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         'MAG_RAWSUFFIX',
  86         'MAG_NEWSECTIONLINK',
  87         'MAG_NUMBEROFPAGES',
  88         'MAG_CURRENTVERSION',
  89         'MAG_BASEPAGENAME',
  90         'MAG_BASEPAGENAMEE',
  91         'MAG_URLENCODE',
  92         'MAG_CURRENTTIMESTAMP',
  93 );
  94 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  95         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  96
  97 for ( $i = 0; $i < count( $magicWords ); ++$i )
  98         define( $magicWords[$i], $i );
  99
 100 $wgVariableIDs = array(
 101         MAG_CURRENTMONTH,
 102         MAG_CURRENTMONTHNAME,
 103         MAG_CURRENTMONTHNAMEGEN,
 104         MAG_CURRENTMONTHABBREV,
 105         MAG_CURRENTDAY,
 106         MAG_CURRENTDAY2,
 107         MAG_CURRENTDAYNAME,
 108         MAG_CURRENTYEAR,
 109         MAG_CURRENTTIME,
 110         MAG_NUMBEROFARTICLES,
 111         MAG_NUMBEROFFILES,
 112         MAG_SITENAME,
 113         MAG_SERVER,
 114         MAG_SERVERNAME,
 115         MAG_SCRIPTPATH,
 116         MAG_PAGENAME,
 117         MAG_PAGENAMEE,
 118         MAG_FULLPAGENAME,
 119         MAG_FULLPAGENAMEE,
 120         MAG_NAMESPACE,
 121         MAG_NAMESPACEE,
 122         MAG_CURRENTWEEK,
 123         MAG_CURRENTDOW,
 124         MAG_REVISIONID,
 125         MAG_SUBPAGENAME,
 126         MAG_SUBPAGENAMEE,
 127         MAG_DISPLAYTITLE,
 128         MAG_TALKSPACE,
 129         MAG_TALKSPACEE,
 130         MAG_SUBJECTSPACE,
 131         MAG_SUBJECTSPACEE,
 132         MAG_TALKPAGENAME,
 133         MAG_TALKPAGENAMEE,
 134         MAG_SUBJECTPAGENAME,
 135         MAG_SUBJECTPAGENAMEE,
 136         MAG_NUMBEROFUSERS,
 137         MAG_RAWSUFFIX,
 138         MAG_NEWSECTIONLINK,
 139         MAG_NUMBEROFPAGES,
 140         MAG_CURRENTVERSION,
 141         MAG_BASEPAGENAME,
 142         MAG_BASEPAGENAMEE,
 143         MAG_URLENCODE,
 144         MAG_CURRENTTIMESTAMP,
 145 );
 146 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
 147         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
 148
 149 /**
 150  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
 151  * Usage:
 152  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
 153  *
 154  * Possible future improvements:
 155  *   * Simultaneous searching for a number of magic words
 156  *   * $wgMagicWords in shared memory
 157  *
 158  * Please avoid reading the data out of one of these objects and then writing
 159  * special case code. If possible, add another match()-like function here.
 160  *
 161  * @package MediaWiki
 162  */
 163 class MagicWord {
 164         /**#@+
 165          * @access private
 166          */
 167         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 168         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 169         var $mModified;
 170         /**#@-*/
 171
 172         function MagicWord($id = 0, $syn = '', $cs = false) {
 173                 $this->mId = $id;
 174                 $this->mSynonyms = (array)$syn;
 175                 $this->mCaseSensitive = $cs;
 176                 $this->mRegex = '';
 177                 $this->mRegexStart = '';
 178                 $this->mVariableRegex = '';
 179                 $this->mVariableStartToEndRegex = '';
 180                 $this->mModified = false;
 181         }
 182
 183         /**
 184          * Factory: creates an object representing an ID
 185          * @static
 186          */
 187         function &get( $id ) {
 188                 global $wgMagicWords;
 189
 190                 if ( !is_array( $wgMagicWords ) ) {
 191                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 192                 }
 193                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 194                         $mw = new MagicWord();
 195                         $mw->load( $id );
 196                         $wgMagicWords[$id] = $mw;
 197                 }
 198                 return $wgMagicWords[$id];
 199         }
 200
 201         # Initialises this object with an ID
 202         function load( $id ) {
 203                 global $wgContLang;
 204                 $this->mId = $id;
 205                 $wgContLang->getMagic( $this );
 206         }
 207
 208         /**
 209          * Preliminary initialisation
 210          * @access private
 211          */
 212         function initRegex() {
 213                 #$variableClass = Title::legalChars();
 214                 # This was used for matching "$1" variables, but different uses of the feature will have
 215                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 216                 # not here. - IMSoP
 217
 218                 $escSyn = array();
 219                 foreach ( $this->mSynonyms as $synonym )
 220                         // In case a magic word contains /, like that's going to happen;)
 221                         $escSyn[] = preg_quote( $synonym, '/' );
 222                 $this->mBaseRegex = implode( '|', $escSyn );
 223
 224                 $case = $this->mCaseSensitive ? '' : 'i';
 225                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 226                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 227                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 228                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 229                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 230         }
 231
 232         /**
 233          * Gets a regex representing matching the word
 234          */
 235         function getRegex() {
 236                 if ($this->mRegex == '' ) {
 237                         $this->initRegex();
 238                 }
 239                 return $this->mRegex;
 240         }
 241
 242         /**
 243          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 244          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 245          * the complete expression
 246          */
 247         function getRegexCase() {
 248                 if ( $this->mRegex === '' )
 249                         $this->initRegex();
 250
 251                 return $this->mCaseSensitive ? '' : 'i';
 252         }
 253
 254         /**
 255          * Gets a regex matching the word, if it is at the string start
 256          */
 257         function getRegexStart() {
 258                 if ($this->mRegex == '' ) {
 259                         $this->initRegex();
 260                 }
 261                 return $this->mRegexStart;
 262         }
 263
 264         /**
 265          * regex without the slashes and what not
 266          */
 267         function getBaseRegex() {
 268                 if ($this->mRegex == '') {
 269                         $this->initRegex();
 270                 }
 271                 return $this->mBaseRegex;
 272         }
 273
 274         /**
 275          * Returns true if the text contains the word
 276          * @return bool
 277          */
 278         function match( $text ) {
 279                 return preg_match( $this->getRegex(), $text );
 280         }
 281
 282         /**
 283          * Returns true if the text starts with the word
 284          * @return bool
 285          */
 286         function matchStart( $text ) {
 287                 return preg_match( $this->getRegexStart(), $text );
 288         }
 289
 290         /**
 291          * Returns NULL if there's no match, the value of $1 otherwise
 292          * The return code is the matched string, if there's no variable
 293          * part in the regex and the matched variable part ($1) if there
 294          * is one.
