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 $magicWords = array(
  16         'MAG_REDIRECT',
  17         'MAG_NOTOC',
  18         'MAG_START',
  19         'MAG_CURRENTMONTH',
  20         'MAG_CURRENTMONTHNAME',
  21         'MAG_CURRENTMONTHNAMEGEN',
  22         'MAG_CURRENTMONTHABBREV',
  23         'MAG_CURRENTDAY',
  24         'MAG_CURRENTDAY2',
  25         'MAG_CURRENTDAYNAME',
  26         'MAG_CURRENTYEAR',
  27         'MAG_CURRENTTIME',
  28         'MAG_NUMBEROFARTICLES',
  29         'MAG_SUBST',
  30         'MAG_MSG',
  31         'MAG_MSGNW',
  32         'MAG_NOEDITSECTION',
  33         'MAG_END',
  34         'MAG_IMG_THUMBNAIL',
  35         'MAG_IMG_RIGHT',
  36         'MAG_IMG_LEFT',
  37         'MAG_IMG_NONE',
  38         'MAG_IMG_WIDTH',
  39         'MAG_IMG_CENTER',
  40         'MAG_INT',
  41         'MAG_FORCETOC',
  42         'MAG_SITENAME',
  43         'MAG_NS',
  44         'MAG_LOCALURL',
  45         'MAG_LOCALURLE',
  46         'MAG_SERVER',
  47         'MAG_IMG_FRAMED',
  48         'MAG_PAGENAME',
  49         'MAG_PAGENAMEE',
  50         'MAG_NAMESPACE',
  51         'MAG_NAMESPACEE',
  52         'MAG_TOC',
  53         'MAG_GRAMMAR',
  54         'MAG_NOTITLECONVERT',
  55         'MAG_NOCONTENTCONVERT',
  56         'MAG_CURRENTWEEK',
  57         'MAG_CURRENTDOW',
  58         'MAG_REVISIONID',
  59         'MAG_SCRIPTPATH',
  60         'MAG_SERVERNAME',
  61         'MAG_NUMBEROFFILES',
  62         'MAG_IMG_MANUALTHUMB',
  63         'MAG_PLURAL',
  64         'MAG_FULLURL',
  65         'MAG_FULLURLE',
  66         'MAG_LCFIRST',
  67         'MAG_UCFIRST',
  68         'MAG_LC',
  69         'MAG_UC',
  70         'MAG_FULLPAGENAME',
  71         'MAG_FULLPAGENAMEE',
  72 );
  73 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  74         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  75
  76 for ( $i = 0; $i < count( $magicWords ); ++$i )
  77         define( $magicWords[$i], $i );
  78
  79 $wgVariableIDs = array(
  80         MAG_CURRENTMONTH,
  81         MAG_CURRENTMONTHNAME,
  82         MAG_CURRENTMONTHNAMEGEN,
  83         MAG_CURRENTMONTHABBREV,
  84         MAG_CURRENTDAY,
  85         MAG_CURRENTDAY2,
  86         MAG_CURRENTDAYNAME,
  87         MAG_CURRENTYEAR,
  88         MAG_CURRENTTIME,
  89         MAG_NUMBEROFARTICLES,
  90         MAG_NUMBEROFFILES,
  91         MAG_SITENAME,
  92         MAG_SERVER,
  93         MAG_SERVERNAME,
  94         MAG_SCRIPTPATH,
  95         MAG_PAGENAME,
  96         MAG_PAGENAMEE,
  97         MAG_FULLPAGENAME,
  98         MAG_FULLPAGENAMEE,
  99         MAG_NAMESPACE,
 100         MAG_NAMESPACEE,
 101         MAG_CURRENTWEEK,
 102         MAG_CURRENTDOW,
 103         MAG_REVISIONID,
 104 );
 105 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 106         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 107
 108 /**
 109  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 110  * Usage:
 111  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 112  *
 113  * Possible future improvements:
 114  *   * Simultaneous searching for a number of magic words
 115  *   * $wgMagicWords in shared memory
 116  *
 117  * Please avoid reading the data out of one of these objects and then writing
 118  * special case code. If possible, add another match()-like function here.
