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 define('MAG_REDIRECT',                  0);
  15 define('MAG_NOTOC',                     1);
  16 define('MAG_START',                     2);
  17 define('MAG_CURRENTMONTH',              3);
  18 define('MAG_CURRENTMONTHNAME',          4);
  19 define('MAG_CURRENTMONTHNAMEGEN',       5);
  20 define('MAG_CURRENTMONTHABBREV',        6);
  21 define('MAG_CURRENTDAY',                7);
  22 define('MAG_CURRENTDAYNAME',            8);
  23 define('MAG_CURRENTYEAR',               9);
  24 define('MAG_CURRENTTIME',               10);
  25 define('MAG_NUMBEROFARTICLES',          11);
  26 define('MAG_SUBST',                     12);
  27 define('MAG_MSG',                       13);
  28 define('MAG_MSGNW',                     14);
  29 define('MAG_NOEDITSECTION',             15);
  30 define('MAG_END',                       16);
  31 define('MAG_IMG_THUMBNAIL',             17);
  32 define('MAG_IMG_RIGHT',                 18);
  33 define('MAG_IMG_LEFT',                  19);
  34 define('MAG_IMG_NONE',                  20);
  35 define('MAG_IMG_WIDTH',                 21);
  36 define('MAG_IMG_CENTER',                22);
  37 define('MAG_INT',                       23);
  38 define('MAG_FORCETOC',                  24);
  39 define('MAG_SITENAME',                  25);
  40 define('MAG_NS',                        26);
  41 define('MAG_LOCALURL',                  27);
  42 define('MAG_LOCALURLE',                 28);
  43 define('MAG_SERVER',                    29);
  44 define('MAG_IMG_FRAMED',                30);
  45 define('MAG_PAGENAME',                  31);
  46 define('MAG_PAGENAMEE',                 32);
  47 define('MAG_NAMESPACE',                 33);
  48 define('MAG_TOC',                       34);
  49 define('MAG_GRAMMAR',                   35);
  50 define('MAG_NOTITLECONVERT',            36);
  51 define('MAG_NOCONTENTCONVERT',          37);
  52 define('MAG_CURRENTWEEK',               38);
  53 define('MAG_CURRENTDOW',                39);
  54 define('MAG_REVISIONID',                40);
  55 define('MAG_SCRIPTPATH',                41);
  56 define('MAG_SERVERNAME',                42);
  57 define('MAG_NUMBEROFFILES',             43);
  58 define('MAG_IMG_MANUALTHUMB',           44);
  59
  60 $wgVariableIDs = array(
  61         MAG_CURRENTMONTH,
  62         MAG_CURRENTMONTHNAME,
  63         MAG_CURRENTMONTHNAMEGEN,
  64         MAG_CURRENTMONTHABBREV,
  65         MAG_CURRENTDAY,
  66         MAG_CURRENTDAYNAME,
  67         MAG_CURRENTYEAR,
  68         MAG_CURRENTTIME,
  69         MAG_NUMBEROFARTICLES,
  70         MAG_NUMBEROFFILES,
  71         MAG_SITENAME,
  72         MAG_SERVER,
  73         MAG_SERVERNAME,
  74         MAG_SCRIPTPATH,
  75         MAG_PAGENAME,
  76         MAG_PAGENAMEE,
  77         MAG_NAMESPACE,
  78         MAG_CURRENTWEEK,
  79         MAG_CURRENTDOW,
  80         MAG_REVISIONID,
  81 );
  82
  83 /**
  84  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  85  * Usage:
  86  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  87  *
  88  * Possible future improvements:
  89  *   * Simultaneous searching for a number of magic words
  90  *   * $wgMagicWords in shared memory
  91  *
  92  * Please avoid reading the data out of one of these objects and then writing
  93  * special case code. If possible, add another match()-like function here.
  94  *
  95  * @package MediaWiki
  96  */
  97 class MagicWord {
  98         /**#@+
  99          * @access private
 100          */
 101         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 102         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 103         var $mModified;
 104         /**#@-*/
 105
 106         function MagicWord($id = 0, $syn = '', $cs = false) {
 107                 $this->mId = $id;
 108                 $this->mSynonyms = (array)$syn;
 109                 $this->mCaseSensitive = $cs;
 110                 $this->mRegex = '';
 111                 $this->mRegexStart = '';
 112                 $this->mVariableRegex = '';
 113                 $this->mVariableStartToEndRegex = '';
 114                 $this->mModified = false;
 115         }
 116
 117         /**
 118          * Factory: creates an object representing an ID
 119          * @static
 120          */
 121         function &get( $id ) {
 122                 global $wgMagicWords;
 123
 124                 if ( !is_array( $wgMagicWords ) ) {
 125                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 126                 }
 127                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 128                         $mw = new MagicWord();
 129                         $mw->load( $id );
 130                         $wgMagicWords[$id] = $mw;
 131                 }
 132                 return $wgMagicWords[$id];
 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                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 152                 $this->mBaseRegex = implode( '|', $escSyn );
 153                 $case = $this->mCaseSensitive ? '' : 'i';
 154                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 155                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 156                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 157                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 158                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 159         }
 160
 161         /**
 162          * Gets a regex representing matching the word
 163          */
 164         function getRegex() {
 165                 if ($this->mRegex == '' ) {
 166                         $this->initRegex();
 167                 }
 168                 return $this->mRegex;
 169         }
 170
 171         /**
 172          * Gets a regex matching the word, if it is at the string start
 173          */
 174         function getRegexStart() {
 175                 if ($this->mRegex == '' ) {
 176                         $this->initRegex();
 177                 }
 178                 return $this->mRegexStart;
 179         }
 180
 181         /**
 182          * regex without the slashes and what not
 183          */
 184         function getBaseRegex() {
 185                 if ($this->mRegex == '') {
 186                         $this->initRegex();
 187                 }
 188                 return $this->mBaseRegex;
 189         }
 190
 191         /**
 192          * Returns true if the text contains the word
 193          * @return bool
 194          */
 195         function match( $text ) {
 196                 return preg_match( $this->getRegex(), $text );
 197         }
 198
 199         /**
 200          * Returns true if the text starts with the word
 201          * @return bool
 202          */
 203         function matchStart( $text ) {
 204                 return preg_match( $this->getRegexStart(), $text );
 205         }
 206
 207         /**
 208          * Returns NULL if there's no match, the value of $1 otherwise
 209          * The return code is the matched string, if there's no variable
 210          * part in the regex and the matched variable part ($1) if there
 211          * is one.
