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
  59 $wgVariableIDs = array(
  60         MAG_CURRENTMONTH,
  61         MAG_CURRENTMONTHNAME,
  62         MAG_CURRENTMONTHNAMEGEN,
  63         MAG_CURRENTMONTHABBREV,
  64         MAG_CURRENTDAY,
  65         MAG_CURRENTDAYNAME,
  66         MAG_CURRENTYEAR,
  67         MAG_CURRENTTIME,
  68         MAG_NUMBEROFARTICLES,
  69         MAG_NUMBEROFFILES,
  70         MAG_SITENAME,
  71         MAG_SERVER,
  72         MAG_SERVERNAME,
  73         MAG_SCRIPTPATH,
  74         MAG_PAGENAME,
  75         MAG_PAGENAMEE,
  76         MAG_NAMESPACE,
  77         MAG_CURRENTWEEK,
  78         MAG_CURRENTDOW,
  79         MAG_REVISIONID,
  80 );
  81
  82 /**
  83  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  84  * Usage:
  85  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  86  *
  87  * Possible future improvements:
  88  *   * Simultaneous searching for a number of magic words
  89  *   * $wgMagicWords in shared memory
  90  *
  91  * Please avoid reading the data out of one of these objects and then writing
  92  * special case code. If possible, add another match()-like function here.
  93  *
  94  * @package MediaWiki
  95  */
  96 class MagicWord {
  97         /**#@+
  98          * @access private
  99          */
 100         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 101         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 102         var $mModified;
 103         /**#@-*/
 104
 105         function MagicWord($id = 0, $syn = '', $cs = false) {
 106                 $this->mId = $id;
 107                 $this->mSynonyms = (array)$syn;
 108                 $this->mCaseSensitive = $cs;
 109                 $this->mRegex = '';
 110                 $this->mRegexStart = '';
 111                 $this->mVariableRegex = '';
 112                 $this->mVariableStartToEndRegex = '';
 113                 $this->mModified = false;
 114         }
 115
 116         /**
 117          * Factory: creates an object representing an ID
 118          * @static
 119          */
 120         function &get( $id ) {
 121                 global $wgMagicWords;
 122
 123                 if ( !is_array( $wgMagicWords ) ) {
 124                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 125                 }
 126                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 127                         $mw = new MagicWord();
 128                         $mw->load( $id );
 129                         $wgMagicWords[$id] = $mw;
 130                 }
 131                 return $wgMagicWords[$id];
 132         }
 133
 134         # Initialises this object with an ID
 135         function load( $id ) {
 136                 global $wgContLang;
 137                 $this->mId = $id;
 138                 $wgContLang->getMagic( $this );
 139         }
 140
 141         /**
 142          * Preliminary initialisation
 143          * @private
 144          */
 145         function initRegex() {
 146                 #$variableClass = Title::legalChars();
 147                 # This was used for matching "$1" variables, but different uses of the feature will have
 148                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 149                 # not here. - IMSoP
 150                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 151                 $this->mBaseRegex = implode( '|', $escSyn );
 152                 $case = $this->mCaseSensitive ? '' : 'i';
 153                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 154                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 155                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 156                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 157                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 158         }
 159
 160         /**
 161          * Gets a regex representing matching the word
 162          */
 163         function getRegex() {
 164                 if ($this->mRegex == '' ) {
 165                         $this->initRegex();
 166                 }
 167                 return $this->mRegex;
 168         }
 169
 170         /**
 171          * Gets a regex matching the word, if it is at the string start
 172          */
 173         function getRegexStart() {
 174                 if ($this->mRegex == '' ) {
 175                         $this->initRegex();
 176                 }
 177                 return $this->mRegexStart;
 178         }
 179
 180         /**
 181          * regex without the slashes and what not
 182          */
 183         function getBaseRegex() {
 184                 if ($this->mRegex == '') {
 185                         $this->initRegex();
 186                 }
 187                 return $this->mBaseRegex;
 188         }
 189
 190         /**
 191          * Returns true if the text contains the word
 192          * @return bool
 193          */
 194         function match( $text ) {
 195                 return preg_match( $this->getRegex(), $text );
 196         }
 197
 198         /**
 199          * Returns true if the text starts with the word
 200          * @return bool
 201          */
 202         function matchStart( $text ) {
 203                 return preg_match( $this->getRegexStart(), $text );
 204         }
 205
 206         /**
 207          * Returns NULL if there's no match, the value of $1 otherwise
 208          * The return code is the matched string, if there's no variable
 209          * part in the regex and the matched variable part ($1) if there
 210          * is one.
