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