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 /**
  14  * To add magic words in an extension, use the LanguageGetMagic hook. For
  15  * magic words which are also Parser variables, add a MagicWordwgVariableIDs
  16  * hook. Use string keys. Do use the interface immediately below,
  17  * MagicWordMagicWords. It is deprecated.
  18  */
  19
  20 $magicWords = array();
  21 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  22         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
  23
  24 for ( $i = 0; $i < count( $magicWords ); ++$i )
  25         define( $magicWords[$i], $magicWords[$i] );
  26
  27
  28 $wgVariableIDs = array(
  29         'currentmonth',
  30         'currentmonthname',
  31         'currentmonthnamegen',
  32         'currentmonthabbrev',
  33         'currentday',
  34         'currentday2',
  35         'currentdayname',
  36         'currentyear',
  37         'currenttime',
  38         'numberofarticles',
  39         'numberoffiles',
  40         'sitename',
  41         'server',
  42         'servername',
  43         'scriptpath',
  44         'pagename',
  45         'pagenamee',
  46         'fullpagename',
  47         'fullpagenamee',
  48         'namespace',
  49         'namespacee',
  50         'currentweek',
  51         'currentdow',
  52         'revisionid',
  53         'subpagename',
  54         'subpagenamee',
  55         'displaytitle',
  56         'talkspace',
  57         'talkspacee',
  58         'subjectspace',
  59         'subjectspacee',
  60         'talkpagename',
  61         'talkpagenamee',
  62         'subjectpagename',
  63         'subjectpagenamee',
  64         'numberofusers',
  65         'rawsuffix',
  66         'newsectionlink',
  67         'numberofpages',
  68         'currentversion',
  69         'basepagename',
  70         'basepagenamee',
  71         'urlencode',
  72         'currenttimestamp',
  73         'directionmark',
  74         'language',
  75         'contentlanguage',
  76         'pagesinnamespace',
  77         'numberofadmins',
  78 );
  79 if ( ! defined( 'MEDIAWIKI_INSTALL' ) )
  80         wfRunHooks( 'MagicWordwgVariableIDs', array( &$wgVariableIDs ) );
  81
  82 /**
  83  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  84  * Usage:
  85  *     if (MagicWord::get( 'redirect' )->match( $text ) )
  86  *
  87  * Possible future improvements:
  88  *   * Simultaneous searching for a number of magic words
  89  *   * $wgMagicWords in shared memory
  90  *
  91  * Please avoid reading the data out of one of these objects and then writing
  92  * special case code. If possible, add another match()-like function here.
  93  *
  94  * @package MediaWiki
  95  */
  96 class MagicWord {
  97         /**#@+
  98          * @private
  99          */
 100         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 101         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 102         var $mModified;
 103         /**#@-*/
 104
 105         function MagicWord($id = 0, $syn = '', $cs = false) {
 106                 $this->mId = $id;
 107                 $this->mSynonyms = (array)$syn;
 108                 $this->mCaseSensitive = $cs;
 109                 $this->mRegex = '';
 110                 $this->mRegexStart = '';
 111                 $this->mVariableRegex = '';
 112                 $this->mVariableStartToEndRegex = '';
 113                 $this->mModified = false;
 114         }
 115
 116         /**
 117          * Factory: creates an object representing an ID
 118          * @static
 119          */
 120         static function &get( $id ) {
 121                 global $wgMagicWords;
 122
 123                 if ( !is_array( $wgMagicWords ) ) {
 124                         throw new MWException( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 125                 }
 126                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 127                         $mw = new MagicWord();
 128                         $mw->load( $id );
 129                         $wgMagicWords[$id] = $mw;
 130                 }
 131                 return $wgMagicWords[$id];
 132         }
 133
 134         # Initialises this object with an ID
 135         function load( $id ) {
 136                 global $wgContLang;
 137                 $this->mId = $id;
 138                 $wgContLang->getMagic( $this );
 139         }
 140
 141         /**
 142          * Preliminary initialisation
 143          * @private
 144          */
 145         function initRegex() {
 146                 #$variableClass = Title::legalChars();
 147                 # This was used for matching "$1" variables, but different uses of the feature will have
 148                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 149                 # not here. - IMSoP
 150
 151                 $escSyn = array();
 152                 foreach ( $this->mSynonyms as $synonym )
 153                         // In case a magic word contains /, like that's going to happen;)
 154                         $escSyn[] = preg_quote( $synonym, '/' );
 155                 $this->mBaseRegex = implode( '|', $escSyn );
 156
 157                 $case = $this->mCaseSensitive ? '' : 'i';
 158                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 159                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 160                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 161                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 162                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 163         }
 164
 165         /**
 166          * Gets a regex representing matching the word
 167          */
 168         function getRegex() {
 169                 if ($this->mRegex == '' ) {
 170                         $this->initRegex();
 171                 }
 172                 return $this->mRegex;
 173         }
 174
 175         /**
 176          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 177          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 178          * the complete expression
 179          */
 180         function getRegexCase() {
 181                 if ( $this->mRegex === '' )
 182                         $this->initRegex();
 183
 184                 return $this->mCaseSensitive ? '' : 'i';
 185         }
 186
 187         /**
 188          * Gets a regex matching the word, if it is at the string start
 189          */
 190         function getRegexStart() {
 191                 if ($this->mRegex == '' ) {
 192                         $this->initRegex();
 193                 }
 194                 return $this->mRegexStart;
 195         }
 196
 197         /**
 198          * regex without the slashes and what not
 199          */
 200         function getBaseRegex() {
 201                 if ($this->mRegex == '') {
 202                         $this->initRegex();
 203                 }
 204                 return $this->mBaseRegex;
 205         }
 206
 207         /**
 208          * Returns true if the text contains the word
 209          * @return bool
 210          */
 211         function match( $text ) {
 212                 return preg_match( $this->getRegex(), $text );
 213         }
 214
 215         /**
 216          * Returns true if the text starts with the word
 217          * @return bool
 218          */
 219         function matchStart( $text ) {
 220                 return preg_match( $this->getRegexStart(), $text );
 221         }
 222
 223         /**
 224          * Returns NULL if there's no match, the value of $1 otherwise
 225          * The return code is the matched string, if there's no variable
 226          * part in the regex and the matched variable part ($1) if there
 227          * is one.
