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 define('MAG_PLURAL',                    45);
  60 define('MAG_FULLURL',                   46);
  61 define('MAG_FULLURLE',                  47);
  62
  63 $wgVariableIDs = array(
  64         MAG_CURRENTMONTH,
  65         MAG_CURRENTMONTHNAME,
  66         MAG_CURRENTMONTHNAMEGEN,
  67         MAG_CURRENTMONTHABBREV,
  68         MAG_CURRENTDAY,
  69         MAG_CURRENTDAYNAME,
  70         MAG_CURRENTYEAR,
  71         MAG_CURRENTTIME,
  72         MAG_NUMBEROFARTICLES,
  73         MAG_NUMBEROFFILES,
  74         MAG_SITENAME,
  75         MAG_SERVER,
  76         MAG_SERVERNAME,
  77         MAG_SCRIPTPATH,
  78         MAG_PAGENAME,
  79         MAG_PAGENAMEE,
  80         MAG_NAMESPACE,
  81         MAG_CURRENTWEEK,
  82         MAG_CURRENTDOW,
  83         MAG_REVISIONID,
  84 );
  85
  86 /**
  87  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  88  * Usage:
  89  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  90  *
  91  * Possible future improvements:
  92  *   * Simultaneous searching for a number of magic words
  93  *   * $wgMagicWords in shared memory
  94  *
  95  * Please avoid reading the data out of one of these objects and then writing
  96  * special case code. If possible, add another match()-like function here.
  97  *
  98  * @package MediaWiki
  99  */
 100 class MagicWord {
 101         /**#@+
 102          * @access private
 103          */
 104         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 105         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 106         var $mModified;
 107         /**#@-*/
 108
 109         function MagicWord($id = 0, $syn = '', $cs = false) {
 110                 $this->mId = $id;
 111                 $this->mSynonyms = (array)$syn;
 112                 $this->mCaseSensitive = $cs;
 113                 $this->mRegex = '';
 114                 $this->mRegexStart = '';
 115                 $this->mVariableRegex = '';
 116                 $this->mVariableStartToEndRegex = '';
 117                 $this->mModified = false;
 118         }
 119
 120         /**
 121          * Factory: creates an object representing an ID
 122          * @static
 123          */
 124         function &get( $id ) {
 125                 global $wgMagicWords;
 126
 127                 if ( !is_array( $wgMagicWords ) ) {
 128                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 129                 }
 130                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 131                         $mw = new MagicWord();
 132                         $mw->load( $id );
 133                         $wgMagicWords[$id] = $mw;
 134                 }
 135                 return $wgMagicWords[$id];
 136         }
 137
 138         # Initialises this object with an ID
 139         function load( $id ) {
 140                 global $wgContLang;
 141                 $this->mId = $id;
 142                 $wgContLang->getMagic( $this );
 143         }
 144
 145         /**
 146          * Preliminary initialisation
 147          * @private
 148          */
 149         function initRegex() {
 150                 #$variableClass = Title::legalChars();
 151                 # This was used for matching "$1" variables, but different uses of the feature will have
 152                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 153                 # not here. - IMSoP
 154                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 155                 $this->mBaseRegex = implode( '|', $escSyn );
 156                 $case = $this->mCaseSensitive ? '' : 'i';
 157                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 158                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 159                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 160                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 161                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 162         }
 163
 164         /**
 165          * Gets a regex representing matching the word
 166          */
 167         function getRegex() {
 168                 if ($this->mRegex == '' ) {
 169                         $this->initRegex();
 170                 }
 171                 return $this->mRegex;
 172         }
 173
 174         /**
 175          * Gets a regex matching the word, if it is at the string start
 176          */
 177         function getRegexStart() {
 178                 if ($this->mRegex == '' ) {
 179                         $this->initRegex();
 180                 }
 181                 return $this->mRegexStart;
 182         }
 183
 184         /**
 185          * regex without the slashes and what not
 186          */
 187         function getBaseRegex() {
 188                 if ($this->mRegex == '') {
 189                         $this->initRegex();
 190                 }
 191                 return $this->mBaseRegex;
 192         }
 193
 194         /**
 195          * Returns true if the text contains the word
 196          * @return bool
 197          */
 198         function match( $text ) {
 199                 return preg_match( $this->getRegex(), $text );
 200         }
 201
 202         /**
 203          * Returns true if the text starts with the word
 204          * @return bool
 205          */
 206         function matchStart( $text ) {
 207                 return preg_match( $this->getRegexStart(), $text );
 208         }
 209
 210         /**
 211          * Returns NULL if there's no match, the value of $1 otherwise
 212          * The return code is the matched string, if there's no variable
 213          * part in the regex and the matched variable part ($1) if there
 214          * is one.
