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_PATH',                      41);
  56
  57 $wgVariableIDs = array(
  58         MAG_CURRENTMONTH,
  59         MAG_CURRENTMONTHNAME,
  60         MAG_CURRENTMONTHNAMEGEN,
  61         MAG_CURRENTMONTHABBREV,
  62         MAG_CURRENTDAY,
  63         MAG_CURRENTDAYNAME,
  64         MAG_CURRENTYEAR,
  65         MAG_CURRENTTIME,
  66         MAG_NUMBEROFARTICLES,
  67         MAG_SITENAME,
  68         MAG_SERVER,
  69         MAG_PATH,
  70         MAG_PAGENAME,
  71         MAG_PAGENAMEE,
  72         MAG_NAMESPACE,
  73         MAG_CURRENTWEEK,
  74         MAG_CURRENTDOW,
  75         MAG_REVISIONID,
  76 );
  77
  78 /**
  79  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  80  * Usage:
  81  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  82  *
  83  * Possible future improvements:
  84  *   * Simultaneous searching for a number of magic words
  85  *   * $wgMagicWords in shared memory
  86  *
  87  * Please avoid reading the data out of one of these objects and then writing
  88  * special case code. If possible, add another match()-like function here.
  89  *
  90  * @package MediaWiki
  91  */
  92 class MagicWord {
  93         /**#@+
  94          * @access private
  95          */
  96         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  97         var $mRegexStart, $mBaseRegex, $mVariableRegex;
  98         var $mModified;
  99         /**#@-*/
 100
 101         function MagicWord($id = 0, $syn = '', $cs = false) {
 102                 $this->mId = $id;
 103                 $this->mSynonyms = (array)$syn;
 104                 $this->mCaseSensitive = $cs;
 105                 $this->mRegex = '';
 106                 $this->mRegexStart = '';
 107                 $this->mVariableRegex = '';
 108                 $this->mVariableStartToEndRegex = '';
 109                 $this->mModified = false;
 110         }
 111
 112         /**
 113          * Factory: creates an object representing an ID
 114          * @static
 115          */
 116         function &get( $id ) {
 117                 global $wgMagicWords;
 118
 119                 if ( !is_array( $wgMagicWords ) ) {
 120                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 121                 }
 122                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 123                         $mw = new MagicWord();
 124                         $mw->load( $id );
 125                         $wgMagicWords[$id] = $mw;
 126                 }
 127                 return $wgMagicWords[$id];
 128         }
 129
 130         # Initialises this object with an ID
 131         function load( $id ) {
 132                 global $wgContLang;
 133                 $this->mId = $id;
 134                 $wgContLang->getMagic( $this );
 135         }
 136
 137         /**
 138          * Preliminary initialisation
 139          * @private
 140          */
 141         function initRegex() {
 142                 #$variableClass = Title::legalChars();
 143                 # This was used for matching "$1" variables, but different uses of the feature will have
 144                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 145                 # not here. - IMSoP
 146                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 147                 $this->mBaseRegex = implode( '|', $escSyn );
 148                 $case = $this->mCaseSensitive ? '' : 'i';
 149                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 150                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 151                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 152                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 153                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 154         }
 155
 156         /**
 157          * Gets a regex representing matching the word
 158          */
 159         function getRegex() {
 160                 if ($this->mRegex == '' ) {
 161                         $this->initRegex();
 162                 }
 163                 return $this->mRegex;
 164         }
 165
 166         /**
 167          * Gets a regex matching the word, if it is at the string start
 168          */
 169         function getRegexStart() {
 170                 if ($this->mRegex == '' ) {
 171                         $this->initRegex();
 172                 }
 173                 return $this->mRegexStart;
 174         }
 175
 176         /**
 177          * regex without the slashes and what not
 178          */
 179         function getBaseRegex() {
 180                 if ($this->mRegex == '') {
 181                         $this->initRegex();
 182                 }
 183                 return $this->mBaseRegex;
 184         }
 185
 186         /**
 187          * Returns true if the text contains the word
 188          * @return bool
 189          */
 190         function match( $text ) {
 191                 return preg_match( $this->getRegex(), $text );
 192         }
 193
 194         /**
 195          * Returns true if the text starts with the word
 196          * @return bool
 197          */
 198         function matchStart( $text ) {
 199                 return preg_match( $this->getRegexStart(), $text );
 200         }
 201
 202         /**
 203          * Returns NULL if there's no match, the value of $1 otherwise
 204          * The return code is the matched string, if there's no variable
 205          * part in the regex and the matched variable part ($1) if there
 206          * is one.