 228          */
 229         function matchVariableStartToEnd( $text ) {
 230                 $matches = array();
 231                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 232                 if ( $matchcount == 0 ) {
 233                         return NULL;
 234                 } else {
 235                         # multiple matched parts (variable match); some will be empty because of
 236                         # synonyms. The variable will be the second non-empty one so remove any
 237                         # blank elements and re-sort the indices.
 238                         # See also bug 6526
 239
 240                         $matches = array_values(array_filter($matches));
 241
 242                         if ( count($matches) == 1 ) { return $matches[0]; }
 243                         else { return $matches[1]; }
 244                 }
 245         }
 246
 247
 248         /**
 249          * Returns true if the text matches the word, and alters the
 250          * input string, removing all instances of the word
 251          */
 252         function matchAndRemove( &$text ) {
 253                 global $wgMagicFound;
 254                 $wgMagicFound = false;
 255                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 256                 return $wgMagicFound;
 257         }
 258
 259         function matchStartAndRemove( &$text ) {
 260                 global $wgMagicFound;
 261                 $wgMagicFound = false;
 262                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 263                 return $wgMagicFound;
 264         }
 265
 266
 267         /**
 268          * Replaces the word with something else
 269          */
 270         function replace( $replacement, $subject, $limit=-1 ) {
 271                 $res = preg_replace( $this->getRegex(), wfRegexReplacement( $replacement ), $subject, $limit );
 272                 $this->mModified = !($res === $subject);
 273                 return $res;
 274         }
 275
 276         /**
 277          * Variable handling: {{SUBST:xxx}} style words
 278          * Calls back a function to determine what to replace xxx with
 279          * Input word must contain $1
 280          */
 281         function substituteCallback( $text, $callback ) {
 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         function getSynonyms() {
 315                 return $this->mSynonyms;
 316         }
 317
 318         /**
 319          * Returns true if the last call to replace() or substituteCallback()
 320          * returned a modified text, otherwise false.
 321          */
 322         function getWasModified(){
 323                 return $this->mModified;
 324         }
 325
 326         /**
 327          * $magicarr is an associative array of (magic word ID => replacement)
 328          * This method uses the php feature to do several replacements at the same time,
 329          * thereby gaining some efficiency. The result is placed in the out variable
 330          * $result. The return value is true if something was replaced.
 331          * @static
 332          **/
 333         function replaceMultiple( $magicarr, $subject, &$result ){
 334                 $search = array();
 335                 $replace = array();
 336                 foreach( $magicarr as $id => $replacement ){
 337                         $mw = MagicWord::get( $id );
 338                         $search[] = $mw->getRegex();
 339                         $replace[] = $replacement;
 340                 }
 341
 342                 $result = preg_replace( $search, $replace, $subject );
 343                 return !($result === $subject);
 344         }
 345
 346         /**
 347          * Adds all the synonyms of this MagicWord to an array, to allow quick
 348          * lookup in a list of magic words
 349          */
 350         function addToArray( &$array, $value ) {
 351                 foreach ( $this->mSynonyms as $syn ) {
 352                         $array[$syn] = $value;
 353                 }
 354         }
 355
 356         function isCaseSensitive() {
 357                 return $this->mCaseSensitive;
 358         }
 359 }
 360
 361 /**
 362  * Used in matchAndRemove()
 363  * @private
 364  **/
 365 function pregRemoveAndRecord( $match ) {
 366         global $wgMagicFound;
 367         $wgMagicFound = true;
 368         return '';
 369 }
 370
 371 ?>