includes/MagicWord.php

   1 <?php
   2 /**
   3  * File for magic words
   4  * @package MediaWiki
   5  */
   6
   7 /**
   8  * private
   9  */
  10 $wgMagicFound = false;
  11
  12 /** Actual keyword to be used is set in Language.php */
  13 define('MAG_REDIRECT', 0);
  14 define('MAG_NOTOC', 1);
  15 define('MAG_START', 2);
  16 define('MAG_CURRENTMONTH', 3);
  17 define('MAG_CURRENTMONTHNAME', 4);
  18 define('MAG_CURRENTMONTHNAMEGEN', 5);
  19 define('MAG_CURRENTMONTHABBREV', 6);
  20 define('MAG_CURRENTDAY', 7);
  21 define('MAG_CURRENTDAYNAME', 8);
  22 define('MAG_CURRENTYEAR', 9);
  23 define('MAG_CURRENTTIME', 10);
  24 define('MAG_NUMBEROFARTICLES', 11);
  25 define('MAG_MSG', 12);
  26 define('MAG_SUBST', 13);
  27 define('MAG_MSGNW', 14);
  28 define('MAG_NOEDITSECTION', 15);
  29 define('MAG_END', 16);
  30 define('MAG_IMG_THUMBNAIL',  17);
  31 define('MAG_IMG_RIGHT',      18);
  32 define('MAG_IMG_LEFT',       19);
  33 define('MAG_IMG_NONE',       20);
  34 define('MAG_IMG_WIDTH',      21);
  35 define('MAG_IMG_CENTER',      22);
  36 define('MAG_INT', 23);
  37 define('MAG_FORCETOC', 24);
  38 define('MAG_SITENAME', 25);
  39 define('MAG_NS', 26);
  40 define('MAG_LOCALURL', 27);
  41 define('MAG_LOCALURLE', 28);
  42 define('MAG_SERVER', 29);
  43 define('MAG_IMG_FRAMED', 30);
  44 define('MAG_PAGENAME', 31);
  45 define('MAG_PAGENAMEE', 32);
  46 define('MAG_NAMESPACE', 33);
  47 define('MAG_TOC', 34);
  48 define('MAG_GRAMMAR', 35);
  49 define('MAG_NOTITLECONVERT', 36);
  50 define('MAG_NOCONTENTCONVERT', 37);
  51 define('MAG_CURRENTWEEK', 38);
  52 define('MAG_CURRENTDOW', 39);
  53 define('MAG_REVISIONID', 40);
  54
  55 $wgVariableIDs = array(
  56         MAG_CURRENTMONTH,
  57         MAG_CURRENTMONTHNAME,
  58         MAG_CURRENTMONTHNAMEGEN,
  59         MAG_CURRENTMONTHABBREV,
  60         MAG_CURRENTDAY,
  61         MAG_CURRENTDAYNAME,
  62         MAG_CURRENTYEAR,
  63         MAG_CURRENTTIME,
  64         MAG_NUMBEROFARTICLES,
  65         MAG_SITENAME,
  66         MAG_SERVER,
  67         MAG_PAGENAME,
  68         MAG_PAGENAMEE,
  69         MAG_NAMESPACE,
  70         MAG_CURRENTWEEK,
  71         MAG_CURRENTDOW,
  72         MAG_REVISIONID,
  73 );
  74
  75 /**
  76  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  77  * Usage:
  78  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  79  *
  80  * Possible future improvements:
  81  *   * Simultaneous searching for a number of magic words
  82  *   * $wgMagicWords in shared memory
  83  *
  84  * Please avoid reading the data out of one of these objects and then writing
  85  * special case code. If possible, add another match()-like function here.
  86  *
  87  * @package MediaWiki
  88  */
  89 class MagicWord {
  90         /**#@+
  91          * @access private
  92          */
  93         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  94         var $mRegexStart, $mBaseRegex, $mVariableRegex;
  95         var $mModified;
  96         /**#@-*/
  97
  98         function MagicWord($id = 0, $syn = '', $cs = false) {
  99                 $this->mId = $id;
 100                 $this->mSynonyms = (array)$syn;
 101                 $this->mCaseSensitive = $cs;
 102                 $this->mRegex = '';
 103                 $this->mRegexStart = '';
 104                 $this->mVariableRegex = '';
 105                 $this->mVariableStartToEndRegex = '';
 106                 $this->mModified = false;
 107         }
 108
 109         /**
 110          * Factory: creates an object representing an ID
 111          * @static
 112          */
 113         function &get( $id ) {
 114                 global $wgMagicWords;
 115
 116                 if ( !is_array( $wgMagicWords ) ) {
 117                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 118                 }
 119                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 120                         $mw = new MagicWord();
 121                         $mw->load( $id );
 122                         $wgMagicWords[$id] = $mw;
 123                 }
 124                 return $wgMagicWords[$id];
 125         }
 126
 127         # Initialises this object with an ID
 128         function load( $id ) {
 129                 global $wgContLang;
 130                 $this->mId = $id;
 131                 $wgContLang->getMagic( $this );
 132         }
 133
 134         /**
 135          * Preliminary initialisation
 136          * @private
 137          */
 138         function initRegex() {
 139                 $variableClass = Title::legalChars();
 140                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 141                 $this->mBaseRegex = implode( '|', $escSyn );
 142                 $case = $this->mCaseSensitive ? '' : 'i';
 143                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 144                 $this->mRegexStart = "/^({$this->mBaseRegex})/{$case}";
 145                 $this->mVariableRegex = str_replace( "\\$1", "([$variableClass]*?)", $this->mRegex );
 146                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "([$variableClass]*?)",
 147                         "/^({$this->mBaseRegex})$/{$case}" );
 148         }
 149
 150         /**
 151          * Gets a regex representing matching the word
 152          */
 153         function getRegex() {
 154                 if ($this->mRegex == '' ) {
 155                         $this->initRegex();
 156                 }
 157                 return $this->mRegex;
 158         }
 159
 160         /**
 161          * Gets a regex matching the word, if it is at the string start
 162          */
 163         function getRegexStart() {
 164                 if ($this->mRegex == '' ) {
 165                         $this->initRegex();
 166                 }
 167                 return $this->mRegexStart;
 168         }
 169
 170         /**