 295          */
 296         function matchVariableStartToEnd( $text ) {
 297                 $matches = array();
 298                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 299                 if ( $matchcount == 0 ) {
 300                         return NULL;
 301                 } elseif ( count($matches) == 1 ) {
 302                         return $matches[0];
 303                 } else {
 304                         # multiple matched parts (variable match); some will be empty because of synonyms
 305                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 306                         $matches = array_values(array_filter($matches));
 307                         return $matches[1];
 308                 }
 309         }
 310
 311
 312         /**
 313          * Returns true if the text matches the word, and alters the
 314          * input string, removing all instances of the word
 315          */
 316         function matchAndRemove( &$text ) {
 317                 global $wgMagicFound;
 318                 $wgMagicFound = false;
 319                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 320                 return $wgMagicFound;
 321         }
 322
 323         function matchStartAndRemove( &$text ) {
 324                 global $wgMagicFound;
 325                 $wgMagicFound = false;
 326                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 327                 return $wgMagicFound;
 328         }
 329
 330
 331         /**
 332          * Replaces the word with something else
 333          */
 334         function replace( $replacement, $subject, $limit=-1 ) {
 335                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject, $limit );
 336                 $this->mModified = !($res === $subject);
 337                 return $res;
 338         }
 339
 340         /**
 341          * Variable handling: {{SUBST:xxx}} style words
 342          * Calls back a function to determine what to replace xxx with
 343          * Input word must contain $1
 344          */
 345         function substituteCallback( $text, $callback ) {
 346                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 347                 $this->mModified = !($res === $text);
 348                 return $res;
 349         }
 350
 351         /**
 352          * Matches the word, where $1 is a wildcard
 353          */
 354         function getVariableRegex()     {
 355                 if ( $this->mVariableRegex == '' ) {
 356                         $this->initRegex();
 357                 }
 358                 return $this->mVariableRegex;
 359         }
 360
 361         /**
 362          * Matches the entire string, where $1 is a wildcard
 363          */
 364         function getVariableStartToEndRegex() {
 365                 if ( $this->mVariableStartToEndRegex == '' ) {
 366                         $this->initRegex();
 367                 }
 368                 return $this->mVariableStartToEndRegex;
 369         }
 370
 371         /**
 372          * Accesses the synonym list directly
 373          */
 374         function getSynonym( $i ) {
 375                 return $this->mSynonyms[$i];
 376         }
 377
 378         /**
 379          * Returns true if the last call to replace() or substituteCallback()
 380          * returned a modified text, otherwise false.
 381          */
 382         function getWasModified(){
 383                 return $this->mModified;
 384         }
 385
 386         /**
 387          * $magicarr is an associative array of (magic word ID => replacement)
 388          * This method uses the php feature to do several replacements at the same time,
 389          * thereby gaining some efficiency. The result is placed in the out variable
 390          * $result. The return value is true if something was replaced.
 391          * @static
 392          **/
 393         function replaceMultiple( $magicarr, $subject, &$result ){
 394                 $search = array();
 395                 $replace = array();
 396                 foreach( $magicarr as $id => $replacement ){
 397                         $mw = MagicWord::get( $id );
 398                         $search[] = $mw->getRegex();
 399                         $replace[] = $replacement;
 400                 }
 401
 402                 $result = preg_replace( $search, $replace, $subject );
 403                 return !($result === $subject);
 404         }
 405
 406         /**
 407          * Adds all the synonyms of this MagicWord to an array, to allow quick
 408          * lookup in a list of magic words
 409          */
 410         function addToArray( &$array, $value ) {
 411                 foreach ( $this->mSynonyms as $syn ) {
 412                         $array[$syn] = $value;
 413                 }
 414         }
 415 }
 416
 417 /**
 418  * Used in matchAndRemove()
 419  * @access private
 420  **/
 421 function pregRemoveAndRecord( $match ) {
 422         global $wgMagicFound;
 423         $wgMagicFound = true;
 424         return '';
 425 }
 426
 427 ?>