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                 'revisionday',
  74                 'revisionday2',
  75                 'revisionmonth',
  76                 'revisionyear',
  77                 'revisiontimestamp',
  78                 'subpagename',
  79                 'subpagenamee',
  80                 'displaytitle',
  81                 'talkspace',
  82                 'talkspacee',
  83                 'subjectspace',
  84                 'subjectspacee',
  85                 'talkpagename',
  86                 'talkpagenamee',
  87                 'subjectpagename',
  88                 'subjectpagenamee',
  89                 'numberofusers',
  90                 'rawsuffix',
  91                 'newsectionlink',
  92                 'numberofpages',
  93                 'currentversion',
  94                 'basepagename',
  95                 'basepagenamee',
  96                 'urlencode',
  97                 'currenttimestamp',
  98                 'localtimestamp',
  99                 'directionmark',
 100                 'language',
 101                 'contentlanguage',
 102                 'pagesinnamespace',
 103                 'numberofadmins',
 104                 'defaultsort',
 105         );
 106
 107         static public $mObjects = array();
 108
 109         /**#@-*/
 110
 111         function MagicWord($id = 0, $syn = '', $cs = false) {
 112                 $this->mId = $id;
 113                 $this->mSynonyms = (array)$syn;
 114                 $this->mCaseSensitive = $cs;
 115                 $this->mRegex = '';
 116                 $this->mRegexStart = '';
 117                 $this->mVariableRegex = '';
 118                 $this->mVariableStartToEndRegex = '';
 119                 $this->mModified = false;
 120         }
 121
 122         /**
 123          * Factory: creates an object representing an ID
 124          * @static
 125          */
 126         static function &get( $id ) {
 127                 if (!array_key_exists( $id, self::$mObjects ) ) {
 128                         $mw = new MagicWord();
 129                         $mw->load( $id );
 130                         self::$mObjects[$id] = $mw;
 131                 }
 132                 return self::$mObjects[$id];
 133         }
 134
 135         /**
 136          * Get an array of parser variable IDs
 137          */
 138         static function getVariableIDs() {
 139                 if ( !self::$mVariableIDsInitialised ) {
 140                         # Deprecated constant definition hook, available for extensions that need it
 141                         $magicWords = array();
 142                         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
 143                         foreach ( $magicWords as $word ) {
 144                                 define( $word, $word );
 145                         }
 146
 147                         # Get variable IDs
 148                         wfRunHooks( 'MagicWordwgVariableIDs', array( &self::$mVariableIDs ) );
 149                         self::$mVariableIDsInitialised = true;
 150                 }
 151                 return self::$mVariableIDs;
 152         }
 153
 154         # Initialises this object with an ID
 155         function load( $id ) {
 156                 global $wgContLang;
 157                 $this->mId = $id;
 158                 $wgContLang->getMagic( $this );
 159                 if ( !$this->mSynonyms ) {
 160                         $this->mSynonyms = array( 'dkjsagfjsgashfajsh' );
 161                         #throw new MWException( "Error: invalid magic word '$id'" );
 162                         wfDebugLog( 'exception', "Error: invalid magic word '$id'\n" );
 163                 }
 164         }
 165
 166         /**
 167          * Preliminary initialisation
 168          * @private
 169          */
 170         function initRegex() {
 171                 #$variableClass = Title::legalChars();
 172                 # This was used for matching "$1" variables, but different uses of the feature will have
 173                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 174                 # not here. - IMSoP
 175
 176                 $escSyn = array();
 177                 foreach ( $this->mSynonyms as $synonym )
 178                         // In case a magic word contains /, like that's going to happen;)
 179                         $escSyn[] = preg_quote( $synonym, '/' );
 180                 $this->mBaseRegex = implode( '|', $escSyn );
 181
 182                 $case = $this->mCaseSensitive ? '' : 'iu';
 183                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 184                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 185                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 186                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 187                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 188         }
 189
 190         /**
 191          * Gets a regex representing matching the word
 192          */
 193         function getRegex() {
 194                 if ($this->mRegex == '' ) {
 195                         $this->initRegex();
 196                 }
 197                 return $this->mRegex;
 198         }
 199
 200         /**
 201          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 202          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 203          * the complete expression
 204          */
 205         function getRegexCase() {
 206                 if ( $this->mRegex === '' )
 207                         $this->initRegex();
 208
 209                 return $this->mCaseSensitive ? '' : 'iu';
 210         }
 211
 212         /**
 213          * Gets a regex matching the word, if it is at the string start
 214          */
 215         function getRegexStart() {
 216                 if ($this->mRegex == '' ) {
 217                         $this->initRegex();
 218                 }
 219                 return $this->mRegexStart;
 220         }
 221
 222         /**
 223          * regex without the slashes and what not
 224          */
 225         function getBaseRegex() {
 226                 if ($this->mRegex == '') {
 227                         $this->initRegex();
 228                 }
 229                 return $this->mBaseRegex;
 230         }
 231
 232         /**
 233          * Returns true if the text contains the word
 234          * @return bool
 235          */
 236         function match( $text ) {
 237                 return preg_match( $this->getRegex(), $text );
 238         }
 239
 240         /**
 241          * Returns true if the text starts with the word
 242          * @return bool
 243          */
 244         function matchStart( $text ) {
 245                 return preg_match( $this->getRegexStart(), $text );
 246         }
 247
 248         /**
 249          * Returns NULL if there's no match, the value of $1 otherwise
 250          * The return code is the matched string, if there's no variable
 251          * part in the regex and the matched variable part ($1) if there
 252          * is one.
