180          * Gets a regex representing matching the word
 181          */
 182         function getRegex() {
 183                 if ($this->mRegex == '' ) {
 184                         $this->initRegex();
 185                 }
 186                 return $this->mRegex;
 187         }
 188
 189         /**
 190          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 191          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 192          * the complete expression
 193          */
 194         function getRegexCase() {
 195                 if ( $this->mRegex === '' )
 196                         $this->initRegex();
 197
 198                 return $this->mCaseSensitive ? '' : 'i';
 199         }
 200
 201         /**
 202          * Gets a regex matching the word, if it is at the string start
 203          */
 204         function getRegexStart() {
 205                 if ($this->mRegex == '' ) {
 206                         $this->initRegex();
 207                 }
 208                 return $this->mRegexStart;
 209         }
 210
 211         /**
 212          * regex without the slashes and what not
 213          */
 214         function getBaseRegex() {
 215                 if ($this->mRegex == '') {
 216                         $this->initRegex();
 217                 }
 218                 return $this->mBaseRegex;
 219         }
 220
 221         /**
 222          * Returns true if the text contains the word
 223          * @return bool
 224          */
 225         function match( $text ) {
 226                 return preg_match( $this->getRegex(), $text );
 227         }
 228
 229         /**
 230          * Returns true if the text starts with the word
 231          * @return bool
 232          */
 233         function matchStart( $text ) {
 234                 return preg_match( $this->getRegexStart(), $text );
 235         }
 236
 237         /**
 238          * Returns NULL if there's no match, the value of $1 otherwise
 239          * The return code is the matched string, if there's no variable
 240          * part in the regex and the matched variable part ($1) if there
 241          * is one.
 242          */
 243         function matchVariableStartToEnd( $text ) {
 244                 $matches = array();
 245                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 246                 if ( $matchcount == 0 ) {
 247                         return NULL;
 248                 } else {
 249                         # multiple matched parts (variable match); some will be empty because of
 250                         # synonyms. The variable will be the second non-empty one so remove any
 251                         # blank elements and re-sort the indices.
 252                         # See also bug 6526
 253
 254                         $matches = array_values(array_filter($matches));
 255
 256                         if ( count($matches) == 1 ) { return $matches[0]; }
 257                         else { return $matches[1]; }
 258                 }
 259         }
 260
 261
 262         /**
 263          * Returns true if the text matches the word, and alters the
 264          * input string, removing all instances of the word
 265          */
 266         function matchAndRemove( &$text ) {
 267                 $this->mFound = false;
 268                 $text = preg_replace_callback( $this->getRegex(), array( &$this, 'pregRemoveAndRecord' ), $text );
 269                 return $this->mFound;
 270         }
 271
 272         function matchStartAndRemove( &$text ) {
 273                 $this->mFound = false;
 274                 $text = preg_replace_callback( $this->getRegexStart(), array( &$this, 'pregRemoveAndRecord' ), $text );
 275                 return $this->mFound;
 276         }
 277
 278         /**
 279          * Used in matchAndRemove()
 280          * @private
 281          **/
 282         function pregRemoveAndRecord( $match ) {
 283                 $this->mFound = true;
 284                 return '';
 285         }
 286
 287         /**
 288          * Replaces the word with something else
 289          */
 290         function replace( $replacement, $subject, $limit=-1 ) {
 291                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject, $limit );
 292                 $this->mModified = !($res === $subject);
 293                 return $res;
 294         }
 295
 296         /**
 297          * Variable handling: {{SUBST:xxx}} style words
 298          * Calls back a function to determine what to replace xxx with
 299          * Input word must contain $1
 300          */
 301         function substituteCallback( $text, $callback ) {
 302                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 303                 $this->mModified = !($res === $text);
 304                 return $res;
 305         }
 306
 307         /**
 308          * Matches the word, where $1 is a wildcard
 309          */
 310         function getVariableRegex()     {
 311                 if ( $this->mVariableRegex == '' ) {
 312                         $this->initRegex();
 313                 }
 314                 return $this->mVariableRegex;
 315         }
 316
 317         /**
 318          * Matches the entire string, where $1 is a wildcard
 319          */
 320         function getVariableStartToEndRegex() {
 321                 if ( $this->mVariableStartToEndRegex == '' ) {
 322                         $this->initRegex();
 323                 }
 324                 return $this->mVariableStartToEndRegex;
 325         }
 326
 327         /**
 328          * Accesses the synonym list directly
 329          */
 330         function getSynonym( $i ) {
 331                 return $this->mSynonyms[$i];
 332         }
 333
 334         function getSynonyms() {
 335                 return $this->mSynonyms;
 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         function isCaseSensitive() {
 377                 return $this->mCaseSensitive;
 378         }
 379 }
 380
 381 ?>