 215          */
 216         function matchVariableStartToEnd( $text ) {
 217                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 218                 if ( $matchcount == 0 ) {
 219                         return NULL;
 220                 } elseif ( count($matches) == 1 ) {
 221                         return $matches[0];
 222                 } else {
 223                         # multiple matched parts (variable match); some will be empty because of synonyms
 224                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 225                         $matches = array_values(array_filter($matches));
 226                         return $matches[1];
 227                 }
 228         }
 229
 230
 231         /**
 232          * Returns true if the text matches the word, and alters the
 233          * input string, removing all instances of the word
 234          */
 235         function matchAndRemove( &$text ) {
 236                 global $wgMagicFound;
 237                 $wgMagicFound = false;
 238                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 239                 return $wgMagicFound;
 240         }
 241
 242         function matchStartAndRemove( &$text ) {
 243                 global $wgMagicFound;
 244                 $wgMagicFound = false;
 245                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 246                 return $wgMagicFound;
 247         }
 248
 249
 250         /**
 251          * Replaces the word with something else
 252          */
 253         function replace( $replacement, $subject ) {
 254                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 255                 $this->mModified = !($res === $subject);
 256                 return $res;
 257         }
 258
 259         /**
 260          * Variable handling: {{SUBST:xxx}} style words
 261          * Calls back a function to determine what to replace xxx with
 262          * Input word must contain $1
 263          */
 264         function substituteCallback( $text, $callback ) {
 265                 $regex = $this->getVariableRegex();
 266                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 267                 $this->mModified = !($res === $text);
 268                 return $res;
 269         }
 270
 271         /**
 272          * Matches the word, where $1 is a wildcard
 273          */
 274         function getVariableRegex()     {
 275                 if ( $this->mVariableRegex == '' ) {
 276                         $this->initRegex();
 277                 }
 278                 return $this->mVariableRegex;
 279         }
 280
 281         /**
 282          * Matches the entire string, where $1 is a wildcard
 283          */
 284         function getVariableStartToEndRegex() {
 285                 if ( $this->mVariableStartToEndRegex == '' ) {
 286                         $this->initRegex();
 287                 }
 288                 return $this->mVariableStartToEndRegex;
 289         }
 290
 291         /**
 292          * Accesses the synonym list directly
 293          */
 294         function getSynonym( $i ) {
 295                 return $this->mSynonyms[$i];
 296         }
 297
 298         /**
 299          * Returns true if the last call to replace() or substituteCallback()
 300          * returned a modified text, otherwise false.
 301          */
 302         function getWasModified(){
 303                 return $this->mModified;
 304         }
 305
 306         /**
 307          * $magicarr is an associative array of (magic word ID => replacement)
 308          * This method uses the php feature to do several replacements at the same time,
 309          * thereby gaining some efficiency. The result is placed in the out variable
 310          * $result. The return value is true if something was replaced.
 311          * @static
 312          **/
 313         function replaceMultiple( $magicarr, $subject, &$result ){
 314                 $search = array();
 315                 $replace = array();
 316                 foreach( $magicarr as $id => $replacement ){
 317                         $mw = MagicWord::get( $id );
 318                         $search[] = $mw->getRegex();
 319                         $replace[] = $replacement;
 320                 }
 321
 322                 $result = preg_replace( $search, $replace, $subject );
 323                 return !($result === $subject);
 324         }
 325
 326         /**
 327          * Adds all the synonyms of this MagicWord to an array, to allow quick
 328          * lookup in a list of magic words
 329          */
 330         function addToArray( &$array, $value ) {
 331                 foreach ( $this->mSynonyms as $syn ) {
 332                         $array[$syn] = $value;
 333                 }
 334         }
 335 }
 336
 337 /**
 338  * Used in matchAndRemove()
 339  * @private
 340  **/
 341 function pregRemoveAndRecord( $match ) {
 342         global $wgMagicFound;
 343         $wgMagicFound = true;
 344         return '';
 345 }
 346
 347 ?>