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