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 define('MAG_REDIRECT',                  0);
  15 define('MAG_NOTOC',                     1);
  16 define('MAG_START',                     2);
  17 define('MAG_CURRENTMONTH',              3);
  18 define('MAG_CURRENTMONTHNAME',          4);
  19 define('MAG_CURRENTMONTHNAMEGEN',       5);
  20 define('MAG_CURRENTMONTHABBREV',        6);
  21 define('MAG_CURRENTDAY',                7);
  22 define('MAG_CURRENTDAYNAME',            8);
  23 define('MAG_CURRENTYEAR',               9);
  24 define('MAG_CURRENTTIME',               10);
  25 define('MAG_NUMBEROFARTICLES',          11);
  26 define('MAG_SUBST',                     12);
  27 define('MAG_MSG',                       13);
  28 define('MAG_MSGNW',                     14);
  29 define('MAG_NOEDITSECTION',             15);
  30 define('MAG_END',                       16);
  31 define('MAG_IMG_THUMBNAIL',             17);
  32 define('MAG_IMG_RIGHT',                 18);
  33 define('MAG_IMG_LEFT',                  19);
  34 define('MAG_IMG_NONE',                  20);
  35 define('MAG_IMG_WIDTH',                 21);
  36 define('MAG_IMG_CENTER',                22);
  37 define('MAG_INT',                       23);
  38 define('MAG_FORCETOC',                  24);
  39 define('MAG_SITENAME',                  25);
  40 define('MAG_NS',                        26);
  41 define('MAG_LOCALURL',                  27);
  42 define('MAG_LOCALURLE',                 28);
  43 define('MAG_SERVER',                    29);
  44 define('MAG_IMG_FRAMED',                30);
  45 define('MAG_PAGENAME',                  31);
  46 define('MAG_PAGENAMEE',                 32);
  47 define('MAG_NAMESPACE',                 33);
  48 define('MAG_TOC',                       34);
  49 define('MAG_GRAMMAR',                   35);
  50 define('MAG_NOTITLECONVERT',            36);
  51 define('MAG_NOCONTENTCONVERT',          37);
  52 define('MAG_CURRENTWEEK',               38);
  53 define('MAG_CURRENTDOW',                39);
  54 define('MAG_REVISIONID',                40);
  55 define('MAG_SCRIPTPATH',                41);
  56 define('MAG_SERVERNAME',                42);
  57 define('MAG_NUMBEROFFILES',             43);
  58 define('MAG_IMG_MANUALTHUMB',           44);
  59 define('MAG_PLURAL',                    45);
  60 define('MAG_FULLURL',                   46);
  61 define('MAG_FULLURLE',                  47);
  62 define('MAG_LCFIRST',                   48);
  63 define('MAG_UCFIRST',                   49);
  64 define('MAG_LC',                        50);
  65 define('MAG_UC',                        51);
  66
  67 $wgVariableIDs = array(
  68         MAG_CURRENTMONTH,
  69         MAG_CURRENTMONTHNAME,
  70         MAG_CURRENTMONTHNAMEGEN,
  71         MAG_CURRENTMONTHABBREV,
  72         MAG_CURRENTDAY,
  73         MAG_CURRENTDAYNAME,
  74         MAG_CURRENTYEAR,
  75         MAG_CURRENTTIME,
  76         MAG_NUMBEROFARTICLES,
  77         MAG_NUMBEROFFILES,
  78         MAG_SITENAME,
  79         MAG_SERVER,
  80         MAG_SERVERNAME,
  81         MAG_SCRIPTPATH,
  82         MAG_PAGENAME,
  83         MAG_PAGENAMEE,
  84         MAG_NAMESPACE,
  85         MAG_CURRENTWEEK,
  86         MAG_CURRENTDOW,
  87         MAG_REVISIONID,
  88 );
  89
  90 /**
  91  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
  92  * Usage:
  93  *     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
  94  *
  95  * Possible future improvements:
  96  *   * Simultaneous searching for a number of magic words
  97  *   * $wgMagicWords in shared memory
  98  *
  99  * Please avoid reading the data out of one of these objects and then writing
 100  * special case code. If possible, add another match()-like function here.
