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
  58 $wgVariableIDs = array(
  59         MAG_CURRENTMONTH,
  60         MAG_CURRENTMONTHNAME,
  61         MAG_CURRENTMONTHNAMEGEN,
  62         MAG_CURRENTMONTHABBREV,
  63         MAG_CURRENTDAY,
  64         MAG_CURRENTDAYNAME,
  65         MAG_CURRENTYEAR,
  66         MAG_CURRENTTIME,
  67         MAG_NUMBEROFARTICLES,
  68         MAG_SITENAME,
  69         MAG_SERVER,
  70         MAG_SERVERNAME,
  71         MAG_SCRIPTPATH,
  72         MAG_PAGENAME,
  73         MAG_PAGENAMEE,
  74         MAG_NAMESPACE,
  75         MAG_CURRENTWEEK,
  76         MAG_CURRENTDOW,
  77         MAG_REVISIONID,
  78 );
  79
  80 /**
  81  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  82  * Usage:
  83  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  84  *
  85  * Possible future improvements:
  86  *   * Simultaneous searching for a number of magic words
  87  *   * $wgMagicWords in shared memory
  88  *
  89  * Please avoid reading the data out of one of these objects and then writing
  90  * special case code. If possible, add another match()-like function here.
  91  *
  92  * @package MediaWiki
  93  */
  94 class MagicWord {
  95         /**#@+
  96          * @access private
  97          */
  98         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  99         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 100         var $mModified;
 101         /**#@-*/
 102
 103         function MagicWord($id = 0, $syn = '', $cs = false) {
 104                 $this->mId = $id;
 105                 $this->mSynonyms = (array)$syn;
 106                 $this->mCaseSensitive = $cs;
 107                 $this->mRegex = '';
 108                 $this->mRegexStart = '';
 109                 $this->mVariableRegex = '';
 110                 $this->mVariableStartToEndRegex = '';
 111                 $this->mModified = false;
 112         }
 113
 114         /**
 115          * Factory: creates an object representing an ID
 116          * @static
 117          */
 118         function &get( $id ) {
 119                 global $wgMagicWords;
 120
 121                 if ( !is_array( $wgMagicWords ) ) {
 122                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 123                 }
 124                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 125                         $mw = new MagicWord();
 126                         $mw->load( $id );
 127                         $wgMagicWords[$id] = $mw;
 128                 }
 129                 return $wgMagicWords[$id];
 130         }
 131
 132         # Initialises this object with an ID
 133         function load( $id ) {
 134                 global $wgContLang;
 135                 $this->mId = $id;
 136                 $wgContLang->getMagic( $this );
 137         }
 138
 139         /**
 140          * Preliminary initialisation
 141          * @private
 142          */
 143         function initRegex() {
 144                 #$variableClass = Title::legalChars();
 145                 # This was used for matching "$1" variables, but different uses of the feature will have
 146                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 147                 # not here. - IMSoP
 148                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 149                 $this->mBaseRegex = implode( '|', $escSyn );
 150                 $case = $this->mCaseSensitive ? '' : 'i';
 151                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 152                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 153                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 154                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 155                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 156         }
 157
 158         /**
 159          * Gets a regex representing matching the word
 160          */
 161         function getRegex() {
 162                 if ($this->mRegex == '' ) {
 163                         $this->initRegex();
 164                 }
 165                 return $this->mRegex;
 166         }
 167
 168         /**
 169          * Gets a regex matching the word, if it is at the string start
 170          */
 171         function getRegexStart() {
 172                 if ($this->mRegex == '' ) {
 173                         $this->initRegex();
 174                 }
 175                 return $this->mRegexStart;
 176         }
 177
 178         /**
 179          * regex without the slashes and what not
 180          */
 181         function getBaseRegex() {
 182                 if ($this->mRegex == '') {
 183                         $this->initRegex();
 184                 }
 185                 return $this->mBaseRegex;
 186         }
 187
 188         /**
 189          * Returns true if the text contains the word
 190          * @return bool
 191          */
 192         function match( $text ) {
 193                 return preg_match( $this->getRegex(), $text );
 194         }
 195
 196         /**
 197          * Returns true if the text starts with the word
 198          * @return bool
 199          */
 200         function matchStart( $text ) {
 201                 return preg_match( $this->getRegexStart(), $text );
 202         }
 203
 204         /**
 205          * Returns NULL if there's no match, the value of $1 otherwise
 206          * The return code is the matched string, if there's no variable
 207          * part in the regex and the matched variable part ($1) if there
 208          * is one.