 253          */
 254         function matchVariableStartToEnd( $text ) {
 255                 $matches = array();
 256                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 257                 if ( $matchcount == 0 ) {
 258                         return NULL;
 259                 } else {
 260                         # multiple matched parts (variable match); some will be empty because of
 261                         # synonyms. The variable will be the second non-empty one so remove any
 262                         # blank elements and re-sort the indices.
 263                         # See also bug 6526
 264
 265                         $matches = array_values(array_filter($matches));
 266
 267                         if ( count($matches) == 1 ) { return $matches[0]; }
 268                         else { return $matches[1]; }
 269                 }
 270         }
 271
 272
 273         /**
 274          * Returns true if the text matches the word, and alters the
 275          * input string, removing all instances of the word
 276          */
 277         function matchAndRemove( &$text ) {
 278                 $this->mFound = false;
 279                 $text = preg_replace_callback( $this->getRegex(), array( &$this, 'pregRemoveAndRecord' ), $text );
 280                 return $this->mFound;
 281         }
 282
 283         function matchStartAndRemove( &$text ) {
 284                 $this->mFound = false;
 285                 $text = preg_replace_callback( $this->getRegexStart(), array( &$this, 'pregRemoveAndRecord' ), $text );
 286                 return $this->mFound;
 287         }
 288
 289         /**
 290          * Used in matchAndRemove()
 291          * @private
 292          **/
 293         function pregRemoveAndRecord( ) {
 294                 $this->mFound = true;
 295                 return '';
 296         }
 297
 298         /**
 299          * Replaces the word with something else
 300          */
 301         function replace( $replacement, $subject, $limit=-1 ) {
 302                 $res = preg_replace( $this->getRegex(), StringUtils::escapeRegexReplacement( $replacement ), $subject, $limit );
 303                 $this->mModified = !($res === $subject);
 304                 return $res;
 305         }
 306
 307         /**
 308          * Variable handling: {{SUBST:xxx}} style words
 309          * Calls back a function to determine what to replace xxx with
 310          * Input word must contain $1
 311          */
 312         function substituteCallback( $text, $callback ) {
 313                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 314                 $this->mModified = !($res === $text);
 315                 return $res;
 316         }
 317
 318         /**
 319          * Matches the word, where $1 is a wildcard
 320          */
 321         function getVariableRegex()     {
 322                 if ( $this->mVariableRegex == '' ) {
 323                         $this->initRegex();
 324                 }
 325                 return $this->mVariableRegex;
 326         }
 327
 328         /**
 329          * Matches the entire string, where $1 is a wildcard
 330          */
 331         function getVariableStartToEndRegex() {
 332                 if ( $this->mVariableStartToEndRegex == '' ) {
 333                         $this->initRegex();
 334                 }
 335                 return $this->mVariableStartToEndRegex;
 336         }
 337
 338         /**
 339          * Accesses the synonym list directly
 340          */
 341         function getSynonym( $i ) {
 342                 return $this->mSynonyms[$i];
 343         }
 344
 345         function getSynonyms() {
 346                 return $this->mSynonyms;
 347         }
 348
 349         /**
 350          * Returns true if the last call to replace() or substituteCallback()
 351          * returned a modified text, otherwise false.
 352          */
 353         function getWasModified(){
 354                 return $this->mModified;
 355         }
 356
 357         /**
 358          * $magicarr is an associative array of (magic word ID => replacement)
 359          * This method uses the php feature to do several replacements at the same time,
 360          * thereby gaining some efficiency. The result is placed in the out variable
 361          * $result. The return value is true if something was replaced.
 362          * @static
 363          **/
 364         function replaceMultiple( $magicarr, $subject, &$result ){
 365                 $search = array();
 366                 $replace = array();
 367                 foreach( $magicarr as $id => $replacement ){
 368                         $mw = MagicWord::get( $id );
 369                         $search[] = $mw->getRegex();
 370                         $replace[] = $replacement;
 371                 }
 372
 373                 $result = preg_replace( $search, $replace, $subject );
 374                 return !($result === $subject);
 375         }
 376
 377         /**
 378          * Adds all the synonyms of this MagicWord to an array, to allow quick
 379          * lookup in a list of magic words
 380          */
 381         function addToArray( &$array, $value ) {
 382                 global $wgContLang;
 383                 foreach ( $this->mSynonyms as $syn ) {
 384                         $array[$wgContLang->lc($syn)] = $value;
 385                 }
 386         }
 387
 388         function isCaseSensitive() {
 389                 return $this->mCaseSensitive;
 390         }
 391 }
 392
 393 ?>