 171          * regex without the slashes and what not
 172          */
 173         function getBaseRegex() {
 174                 if ($this->mRegex == '') {
 175                         $this->initRegex();
 176                 }
 177                 return $this->mBaseRegex;
 178         }
 179
 180         /**
 181          * Returns true if the text contains the word
 182          * @return bool
 183          */
 184         function match( $text ) {
 185                 return preg_match( $this->getRegex(), $text );
 186         }
 187
 188         /**
 189          * Returns true if the text starts with the word
 190          * @return bool
 191          */
 192         function matchStart( $text ) {
 193                 return preg_match( $this->getRegexStart(), $text );
 194         }
 195
 196         /**
 197          * Returns NULL if there's no match, the value of $1 otherwise
 198          * The return code is the matched string, if there's no variable
 199          * part in the regex and the matched variable part ($1) if there
 200          * is one.
 201          */
 202         function matchVariableStartToEnd( $text ) {
 203                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 204                 if ( $matchcount == 0 ) {
 205                         return NULL;
 206                 } elseif ( count($matches) == 2 ) {
 207                         return $matches[0];
 208                 } else {
 209                         return $matches[2];
 210                 }
 211         }
 212
 213
 214         /**
 215          * Returns true if the text matches the word, and alters the
 216          * input string, removing all instances of the word
 217          */
 218         function matchAndRemove( &$text ) {
 219                 global $wgMagicFound;
 220                 $wgMagicFound = false;
 221                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 222                 return $wgMagicFound;
 223         }
 224
 225         function matchStartAndRemove( &$text ) {
 226                 global $wgMagicFound;
 227                 $wgMagicFound = false;
 228                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 229                 return $wgMagicFound;
 230         }
 231
 232
 233         /**
 234          * Replaces the word with something else
 235          */
 236         function replace( $replacement, $subject ) {
 237                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 238                 $this->mModified = !($res === $subject);
 239                 return $res;
 240         }
 241
 242         /**
 243          * Variable handling: {{SUBST:xxx}} style words
 244          * Calls back a function to determine what to replace xxx with
 245          * Input word must contain $1
 246          */
 247         function substituteCallback( $text, $callback ) {
 248                 $regex = $this->getVariableRegex();
 249                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 250                 $this->mModified = !($res === $text);
 251                 return $res;
 252         }
 253
 254         /**
 255          * Matches the word, where $1 is a wildcard
 256          */
 257         function getVariableRegex()     {
 258                 if ( $this->mVariableRegex == '' ) {
 259                         $this->initRegex();
 260                 }
 261                 return $this->mVariableRegex;
 262         }
 263
 264         /**
 265          * Matches the entire string, where $1 is a wildcard
 266          */
 267         function getVariableStartToEndRegex() {
 268                 if ( $this->mVariableStartToEndRegex == '' ) {
 269                         $this->initRegex();
 270                 }
 271                 return $this->mVariableStartToEndRegex;
 272         }
 273
 274         /**
 275          * Accesses the synonym list directly
 276          */
 277         function getSynonym( $i ) {
 278                 return $this->mSynonyms[$i];
 279         }
 280
 281         /**
 282          * Returns true if the last call to replace() or substituteCallback()
 283          * returned a modified text, otherwise false.
 284          */
 285         function getWasModified(){
 286                 return $this->mModified;
 287         }
 288
 289         /**
 290          * $magicarr is an associative array of (magic word ID => replacement)
 291          * This method uses the php feature to do several replacements at the same time,
 292          * thereby gaining some efficiency. The result is placed in the out variable
 293          * $result. The return value is true if something was replaced.
 294          * @static
 295          **/
 296         function replaceMultiple( $magicarr, $subject, &$result ){
 297                 $search = array();
 298                 $replace = array();
 299                 foreach( $magicarr as $id => $replacement ){
 300                         $mw = MagicWord::get( $id );
 301                         $search[] = $mw->getRegex();
 302                         $replace[] = $replacement;
 303                 }
 304
 305                 $result = preg_replace( $search, $replace, $subject );
 306                 return !($result === $subject);
 307         }
 308
 309         /**
 310          * Adds all the synonyms of this MagicWord to an array, to allow quick
 311          * lookup in a list of magic words
 312          */
 313         function addToArray( &$array, $value ) {
 314                 foreach ( $this->mSynonyms as $syn ) {
 315                         $array[$syn] = $value;
 316                 }
 317         }
 318 }
 319
 320 /**
 321  * Used in matchAndRemove()
 322  * @private
 323  **/
 324 function pregRemoveAndRecord( $match ) {
 325         global $wgMagicFound;
 326         $wgMagicFound = true;
 327         return '';
 328 }
 329
 330 ?>