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_MSGNW',                     13);
  28 define('MAG_NOEDITSECTION',             14);
  29 define('MAG_END',                       15);
  30 define('MAG_IMG_THUMBNAIL',             16);
  31 define('MAG_IMG_RIGHT',                 17);
  32 define('MAG_IMG_LEFT',                  18);
  33 define('MAG_IMG_NONE',                  19);
  34 define('MAG_IMG_WIDTH',                 20);
  35 define('MAG_IMG_CENTER',                21);
  36 define('MAG_INT',                       22);
  37 define('MAG_FORCETOC',                  23);
  38 define('MAG_SITENAME',                  24);
  39 define('MAG_NS',                        25);
  40 define('MAG_LOCALURL',                  26);
  41 define('MAG_LOCALURLE',                 27);
  42 define('MAG_SERVER',                    28);
  43 define('MAG_IMG_FRAMED',                29);
  44 define('MAG_PAGENAME',                  30);
  45 define('MAG_PAGENAMEE',                 31);
  46 define('MAG_NAMESPACE',                 32);
  47 define('MAG_TOC',                       33);
  48 define('MAG_GRAMMAR',                   34);
  49 define('MAG_NOTITLECONVERT',            35);
  50 define('MAG_NOCONTENTCONVERT',          36);
  51 define('MAG_CURRENTWEEK',               37);
  52 define('MAG_CURRENTDOW',                38);
  53 define('MAG_REVISIONID',                39);
  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                 # This was used for matching "$1" variables, but different uses of the feature will have
 141                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 142                 # not here. - IMSoP
 143                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 144                 $this->mBaseRegex = implode( '|', $escSyn );
 145                 $case = $this->mCaseSensitive ? '' : 'i';
 146                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 147                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 148                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 149                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 150                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 151         }
 152
 153         /**
 154          * Gets a regex representing matching the word
 155          */
 156         function getRegex() {
 157                 if ($this->mRegex == '' ) {
 158                         $this->initRegex();
 159                 }
 160                 return $this->mRegex;
 161         }
 162
 163         /**
 164          * Gets a regex matching the word, if it is at the string start
 165          */
 166         function getRegexStart() {
 167                 if ($this->mRegex == '' ) {
 168                         $this->initRegex();
 169                 }
 170                 return $this->mRegexStart;
 171         }
 172
 173         /**
 174          * regex without the slashes and what not
 175          */
 176         function getBaseRegex() {
 177                 if ($this->mRegex == '') {
 178                         $this->initRegex();
 179                 }
 180                 return $this->mBaseRegex;
 181         }
 182
 183         /**
 184          * Returns true if the text contains the word
 185          * @return bool
 186          */
 187         function match( $text ) {
 188                 return preg_match( $this->getRegex(), $text );
 189         }
 190
 191         /**
 192          * Returns true if the text starts with the word
 193          * @return bool
 194          */
 195         function matchStart( $text ) {
 196                 return preg_match( $this->getRegexStart(), $text );
 197         }
 198
 199         /**
 200          * Returns NULL if there's no match, the value of $1 otherwise
 201          * The return code is the matched string, if there's no variable
 202          * part in the regex and the matched variable part ($1) if there
 203          * is one.
 204          */
 205         function matchVariableStartToEnd( $text ) {
 206                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 207                 if ( $matchcount == 0 ) {
 208                         return NULL;
 209                 } elseif ( count($matches) == 1 ) {
 210                         return $matches[0];
 211                 } else {
 212                         # multiple matched parts (variable match); some will be empty because of synonyms
 213                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 214                         $matches = array_values(array_filter($matches));
 215                         return $matches[1];
 216                 }
 217         }
 218
 219
 220         /**
 221          * Returns true if the text matches the word, and alters the
 222          * input string, removing all instances of the word
 223          */
 224         function matchAndRemove( &$text ) {
 225                 global $wgMagicFound;
 226                 $wgMagicFound = false;
 227                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 228                 return $wgMagicFound;
 229         }
 230
 231         function matchStartAndRemove( &$text ) {
 232                 global $wgMagicFound;
 233                 $wgMagicFound = false;
 234                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 235                 return $wgMagicFound;
 236         }
 237
 238
 239         /**
 240          * Replaces the word with something else
 241          */
 242         function replace( $replacement, $subject ) {
 243                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 244                 $this->mModified = !($res === $subject);
 245                 return $res;
 246         }
 247
 248         /**
 249          * Variable handling: {{SUBST:xxx}} style words
 250          * Calls back a function to determine what to replace xxx with
 251          * Input word must contain $1
 252          */
 253         function substituteCallback( $text, $callback ) {
 254                 $regex = $this->getVariableRegex();
 255                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 256                 $this->mModified = !($res === $text);
 257                 return $res;
 258         }
 259
 260         /**
 261          * Matches the word, where $1 is a wildcard
 262          */
 263         function getVariableRegex()     {
 264                 if ( $this->mVariableRegex == '' ) {
 265                         $this->initRegex();
 266                 }
 267                 return $this->mVariableRegex;
 268         }
 269
 270         /**
 271          * Matches the entire string, where $1 is a wildcard
 272          */
 273         function getVariableStartToEndRegex() {
 274                 if ( $this->mVariableStartToEndRegex == '' ) {
 275                         $this->initRegex();
 276                 }
 277                 return $this->mVariableStartToEndRegex;
 278         }
 279
 280         /**
 281          * Accesses the synonym list directly
 282          */
 283         function getSynonym( $i ) {
 284                 return $this->mSynonyms[$i];
 285         }
 286
 287         /**
 288          * Returns true if the last call to replace() or substituteCallback()
 289          * returned a modified text, otherwise false.
 290          */
 291         function getWasModified(){
 292                 return $this->mModified;
 293         }
 294
 295         /**
 296          * $magicarr is an associative array of (magic word ID => replacement)
 297          * This method uses the php feature to do several replacements at the same time,
 298          * thereby gaining some efficiency. The result is placed in the out variable
 299          * $result. The return value is true if something was replaced.
 300          * @static
 301          **/
 302         function replaceMultiple( $magicarr, $subject, &$result ){
 303                 $search = array();
 304                 $replace = array();
 305                 foreach( $magicarr as $id => $replacement ){
 306                         $mw = MagicWord::get( $id );
 307                         $search[] = $mw->getRegex();
 308                         $replace[] = $replacement;
 309                 }
 310
 311                 $result = preg_replace( $search, $replace, $subject );
 312                 return !($result === $subject);
 313         }
 314
 315         /**
 316          * Adds all the synonyms of this MagicWord to an array, to allow quick
 317          * lookup in a list of magic words
 318          */
 319         function addToArray( &$array, $value ) {
 320                 foreach ( $this->mSynonyms as $syn ) {
 321                         $array[$syn] = $value;
 322                 }
 323         }
 324 }
 325
 326 /**
 327  * Used in matchAndRemove()
 328  * @private
 329  **/
 330 function pregRemoveAndRecord( $match ) {
 331         global $wgMagicFound;
 332         $wgMagicFound = true;
 333         return '';
 334 }
 335
 336 ?>