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