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