 207          */
 208         function matchVariableStartToEnd( $text ) {
 209                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 210                 if ( $matchcount == 0 ) {
 211                         return NULL;
 212                 } elseif ( count($matches) == 1 ) {
 213                         return $matches[0];
 214                 } else {
 215                         # multiple matched parts (variable match); some will be empty because of synonyms
 216                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 217                         $matches = array_values(array_filter($matches));
 218                         return $matches[1];
 219                 }
 220         }
 221
 222
 223         /**
 224          * Returns true if the text matches the word, and alters the
 225          * input string, removing all instances of the word
 226          */
 227         function matchAndRemove( &$text ) {
 228                 global $wgMagicFound;
 229                 $wgMagicFound = false;
 230                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 231                 return $wgMagicFound;
 232         }
 233
 234         function matchStartAndRemove( &$text ) {
 235                 global $wgMagicFound;
 236                 $wgMagicFound = false;
 237                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 238                 return $wgMagicFound;
 239         }
 240
 241
 242         /**
 243          * Replaces the word with something else
 244          */
 245         function replace( $replacement, $subject ) {
 246                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 247                 $this->mModified = !($res === $subject);
 248                 return $res;
 249         }
 250
 251         /**
 252          * Variable handling: {{SUBST:xxx}} style words
 253          * Calls back a function to determine what to replace xxx with
 254          * Input word must contain $1
 255          */
 256         function substituteCallback( $text, $callback ) {
 257                 $regex = $this->getVariableRegex();
 258                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 259                 $this->mModified = !($res === $text);
 260                 return $res;
 261         }
 262
 263         /**
 264          * Matches the word, where $1 is a wildcard
 265          */
 266         function getVariableRegex()     {
 267                 if ( $this->mVariableRegex == '' ) {
 268                         $this->initRegex();
 269                 }
 270                 return $this->mVariableRegex;
 271         }
 272
 273         /**
 274          * Matches the entire string, where $1 is a wildcard
 275          */
 276         function getVariableStartToEndRegex() {
 277                 if ( $this->mVariableStartToEndRegex == '' ) {
 278                         $this->initRegex();
 279                 }
 280                 return $this->mVariableStartToEndRegex;
 281         }
 282
 283         /**
 284          * Accesses the synonym list directly
 285          */
 286         function getSynonym( $i ) {
 287                 return $this->mSynonyms[$i];
 288         }
 289
 290         /**
 291          * Returns true if the last call to replace() or substituteCallback()
 292          * returned a modified text, otherwise false.
 293          */
 294         function getWasModified(){
 295                 return $this->mModified;
 296         }
 297
 298         /**
 299          * $magicarr is an associative array of (magic word ID => replacement)
 300          * This method uses the php feature to do several replacements at the same time,
 301          * thereby gaining some efficiency. The result is placed in the out variable
 302          * $result. The return value is true if something was replaced.
 303          * @static
 304          **/
 305         function replaceMultiple( $magicarr, $subject, &$result ){
 306                 $search = array();
 307                 $replace = array();
 308                 foreach( $magicarr as $id => $replacement ){
 309                         $mw = MagicWord::get( $id );
 310                         $search[] = $mw->getRegex();
 311                         $replace[] = $replacement;
 312                 }
 313
 314                 $result = preg_replace( $search, $replace, $subject );
 315                 return !($result === $subject);
 316         }
 317
 318         /**
 319          * Adds all the synonyms of this MagicWord to an array, to allow quick
 320          * lookup in a list of magic words
 321          */
 322         function addToArray( &$array, $value ) {
 323                 foreach ( $this->mSynonyms as $syn ) {
 324                         $array[$syn] = $value;
 325                 }
 326         }
 327 }
 328
 329 /**
 330  * Used in matchAndRemove()
 331  * @private
 332  **/
 333 function pregRemoveAndRecord( $match ) {
 334         global $wgMagicFound;
 335         $wgMagicFound = true;
 336         return '';
 337 }
 338
 339 ?>