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