includes/MagicWord.php

   1 <?php
   2
   3 # This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
   4 # Usage:
   5 #     if (MagicWord::get( MAG_REDIRECT )->match( $text ) )
   6 #
   7 # Possible future improvements:
   8 #   * Simultaneous searching for a number of magic words
   9 #   * $wgMagicWords in shared memory
  10 #
  11 # Please avoid reading the data out of one of these objects and then writing
  12 # special case code. If possible, add another match()-like function here.
  13
  14 /*private*/ $wgMagicFound = false;
  15
  16 define("MAG_REDIRECT", 0);
  17 define("MAG_NOTOC", 1);
  18 define("MAG_START", 2);
  19 define("MAG_CURRENTMONTH", 3);
  20 define("MAG_CURRENTMONTHNAME", 4);
  21 define("MAG_CURRENTDAY", 5);
  22 define("MAG_CURRENTDAYNAME", 6);
  23 define("MAG_CURRENTYEAR", 7);
  24 define("MAG_CURRENTTIME", 8);
  25 define("MAG_NUMBEROFARTICLES", 9);
  26 define("MAG_CURRENTMONTHNAMEGEN", 10);
  27 define("MAG_MSG", 11);
  28 define("MAG_SUBST", 12);
  29 define("MAG_MSGNW", 13);
  30 define("MAG_NOEDITSECTION", 14);
  31 define("MAG_END", 15);
  32 define("MAG_IMG_THUMBNAIL",  16);
  33 define("MAG_IMG_RIGHT",      17);
  34 define("MAG_IMG_LEFT",       18);
  35 define("MAG_IMG_NONE",       19);
  36 define("MAG_IMG_WIDTH",      20);
  37 define("MAG_IMG_CENTER",      21);
  38 define("MAG_INT", 22);
  39 define("MAG_FORCETOC", 23);
  40 define("MAG_SITENAME", 24);
  41 define("MAG_NS", 25);
  42 define("MAG_LOCALURL", 26);
  43 define("MAG_LOCALURLE", 27);
  44 define("MAG_SERVER", 28);
  45 define("MAG_IMG_FRAMED", 29);
  46 define("MAG_PAGENAME", 30);
  47 define("MAG_NAMESPACE", 31);
  48 define("MAG_TOC", 32);
  49
  50 $wgVariableIDs = array(
  51         MAG_CURRENTMONTH,
  52         MAG_CURRENTMONTHNAME,
  53         MAG_CURRENTDAY,
  54         MAG_CURRENTDAYNAME,
  55         MAG_CURRENTYEAR,
  56         MAG_CURRENTTIME,
  57         MAG_NUMBEROFARTICLES,
  58         MAG_CURRENTMONTHNAMEGEN,
  59     MAG_SITENAME,
  60         MAG_SERVER,
  61         MAG_PAGENAME,
  62         MAG_NAMESPACE
  63 );
  64
  65 class MagicWord {
  66         /*private*/ var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  67         /*private*/ var $mRegexStart, $mBaseRegex, $mVariableRegex;
  68         /*private*/ var $mModified;
  69
  70         function MagicWord($id = 0, $syn = "", $cs = false)
  71         {
  72                 $this->mId = $id;
  73                 $this->mSynonyms = (array)$syn;
  74                 $this->mCaseSensitive = $cs;
  75                 $this->mRegex = "";
  76                 $this->mRegexStart = "";
  77                 $this->mVariableRegex = "";
  78                 $this->mVariableStartToEndRegex = "";
  79                 $this->mModified = false;
  80         }
  81
  82         # Factory: creates an object representing an ID
  83         /*static*/ function &get( $id )
  84         {
  85                 global $wgMagicWords;
  86
  87                 if (!array_key_exists( $id, $wgMagicWords ) ) {
  88                         $mw = new MagicWord();
  89                         $mw->load( $id );
  90                         $wgMagicWords[$id] = $mw;
  91                 }
  92                 return $wgMagicWords[$id];
  93         }
  94
  95         # Initialises this object with an ID
  96         function load( $id )
  97         {
  98                 global $wgLang;
  99                 $this->mId = $id;
 100                 $wgLang->getMagic( $this );
 101         }
 102
 103         # Preliminary initialisation
 104         /* private */ function initRegex()
 105         {
 106                 $variableClass = Title::legalChars();
 107                 $escSyn = array_map( "preg_quote", $this->mSynonyms );
 108                 $this->mBaseRegex = implode( "|", $escSyn );
 109                 $case = $this->mCaseSensitive ? "" : "i";
 110                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 111                 $this->mRegexStart = "/^{$this->mBaseRegex}/{$case}";
 112                 $this->mVariableRegex = str_replace( "\\$1", "([$variableClass]*?)", $this->mRegex );
 113                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "([$variableClass]*?)",
 114                         "/^({$this->mBaseRegex})$/{$case}" );
 115         }
 116
 117         # Gets a regex representing matching the word
 118         function getRegex()
 119         {
 120                 if ($this->mRegex == "" ) {
 121                         $this->initRegex();
 122                 }
 123                 return $this->mRegex;
 124         }
 125
 126         # Gets a regex matching the word, if it is at the
 127         # string start
 128         function getRegexStart()
 129         {
 130                 if ($this->mRegex == "" ) {
 131                         $this->initRegex();
 132                 }
 133                 return $this->mRegexStart;
 134         }
 135
 136         # regex without the slashes and what not
 137         function getBaseRegex()
 138         {
 139                 if ($this->mRegex == "") {
 140                         $this->initRegex();
 141                 }
 142                 return $this->mBaseRegex;
 143         }
 144
 145         # Returns true if the text contains the word
 146         function match( $text ) {
 147                 return preg_match( $this->getRegex(), $text );
 148         }
 149
 150         # Returns true if the text starts with the word
 151         function matchStart( $text )
 152         {
 153                 return preg_match( $this->getRegexStart(), $text );
 154         }
 155
 156         # Returns NULL if there's no match, the value of $1 otherwise
 157         # The return code is the matched string, if there's no variable
 158         # part in the regex and the matched variable part ($1) if there
 159         # is one.