 212          */
 213         function matchVariableStartToEnd( $text ) {
 214                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 215                 if ( $matchcount == 0 ) {
 216                         return NULL;
 217                 } elseif ( count($matches) == 1 ) {
 218                         return $matches[0];
 219                 } else {
 220                         # multiple matched parts (variable match); some will be empty because of synonyms
 221                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 222                         $matches = array_values(array_filter($matches));
 223                         return $matches[1];
 224                 }
 225         }
 226
 227
 228         /**
 229          * Returns true if the text matches the word, and alters the
 230          * input string, removing all instances of the word
 231          */
 232         function matchAndRemove( &$text ) {
 233                 global $wgMagicFound;
 234                 $wgMagicFound = false;
 235                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 236                 return $wgMagicFound;
 237         }
 238
 239         function matchStartAndRemove( &$text ) {
 240                 global $wgMagicFound;
 241                 $wgMagicFound = false;
 242                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 243                 return $wgMagicFound;
 244         }
 245
 246
 247         /**
 248          * Replaces the word with something else
 249          */
 250         function replace( $replacement, $subject ) {
 251                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 252                 $this->mModified = !($res === $subject);
 253                 return $res;
 254         }
 255
 256         /**
 257          * Variable handling: {{SUBST:xxx}} style words
 258          * Calls back a function to determine what to replace xxx with
 259          * Input word must contain $1
 260          */
 261         function substituteCallback( $text, $callback ) {
 262                 $regex = $this->getVariableRegex();
 263                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 264                 $this->mModified = !($res === $text);
 265                 return $res;
 266         }
 267
 268         /**
 269          * Matches the word, where $1 is a wildcard
 270          */
 271         function getVariableRegex()     {
 272                 if ( $this->mVariableRegex == '' ) {
 273                         $this->initRegex();
 274                 }
 275                 return $this->mVariableRegex;
 276         }
 277
 278         /**
 279          * Matches the entire string, where $1 is a wildcard
 280          */
 281         function getVariableStartToEndRegex() {
 282                 if ( $this->mVariableStartToEndRegex == '' ) {
 283                         $this->initRegex();
 284                 }
 285                 return $this->mVariableStartToEndRegex;
 286         }
 287
 288         /**
 289          * Accesses the synonym list directly
 290          */
 291         function getSynonym( $i ) {
 292                 return $this->mSynonyms[$i];
 293         }
 294
 295         /**
 296          * Returns true if the last call to replace() or substituteCallback()
 297          * returned a modified text, otherwise false.
 298          */
 299         function getWasModified(){
 300                 return $this->mModified;
 301         }
 302
 303         /**
 304          * $magicarr is an associative array of (magic word ID => replacement)
 305          * This method uses the php feature to do several replacements at the same time,
 306          * thereby gaining some efficiency. The result is placed in the out variable
 307          * $result. The return value is true if something was replaced.
 308          * @static
 309          **/
 310         function replaceMultiple( $magicarr, $subject, &$result ){
 311                 $search = array();
 312                 $replace = array();
 313                 foreach( $magicarr as $id => $replacement ){
 314                         $mw = MagicWord::get( $id );
 315                         $search[] = $mw->getRegex();
 316                         $replace[] = $replacement;
 317                 }
 318
 319                 $result = preg_replace( $search, $replace, $subject );
 320                 return !($result === $subject);
 321         }
 322
 323         /**
 324          * Adds all the synonyms of this MagicWord to an array, to allow quick
 325          * lookup in a list of magic words
 326          */
 327         function addToArray( &$array, $value ) {
 328                 foreach ( $this->mSynonyms as $syn ) {
 329                         $array[$syn] = $value;
 330                 }
 331         }
 332 }
 333
 334 /**
 335  * Used in matchAndRemove()
 336  * @private
 337  **/
 338 function pregRemoveAndRecord( $match ) {
 339         global $wgMagicFound;
 340         $wgMagicFound = true;
 341         return '';
 342 }
 343
 344 ?>