 209          */
 210         function matchVariableStartToEnd( $text ) {
 211                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 212                 if ( $matchcount == 0 ) {
 213                         return NULL;
 214                 } elseif ( count($matches) == 1 ) {
 215                         return $matches[0];
 216                 } else {
 217                         # multiple matched parts (variable match); some will be empty because of synonyms
 218                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 219                         $matches = array_values(array_filter($matches));
 220                         return $matches[1];
 221                 }
 222         }
 223
 224
 225         /**
 226          * Returns true if the text matches the word, and alters the
 227          * input string, removing all instances of the word
 228          */
 229         function matchAndRemove( &$text ) {
 230                 global $wgMagicFound;
 231                 $wgMagicFound = false;
 232                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 233                 return $wgMagicFound;
 234         }
 235
 236         function matchStartAndRemove( &$text ) {
 237                 global $wgMagicFound;
 238                 $wgMagicFound = false;
 239                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 240                 return $wgMagicFound;
 241         }
 242
 243
 244         /**
 245          * Replaces the word with something else
 246          */
 247         function replace( $replacement, $subject ) {
 248                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 249                 $this->mModified = !($res === $subject);
 250                 return $res;
 251         }
 252
 253         /**
 254          * Variable handling: {{SUBST:xxx}} style words
 255          * Calls back a function to determine what to replace xxx with
 256          * Input word must contain $1
 257          */
 258         function substituteCallback( $text, $callback ) {
 259                 $regex = $this->getVariableRegex();
 260                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 261                 $this->mModified = !($res === $text);
 262                 return $res;
 263         }
 264
 265         /**
 266          * Matches the word, where $1 is a wildcard
 267          */
 268         function getVariableRegex()     {
 269                 if ( $this->mVariableRegex == '' ) {
 270                         $this->initRegex();
 271                 }
 272                 return $this->mVariableRegex;
 273         }
 274
 275         /**
 276          * Matches the entire string, where $1 is a wildcard
 277          */
 278         function getVariableStartToEndRegex() {
 279                 if ( $this->mVariableStartToEndRegex == '' ) {
 280                         $this->initRegex();
 281                 }
 282                 return $this->mVariableStartToEndRegex;
 283         }
 284
 285         /**
 286          * Accesses the synonym list directly
 287          */
 288         function getSynonym( $i ) {
 289                 return $this->mSynonyms[$i];
 290         }
 291
 292         /**
 293          * Returns true if the last call to replace() or substituteCallback()
 294          * returned a modified text, otherwise false.
 295          */
 296         function getWasModified(){
 297                 return $this->mModified;
 298         }
 299
 300         /**
 301          * $magicarr is an associative array of (magic word ID => replacement)
 302          * This method uses the php feature to do several replacements at the same time,
 303          * thereby gaining some efficiency. The result is placed in the out variable
 304          * $result. The return value is true if something was replaced.
 305          * @static
 306          **/
 307         function replaceMultiple( $magicarr, $subject, &$result ){
 308                 $search = array();
 309                 $replace = array();
 310                 foreach( $magicarr as $id => $replacement ){
 311                         $mw = MagicWord::get( $id );
 312                         $search[] = $mw->getRegex();
 313                         $replace[] = $replacement;
 314                 }
 315
 316                 $result = preg_replace( $search, $replace, $subject );
 317                 return !($result === $subject);
 318         }
 319
 320         /**
 321          * Adds all the synonyms of this MagicWord to an array, to allow quick
 322          * lookup in a list of magic words
 323          */
 324         function addToArray( &$array, $value ) {
 325                 foreach ( $this->mSynonyms as $syn ) {
 326                         $array[$syn] = $value;
 327                 }
 328         }
 329 }
 330
 331 /**
 332  * Used in matchAndRemove()
 333  * @private
 334  **/
 335 function pregRemoveAndRecord( $match ) {
 336         global $wgMagicFound;
 337         $wgMagicFound = true;
 338         return '';
 339 }
 340
 341 ?>