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
  15 $magicWords = array(
  16         'MAG_REDIRECT',
  17         'MAG_NOTOC',
  18         'MAG_START',
  19         'MAG_CURRENTMONTH',
  20         'MAG_CURRENTMONTHNAME',
  21         'MAG_CURRENTMONTHNAMEGEN',
  22         'MAG_CURRENTMONTHABBREV',
  23         'MAG_CURRENTDAY',
  24         'MAG_CURRENTDAY2',
  25         'MAG_CURRENTDAYNAME',
  26         'MAG_CURRENTYEAR',
  27         'MAG_CURRENTTIME',
  28         'MAG_NUMBEROFARTICLES',
  29         'MAG_SUBST',
  30         'MAG_MSG',
  31         'MAG_MSGNW',
  32         'MAG_NOEDITSECTION',
  33         'MAG_END',
  34         'MAG_IMG_THUMBNAIL',
  35         'MAG_IMG_RIGHT',
  36         'MAG_IMG_LEFT',
  37         'MAG_IMG_NONE',
  38         'MAG_IMG_WIDTH',
  39         'MAG_IMG_CENTER',
  40         'MAG_INT',
  41         'MAG_FORCETOC',
  42         'MAG_SITENAME',
  43         'MAG_NS',
  44         'MAG_LOCALURL',
  45         'MAG_LOCALURLE',
  46         'MAG_SERVER',
  47         'MAG_IMG_FRAMED',
  48         'MAG_PAGENAME',
  49         'MAG_PAGENAMEE',
  50         'MAG_NAMESPACE',
  51         'MAG_NAMESPACEE',
  52         'MAG_TOC',
  53         'MAG_GRAMMAR',
  54         'MAG_NOTITLECONVERT',
  55         'MAG_NOCONTENTCONVERT',
  56         'MAG_CURRENTWEEK',
  57         'MAG_CURRENTDOW',
  58         'MAG_REVISIONID',
  59         'MAG_SCRIPTPATH',
  60         'MAG_SERVERNAME',
  61         'MAG_NUMBEROFFILES',
  62         'MAG_IMG_MANUALTHUMB',
  63         'MAG_PLURAL',
  64         'MAG_FULLURL',
  65         'MAG_FULLURLE',
  66         'MAG_LCFIRST',
  67         'MAG_UCFIRST',
  68         'MAG_LC',
  69         'MAG_UC',
  70         'MAG_FULLPAGENAME',
  71         'MAG_FULLPAGENAMEE',
  72         'MAG_RAW',
  73         'MAG_SUBPAGENAME',
  74         'MAG_SUBPAGENAMEE',
  75         'MAG_DISPLAYTITLE',
  76         'MAG_TALKSPACE',
  77         'MAG_TALKSPACEE',
  78         'MAG_SUBJECTSPACE',
  79         'MAG_SUBJECTSPACEE',
  80         'MAG_TALKPAGENAME',
  81         'MAG_TALKPAGENAMEE',
  82         'MAG_SUBJECTPAGENAME',
  83         'MAG_SUBJECTPAGENAMEE',
  84         'MAG_NUMBEROFUSERS',
  85 );
  86 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  87         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  88
  89 for ( $i = 0; $i < count( $magicWords ); ++$i )
  90         define( $magicWords[$i], $i );
  91
  92 $wgVariableIDs = array(
  93         MAG_CURRENTMONTH,
  94         MAG_CURRENTMONTHNAME,
  95         MAG_CURRENTMONTHNAMEGEN,
  96         MAG_CURRENTMONTHABBREV,
  97         MAG_CURRENTDAY,
  98         MAG_CURRENTDAY2,
  99         MAG_CURRENTDAYNAME,
 100         MAG_CURRENTYEAR,
 101         MAG_CURRENTTIME,
 102         MAG_NUMBEROFARTICLES,
 103         MAG_NUMBEROFFILES,
 104         MAG_SITENAME,
 105         MAG_SERVER,
 106         MAG_SERVERNAME,
 107         MAG_SCRIPTPATH,
 108         MAG_PAGENAME,
 109         MAG_PAGENAMEE,
 110         MAG_FULLPAGENAME,
 111         MAG_FULLPAGENAMEE,
 112         MAG_NAMESPACE,
 113         MAG_NAMESPACEE,
 114         MAG_CURRENTWEEK,
 115         MAG_CURRENTDOW,
 116         MAG_REVISIONID,
 117         MAG_SUBPAGENAME,
 118         MAG_SUBPAGENAMEE,
 119         MAG_DISPLAYTITLE,
 120         MAG_TALKSPACE,
 121         MAG_TALKSPACEE,
 122         MAG_SUBJECTSPACE,
 123         MAG_SUBJECTSPACEE,
 124         MAG_TALKPAGENAME,
 125         MAG_TALKPAGENAMEE,
 126         MAG_SUBJECTPAGENAME,
 127         MAG_SUBJECTPAGENAMEE,
 128         MAG_NUMBEROFUSERS,
 129 );
 130 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 131         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 132
 133 /**
 134  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 135  * Usage:
 136  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 137  *
 138  * Possible future improvements:
 139  *   * Simultaneous searching for a number of magic words
 140  *   * $wgMagicWords in shared memory
 141  *
 142  * Please avoid reading the data out of one of these objects and then writing
 143  * special case code. If possible, add another match()-like function here.
