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