 101  *
 102  * @package MediaWiki
 103  */
 104 class MagicWord {
 105         /**#@+
 106          * @access private
 107          */
 108         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
 109         var $mRegexStart, $mBaseRegex, $mVariableRegex;
 110         var $mModified;
 111         /**#@-*/
 112
 113         function MagicWord($id = 0, $syn = '', $cs = false) {
 114                 $this->mId = $id;
 115                 $this->mSynonyms = (array)$syn;
 116                 $this->mCaseSensitive = $cs;
 117                 $this->mRegex = '';
 118                 $this->mRegexStart = '';
 119                 $this->mVariableRegex = '';
 120                 $this->mVariableStartToEndRegex = '';
 121                 $this->mModified = false;
 122         }
 123
 124         /**
 125          * Factory: creates an object representing an ID
 126          * @static
 127          */
 128         function &get( $id ) {
 129                 global $wgMagicWords;
 130
 131                 if ( !is_array( $wgMagicWords ) ) {
 132                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
 133                 }
 134                 if (!array_key_exists( $id, $wgMagicWords ) ) {
 135                         $mw = new MagicWord();
 136                         $mw->load( $id );
 137                         $wgMagicWords[$id] = $mw;
 138                 }
 139                 return $wgMagicWords[$id];
 140         }
 141
 142         # Initialises this object with an ID
 143         function load( $id ) {
 144                 global $wgContLang;
 145                 $this->mId = $id;
 146                 $wgContLang->getMagic( $this );
 147         }
 148
 149         /**
 150          * Preliminary initialisation
 151          * @private
 152          */
 153         function initRegex() {
 154                 #$variableClass = Title::legalChars();
 155                 # This was used for matching "$1" variables, but different uses of the feature will have
 156                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 157                 # not here. - IMSoP
 158                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 159                 $this->mBaseRegex = implode( '|', $escSyn );
 160                 $case = $this->mCaseSensitive ? '' : 'i';
 161                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 162                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 163                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 164                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 165                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 166         }
 167
 168         /**
 169          * Gets a regex representing matching the word
 170          */
 171         function getRegex() {
 172                 if ($this->mRegex == '' ) {
 173                         $this->initRegex();
 174                 }
 175                 return $this->mRegex;
 176         }
 177
 178         /**
 179          * Gets a regex matching the word, if it is at the string start
 180          */
 181         function getRegexStart() {
 182                 if ($this->mRegex == '' ) {
 183                         $this->initRegex();
 184                 }
 185                 return $this->mRegexStart;
 186         }
 187
 188         /**
 189          * regex without the slashes and what not
 190          */
 191         function getBaseRegex() {
 192                 if ($this->mRegex == '') {
 193                         $this->initRegex();
 194                 }
 195                 return $this->mBaseRegex;
 196         }
 197
 198         /**
 199          * Returns true if the text contains the word
 200          * @return bool
 201          */
 202         function match( $text ) {
 203                 return preg_match( $this->getRegex(), $text );
 204         }
 205
 206         /**
 207          * Returns true if the text starts with the word
 208          * @return bool
 209          */
 210         function matchStart( $text ) {
 211                 return preg_match( $this->getRegexStart(), $text );
 212         }
 213
 214         /**
 215          * Returns NULL if there's no match, the value of $1 otherwise
 216          * The return code is the matched string, if there's no variable
 217          * part in the regex and the matched variable part ($1) if there
 218          * is one.
