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