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