 119  *
 120  * @package MediaWiki
 121  */
 122 class MagicWord {
 123         /**#@+
 124          * @access private
 125          */
 126         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 127         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 128         var $mModified;
 129         /**#@-*/
 130
 131         function MagicWord($id = 0, $syn = '', $cs = false) {
 132                 $this->mId = $id;
 133                 $this->mSynonyms = (array)$syn;
 134                 $this->mCaseSensitive = $cs;
 135                 $this->mRegex = '';
 136                 $this->mRegexStart = '';
 137                 $this->mVariableRegex = '';
 138                 $this->mVariableStartToEndRegex = '';
 139                 $this->mModified = false;
 140         }
 141
 142         /**
 143          * Factory: creates an object representing an ID
 144          * @static
 145          */
 146         function &get( $id ) {
 147                 global $wgMagicWords;
 148
 149                 if ( !is_array( $wgMagicWords ) ) {
 150                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 151                 }
 152                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 153                         $mw = new MagicWord();
 154                         $mw->load( $id );
 155                         $wgMagicWords[$id] = $mw;
 156                 }
 157                 return $wgMagicWords[$id];
 158         }
 159
 160         # Initialises this object with an ID
 161         function load( $id ) {
 162                 global $wgContLang;
 163                 $this->mId = $id;
 164                 $wgContLang->getMagic( $this );
 165         }
 166
 167         /**
 168          * Preliminary initialisation
 169          * @access private
 170          */
 171         function initRegex() {
 172                 #$variableClass = Title::legalChars();
 173                 # This was used for matching "$1" variables, but different uses of the feature will have
 174                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 175                 # not here. - IMSoP
 176
 177                 $escSyn = array();
 178                 foreach ( $this->mSynonyms as $synonym )
 179                         // In case a magic word contains /, like that's going to happen;)
 180                         $escSyn[] = preg_quote( $synonym, '/' );
 181                 $this->mBaseRegex = implode( '|', $escSyn );
 182
 183                 $case = $this->mCaseSensitive ? '' : 'i';
 184                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 185                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 186                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 187                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 188                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 189         }
 190
 191         /**
 192          * Gets a regex representing matching the word
 193          */
 194         function getRegex() {
 195                 if ($this->mRegex == '' ) {
 196                         $this->initRegex();
 197                 }
 198                 return $this->mRegex;
 199         }
 200
 201         /**
 202          * Gets a regex matching the word, if it is at the string start
 203          */
 204         function getRegexStart() {
 205                 if ($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 ( $matchcount == 0 ) {
 247                         return NULL;
 248                 } elseif ( count($matches) == 1 ) {
 249                         return $matches[0];
 250                 } else {
 251                         # multiple matched parts (variable match); some will be empty because of synonyms
 252                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 253                         $matches = array_values(array_filter($matches));
 254                         return $matches[1];
 255                 }
 256         }
 257
 258
 259         /**
 260          * Returns true if the text matches the word, and alters the
 261          * input string, removing all instances of the word
 262          */
 263         function matchAndRemove( &$text ) {
 264                 global $wgMagicFound;
 265                 $wgMagicFound = false;
 266                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 267                 return $wgMagicFound;
 268         }
 269
 270         function matchStartAndRemove( &$text ) {
 271                 global $wgMagicFound;
 272                 $wgMagicFound = false;
 273                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 274                 return $wgMagicFound;
 275         }
 276
 277
 278         /**
 279          * Replaces the word with something else
 280          */
 281         function replace( $replacement, $subject ) {
 282                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject );
 283                 $this->mModified = !($res === $subject);
 284                 return $res;
 285         }
 286
 287         /**
 288          * Variable handling: {{SUBST:xxx}} style words
 289          * Calls back a function to determine what to replace xxx with
 290          * Input word must contain $1
 291          */
 292         function substituteCallback( $text, $callback ) {
 293                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 294                 $this->mModified = !($res === $text);
 295                 return $res;
 296         }
 297
 298         /**
 299          * Matches the word, where $1 is a wildcard
 300          */
 301         function getVariableRegex()     {
 302                 if ( $this->mVariableRegex == '' ) {
 303                         $this->initRegex();
 304                 }
 305                 return $this->mVariableRegex;
 306         }
 307
 308         /**
 309          * Matches the entire string, where $1 is a wildcard
 310          */
 311         function getVariableStartToEndRegex() {
 312                 if ( $this->mVariableStartToEndRegex == '' ) {
 313                         $this->initRegex();
 314                 }
 315                 return $this->mVariableStartToEndRegex;
 316         }
 317
 318         /**
 319          * Accesses the synonym list directly
 320          */
 321         function getSynonym( $i ) {
 322                 return $this->mSynonyms[$i];
 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
 364 /**
 365  * Used in matchAndRemove()
 366  * @access private
 367  **/
 368 function pregRemoveAndRecord( $match ) {
 369         global $wgMagicFound;
 370         $wgMagicFound = true;
 371         return '';
 372 }
 373
 374 ?>