 211          */
 212         function matchVariableStartToEnd( $text ) {
 213                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 214                 if ( $matchcount == 0 ) {
 215                         return NULL;
 216                 } elseif ( count($matches) == 1 ) {
 217                         return $matches[0];
 218                 } else {
 219                         # multiple matched parts (variable match); some will be empty because of synonyms
 220                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 221                         $matches = array_values(array_filter($matches));
 222                         return $matches[1];
 223                 }
 224         }
 225
 226
 227         /**
 228          * Returns true if the text matches the word, and alters the
 229          * input string, removing all instances of the word
 230          */
 231         function matchAndRemove( &$text ) {
 232                 global $wgMagicFound;
 233                 $wgMagicFound = false;
 234                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 235                 return $wgMagicFound;
 236         }
 237
 238         function matchStartAndRemove( &$text ) {
 239                 global $wgMagicFound;
 240                 $wgMagicFound = false;
 241                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 242                 return $wgMagicFound;
 243         }
 244
 245
 246         /**
 247          * Replaces the word with something else
 248          */
 249         function replace( $replacement, $subject ) {
 250                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 251                 $this->mModified = !($res === $subject);
 252                 return $res;
 253         }
 254
 255         /**
 256          * Variable handling: {{SUBST:xxx}} style words
 257          * Calls back a function to determine what to replace xxx with
 258          * Input word must contain $1
 259          */
 260         function substituteCallback( $text, $callback ) {
 261                 $regex = $this->getVariableRegex();
 262                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 263                 $this->mModified = !($res === $text);
 264                 return $res;
 265         }
 266
 267         /**
 268          * Matches the word, where $1 is a wildcard
 269          */
 270         function getVariableRegex()     {
 271                 if ( $this->mVariableRegex == '' ) {
 272                         $this->initRegex();
 273                 }
 274                 return $this->mVariableRegex;
 275         }
 276
 277         /**
 278          * Matches the entire string, where $1 is a wildcard
 279          */
 280         function getVariableStartToEndRegex() {
 281                 if ( $this->mVariableStartToEndRegex == '' ) {
 282                         $this->initRegex();
 283                 }
 284                 return $this->mVariableStartToEndRegex;
 285         }
 286
 287         /**
 288          * Accesses the synonym list directly
 289          */
 290         function getSynonym( $i ) {
 291                 return $this->mSynonyms[$i];
 292         }
 293
 294         /**
 295          * Returns true if the last call to replace() or substituteCallback()
 296          * returned a modified text, otherwise false.
 297          */
 298         function getWasModified(){
 299                 return $this->mModified;
 300         }
 301
 302         /**
 303          * $magicarr is an associative array of (magic word ID => replacement)
 304          * This method uses the php feature to do several replacements at the same time,
 305          * thereby gaining some efficiency. The result is placed in the out variable
 306          * $result. The return value is true if something was replaced.
 307          * @static
 308          **/
 309         function replaceMultiple( $magicarr, $subject, &$result ){
 310                 $search = array();
 311                 $replace = array();
 312                 foreach( $magicarr as $id => $replacement ){
 313                         $mw = MagicWord::get( $id );
 314                         $search[] = $mw->getRegex();
 315                         $replace[] = $replacement;
 316                 }
 317
 318                 $result = preg_replace( $search, $replace, $subject );
 319                 return !($result === $subject);
 320         }
 321
 322         /**
 323          * Adds all the synonyms of this MagicWord to an array, to allow quick
 324          * lookup in a list of magic words
 325          */
 326         function addToArray( &$array, $value ) {
 327                 foreach ( $this->mSynonyms as $syn ) {
 328                         $array[$syn] = $value;
 329                 }
 330         }
 331 }
 332
 333 /**
 334  * Used in matchAndRemove()
 335  * @private
 336  **/
 337 function pregRemoveAndRecord( $match ) {
 338         global $wgMagicFound;
 339         $wgMagicFound = true;
 340         return '';
 341 }
 342
 343 ?>