 144  *
 145  * @package MediaWiki
 146  */
 147 class MagicWord {
 148         /**#@+
 149          * @access private
 150          */
 151         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 152         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 153         var $mModified;
 154         /**#@-*/
 155
 156         function MagicWord($id = 0, $syn = '', $cs = false) {
 157                 $this->mId = $id;
 158                 $this->mSynonyms = (array)$syn;
 159                 $this->mCaseSensitive = $cs;
 160                 $this->mRegex = '';
 161                 $this->mRegexStart = '';
 162                 $this->mVariableRegex = '';
 163                 $this->mVariableStartToEndRegex = '';
 164                 $this->mModified = false;
 165         }
 166
 167         /**
 168          * Factory: creates an object representing an ID
 169          * @static
 170          */
 171         function &get( $id ) {
 172                 global $wgMagicWords;
 173
 174                 if ( !is_array( $wgMagicWords ) ) {
 175                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 176                 }
 177                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 178                         $mw = new MagicWord();
 179                         $mw->load( $id );
 180                         $wgMagicWords[$id] = $mw;
 181                 }
 182                 return $wgMagicWords[$id];
 183         }
 184
 185         # Initialises this object with an ID
 186         function load( $id ) {
 187                 global $wgContLang;
 188                 $this->mId = $id;
 189                 $wgContLang->getMagic( $this );
 190         }
 191
 192         /**
 193          * Preliminary initialisation
 194          * @access private
 195          */
 196         function initRegex() {
 197                 #$variableClass = Title::legalChars();
 198                 # This was used for matching "$1" variables, but different uses of the feature will have
 199                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 200                 # not here. - IMSoP
 201
 202                 $escSyn = array();
 203                 foreach ( $this->mSynonyms as $synonym )
 204                         // In case a magic word contains /, like that's going to happen;)
 205                         $escSyn[] = preg_quote( $synonym, '/' );
 206                 $this->mBaseRegex = implode( '|', $escSyn );
 207
 208                 $case = $this->mCaseSensitive ? '' : 'i';
 209                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 210                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 211                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 212                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 213                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 214         }
 215
 216         /**
 217          * Gets a regex representing matching the word
 218          */
 219         function getRegex() {
 220                 if ($this->mRegex == '' ) {
 221                         $this->initRegex();
 222                 }
 223                 return $this->mRegex;
 224         }
 225
 226         /**
 227          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 228          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 229          * the complete expression
 230          */
 231         function getRegexCase() {
 232                 if ( $this->mRegex === '' )
 233                         $this->initRegex();
 234
 235                 return $this->mCaseSensitive ? '' : 'i';
 236         }
 237
 238         /**
 239          * Gets a regex matching the word, if it is at the string start
 240          */
 241         function getRegexStart() {
 242                 if ($this->mRegex == '' ) {
 243                         $this->initRegex();
 244                 }
 245                 return $this->mRegexStart;
 246         }
 247
 248         /**
 249          * regex without the slashes and what not
 250          */
 251         function getBaseRegex() {
 252                 if ($this->mRegex == '') {
 253                         $this->initRegex();
 254                 }
 255                 return $this->mBaseRegex;
 256         }
 257
 258         /**
 259          * Returns true if the text contains the word
 260          * @return bool
 261          */
 262         function match( $text ) {
 263                 return preg_match( $this->getRegex(), $text );
 264         }
 265
 266         /**
 267          * Returns true if the text starts with the word
 268          * @return bool
 269          */
 270         function matchStart( $text ) {
 271                 return preg_match( $this->getRegexStart(), $text );
 272         }
 273
 274         /**
 275          * Returns NULL if there's no match, the value of $1 otherwise
 276          * The return code is the matched string, if there's no variable
 277          * part in the regex and the matched variable part ($1) if there
 278          * is one.
