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