 219          */
 220         function matchVariableStartToEnd( $text ) {
 221                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 222                 if ( $matchcount == 0 ) {
 223                         return NULL;
 224                 } elseif ( count($matches) == 1 ) {
 225                         return $matches[0];
 226                 } else {
 227                         # multiple matched parts (variable match); some will be empty because of synonyms
 228                         # the variable will be the second non-empty one so remove any blank elements and re-sort the indices
 229                         $matches = array_values(array_filter($matches));
 230                         return $matches[1];
 231                 }
 232         }
 233
 234
 235         /**
 236          * Returns true if the text matches the word, and alters the
 237          * input string, removing all instances of the word
 238          */
 239         function matchAndRemove( &$text ) {
 240                 global $wgMagicFound;
 241                 $wgMagicFound = false;
 242                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 243                 return $wgMagicFound;
 244         }
 245
 246         function matchStartAndRemove( &$text ) {
 247                 global $wgMagicFound;
 248                 $wgMagicFound = false;
 249                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 250                 return $wgMagicFound;
 251         }
 252
 253
 254         /**
 255          * Replaces the word with something else
 256          */
 257         function replace( $replacement, $subject ) {
 258                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 259                 $this->mModified = !($res === $subject);
 260                 return $res;
 261         }
 262
 263         /**
 264          * Variable handling: {{SUBST:xxx}} style words
 265          * Calls back a function to determine what to replace xxx with
 266          * Input word must contain $1
 267          */
 268         function substituteCallback( $text, $callback ) {
 269                 $regex = $this->getVariableRegex();
 270                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 271                 $this->mModified = !($res === $text);
 272                 return $res;
 273         }
 274
 275         /**
 276          * Matches the word, where $1 is a wildcard
 277          */
 278         function getVariableRegex()     {
 279                 if ( $this->mVariableRegex == '' ) {
 280                         $this->initRegex();
 281                 }
 282                 return $this->mVariableRegex;
 283         }
 284
 285         /**
 286          * Matches the entire string, where $1 is a wildcard
 287          */
 288         function getVariableStartToEndRegex() {
 289                 if ( $this->mVariableStartToEndRegex == '' ) {
 290                         $this->initRegex();
 291                 }
 292                 return $this->mVariableStartToEndRegex;
 293         }
 294
 295         /**
 296          * Accesses the synonym list directly
 297          */
 298         function getSynonym( $i ) {
 299                 return $this->mSynonyms[$i];
 300         }
 301
 302         /**
 303          * Returns true if the last call to replace() or substituteCallback()
 304          * returned a modified text, otherwise false.
 305          */
 306         function getWasModified(){
 307                 return $this->mModified;
 308         }
 309
 310         /**
 311          * $magicarr is an associative array of (magic word ID => replacement)
 312          * This method uses the php feature to do several replacements at the same time,
 313          * thereby gaining some efficiency. The result is placed in the out variable
 314          * $result. The return value is true if something was replaced.
 315          * @static
 316          **/
 317         function replaceMultiple( $magicarr, $subject, &$result ){
 318                 $search = array();
 319                 $replace = array();
 320                 foreach( $magicarr as $id => $replacement ){
 321                         $mw = MagicWord::get( $id );
 322                         $search[] = $mw->getRegex();
 323                         $replace[] = $replacement;
 324                 }
 325
 326                 $result = preg_replace( $search, $replace, $subject );
 327                 return !($result === $subject);
 328         }
 329
 330         /**
 331          * Adds all the synonyms of this MagicWord to an array, to allow quick
 332          * lookup in a list of magic words
 333          */
 334         function addToArray( &$array, $value ) {
 335                 foreach ( $this->mSynonyms as $syn ) {
 336                         $array[$syn] = $value;
 337                 }
 338         }
 339 }
 340
 341 /**
 342  * Used in matchAndRemove()
 343  * @private
 344  **/
 345 function pregRemoveAndRecord( $match ) {
 346         global $wgMagicFound;
 347         $wgMagicFound = true;
 348         return '';
 349 }
 350
 351 ?>