 160         function matchVariableStartToEnd( $text ) {
 161                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 162                 if ( $matchcount == 0 ) {
 163                         return NULL;
 164                 } elseif ( count($matches) == 1 ) {
 165                         return $matches[0];
 166                 } else {
 167                         return $matches[1];
 168                 }
 169         }
 170
 171
 172         # Returns true if the text matches the word, and alters the
 173         # input string, removing all instances of the word
 174         function matchAndRemove( &$text )
 175         {
 176                 global $wgMagicFound;
 177                 $wgMagicFound = false;
 178                 $text = preg_replace_callback( $this->getRegex(), "pregRemoveAndRecord", $text );
 179                 return $wgMagicFound;
 180         }
 181
 182         function matchStartAndRemove( &$text ) {
 183                 global $wgMagicFound;
 184                 $wgMagicFound = false;
 185                 $text = preg_replace_callback( $this->getRegexStart(), "pregRemoveAndRecord", $text );
 186                 return $wgMagicFound;
 187         }
 188
 189
 190         # Replaces the word with something else
 191         function replace( $replacement, $subject )
 192         {
 193                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 194                 $this->mModified = !($res === $subject);
 195                 return $res;
 196         }
 197
 198         # Variable handling: {{SUBST:xxx}} style words
 199         # Calls back a function to determine what to replace xxx with
 200         # Input word must contain $1
 201         function substituteCallback( $text, $callback ) {
 202                 $regex = $this->getVariableRegex();
 203                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 204                 $this->mModified = !($res === $text);
 205                 return $res;
 206         }
 207
 208         # Matches the word, where $1 is a wildcard
 209         function getVariableRegex()
 210         {
 211                 if ( $this->mVariableRegex == "" ) {
 212                         $this->initRegex();
 213                 }
 214                 return $this->mVariableRegex;
 215         }
 216
 217         # Matches the entire string, where $1 is a wildcard
 218         function getVariableStartToEndRegex()
 219         {
 220                 if ( $this->mVariableStartToEndRegex == "" ) {
 221                         $this->initRegex();
 222                 }
 223                 return $this->mVariableStartToEndRegex;
 224         }
 225
 226         # Accesses the synonym list directly
 227         function getSynonym( $i ) {
 228                 return $this->mSynonyms[$i];
 229         }
 230
 231         # Returns true if the last call to replace() or substituteCallback()
 232         # returned a modified text, otherwise false.
 233         function getWasModified(){
 234                 return $this->mModified;
 235         }
 236
 237         # $magicarr is an associative array of (magic word ID => replacement)
 238         # This method uses the php feature to do several replacements at the same time,
 239         # thereby gaining some efficiency. The result is placed in the out variable
 240         # $result. The return value is true if something was replaced.
 241
 242         /* static */ function replaceMultiple( $magicarr, $subject, &$result ){
 243                 $search = array();
 244                 $replace = array();
 245                 foreach( $magicarr as $id => $replacement ){
 246                         $mw = MagicWord::get( $id );
 247                         $search[] = $mw->getRegex();
 248                         $replace[] = $replacement;
 249                 }
 250
 251                 $result = preg_replace( $search, $replace, $subject );
 252                 return !($result === $subject);
 253         }
 254
 255         # Adds all the synonyms of this MagicWord to an array, to allow quick lookup in a list of magic words
 256         function addToArray( &$array, $value )
 257         {
 258                 foreach ( $this->mSynonyms as $syn ) {
 259                         $array[$syn] = $value;
 260                 }
 261         }
 262 }
 263
 264 # Used in matchAndRemove()
 265 /*private*/ function pregRemoveAndRecord( $match )
 266 {
 267         global $wgMagicFound;
 268         $wgMagicFound = true;
 269         return "";
 270 }
 271
 272 ?>