 279          */
 280         function matchVariableStartToEnd( $text ) {
 281                 $matches = array();
 282                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 283                 if ( $matchcount == 0 ) {
 284                         return NULL;
 285                 } elseif ( count($matches) == 1 ) {
 286                         return $matches[0];
 287                 } else {
 288                         # multiple matched parts (variable match); some will be empty because of synonyms
 289                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 290                         $matches = array_values(array_filter($matches));
 291                         return $matches[1];
 292                 }
 293         }
 294
 295
 296         /**
 297          * Returns true if the text matches the word, and alters the
 298          * input string, removing all instances of the word
 299          */
 300         function matchAndRemove( &$text ) {
 301                 global $wgMagicFound;
 302                 $wgMagicFound = false;
 303                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 304                 return $wgMagicFound;
 305         }
 306
 307         function matchStartAndRemove( &$text ) {
 308                 global $wgMagicFound;
 309                 $wgMagicFound = false;
 310                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 311                 return $wgMagicFound;
 312         }
 313
 314
 315         /**
 316          * Replaces the word with something else
 317          */
 318         function replace( $replacement, $subject ) {
 319                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject );
 320                 $this->mModified = !($res === $subject);
 321                 return $res;
 322         }
 323
 324         /**
 325          * Variable handling: {{SUBST:xxx}} style words
 326          * Calls back a function to determine what to replace xxx with
 327          * Input word must contain $1
 328          */
 329         function substituteCallback( $text, $callback ) {
 330                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 331                 $this->mModified = !($res === $text);
 332                 return $res;
 333         }
 334
 335         /**
 336          * Matches the word, where $1 is a wildcard
 337          */
 338         function getVariableRegex()     {
 339                 if ( $this->mVariableRegex == '' ) {
 340                         $this->initRegex();
 341                 }
 342                 return $this->mVariableRegex;
 343         }
 344
 345         /**
 346          * Matches the entire string, where $1 is a wildcard
 347          */
 348         function getVariableStartToEndRegex() {
 349                 if ( $this->mVariableStartToEndRegex == '' ) {
 350                         $this->initRegex();
 351                 }
 352                 return $this->mVariableStartToEndRegex;
 353         }
 354
 355         /**
 356          * Accesses the synonym list directly
 357          */
 358         function getSynonym( $i ) {
 359                 return $this->mSynonyms[$i];
 360         }
 361
 362         /**
 363          * Returns true if the last call to replace() or substituteCallback()
 364          * returned a modified text, otherwise false.
 365          */
 366         function getWasModified(){
 367                 return $this->mModified;
 368         }
 369
 370         /**
 371          * $magicarr is an associative array of (magic word ID => replacement)
 372          * This method uses the php feature to do several replacements at the same time,
 373          * thereby gaining some efficiency. The result is placed in the out variable
 374          * $result. The return value is true if something was replaced.
 375          * @static
 376          **/
 377         function replaceMultiple( $magicarr, $subject, &$result ){
 378                 $search = array();
 379                 $replace = array();
 380                 foreach( $magicarr as $id => $replacement ){
 381                         $mw = MagicWord::get( $id );
 382                         $search[] = $mw->getRegex();
 383                         $replace[] = $replacement;
 384                 }
 385
 386                 $result = preg_replace( $search, $replace, $subject );
 387                 return !($result === $subject);
 388         }
 389
 390         /**
 391          * Adds all the synonyms of this MagicWord to an array, to allow quick
 392          * lookup in a list of magic words
 393          */
 394         function addToArray( &$array, $value ) {
 395                 foreach ( $this->mSynonyms as $syn ) {
 396                         $array[$syn] = $value;
 397                 }
 398         }
 399 }
 400
 401 /**
 402  * Used in matchAndRemove()
 403  * @access private
 404  **/
 405 function pregRemoveAndRecord( $match ) {
 406         global $wgMagicFound;
 407         $wgMagicFound = true;
 408         return '';
 409 }
 410
 411 ?>