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