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 ( !is_array( $wgMagicWords ) ) {
  88                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
  89                 }
  90                 if (!array_key_exists( $id, $wgMagicWords ) ) {
  91                         $mw = new MagicWord();
  92                         $mw->load( $id );
  93                         $wgMagicWords[$id] = $mw;
  94                 }
  95                 return $wgMagicWords[$id];
  96         }
  97
  98         # Initialises this object with an ID
  99         function load( $id )
 100         {
 101                 global $wgLang;
 102                 $this->mId = $id;
 103                 $wgLang->getMagic( $this );
 104         }
 105
 106         # Preliminary initialisation
 107         /* private */ function initRegex()
 108         {
 109                 $variableClass = Title::legalChars();
 110                 $escSyn = array_map( "preg_quote", $this->mSynonyms );
 111                 $this->mBaseRegex = implode( "|", $escSyn );
 112                 $case = $this->mCaseSensitive ? "" : "i";
 113                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 114                 $this->mRegexStart = "/^{$this->mBaseRegex}/{$case}";
 115                 $this->mVariableRegex = str_replace( "\\$1", "([$variableClass]*?)", $this->mRegex );
 116                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "([$variableClass]*?)",
 117                         "/^({$this->mBaseRegex})$/{$case}" );
 118         }
 119
 120         # Gets a regex representing matching the word
 121         function getRegex()
 122         {
 123                 if ($this->mRegex == "" ) {
 124                         $this->initRegex();
 125                 }
 126                 return $this->mRegex;
 127         }
 128
 129         # Gets a regex matching the word, if it is at the
 130         # string start
 131         function getRegexStart()
 132         {
 133                 if ($this->mRegex == "" ) {
 134                         $this->initRegex();
 135                 }
 136                 return $this->mRegexStart;
 137         }
 138
 139         # regex without the slashes and what not
 140         function getBaseRegex()
 141         {
 142                 if ($this->mRegex == "") {
 143                         $this->initRegex();
 144                 }
 145                 return $this->mBaseRegex;
 146         }
 147
 148         # Returns true if the text contains the word
 149         function match( $text ) {
 150                 return preg_match( $this->getRegex(), $text );
 151         }
 152
 153         # Returns true if the text starts with the word
 154         function matchStart( $text )
 155         {
 156                 return preg_match( $this->getRegexStart(), $text );
 157         }
 158
 159         # Returns NULL if there's no match, the value of $1 otherwise
 160         # The return code is the matched string, if there's no variable
 161         # part in the regex and the matched variable part ($1) if there
 162         # is one.
 163         function matchVariableStartToEnd( $text ) {
 164                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 165                 if ( $matchcount == 0 ) {
 166                         return NULL;
 167                 } elseif ( count($matches) == 1 ) {
 168                         return $matches[0];
 169                 } else {
 170                         return $matches[1];
 171                 }
 172         }
 173
 174
 175         # Returns true if the text matches the word, and alters the
 176         # input string, removing all instances of the word
 177         function matchAndRemove( &$text )
 178         {
 179                 global $wgMagicFound;
 180                 $wgMagicFound = false;
 181                 $text = preg_replace_callback( $this->getRegex(), "pregRemoveAndRecord", $text );
 182                 return $wgMagicFound;
 183         }
 184
 185         function matchStartAndRemove( &$text ) {
 186                 global $wgMagicFound;
 187                 $wgMagicFound = false;
 188                 $text = preg_replace_callback( $this->getRegexStart(), "pregRemoveAndRecord", $text );
 189                 return $wgMagicFound;
 190         }
 191
 192
 193         # Replaces the word with something else
 194         function replace( $replacement, $subject )
 195         {
 196                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 197                 $this->mModified = !($res === $subject);
 198                 return $res;
 199         }
 200
 201         # Variable handling: {{SUBST:xxx}} style words
 202         # Calls back a function to determine what to replace xxx with
 203         # Input word must contain $1
 204         function substituteCallback( $text, $callback ) {
 205                 $regex = $this->getVariableRegex();
 206                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 207                 $this->mModified = !($res === $text);
 208                 return $res;
 209         }
 210
 211         # Matches the word, where $1 is a wildcard
 212         function getVariableRegex()
 213         {
 214                 if ( $this->mVariableRegex == "" ) {
 215                         $this->initRegex();
 216                 }
 217                 return $this->mVariableRegex;
 218         }
 219
 220         # Matches the entire string, where $1 is a wildcard
 221         function getVariableStartToEndRegex()
 222         {
 223                 if ( $this->mVariableStartToEndRegex == "" ) {
 224                         $this->initRegex();
 225                 }
 226                 return $this->mVariableStartToEndRegex;
 227         }
 228
 229         # Accesses the synonym list directly
 230         function getSynonym( $i ) {
 231                 return $this->mSynonyms[$i];
 232         }
 233
 234         # Returns true if the last call to replace() or substituteCallback()
 235         # returned a modified text, otherwise false.
 236         function getWasModified(){
 237                 return $this->mModified;
 238         }
 239
 240         # $magicarr is an associative array of (magic word ID => replacement)
 241         # This method uses the php feature to do several replacements at the same time,
 242         # thereby gaining some efficiency. The result is placed in the out variable
 243         # $result. The return value is true if something was replaced.
 244
 245         /* static */ function replaceMultiple( $magicarr, $subject, &$result ){
 246                 $search = array();
 247                 $replace = array();
 248                 foreach( $magicarr as $id => $replacement ){
 249                         $mw = MagicWord::get( $id );
 250                         $search[] = $mw->getRegex();
 251                         $replace[] = $replacement;
 252                 }
 253
 254                 $result = preg_replace( $search, $replace, $subject );
 255                 return !($result === $subject);
 256         }
 257
 258         # Adds all the synonyms of this MagicWord to an array, to allow quick lookup in a list of magic words
 259         function addToArray( &$array, $value )
 260         {
 261                 foreach ( $this->mSynonyms as $syn ) {
 262                         $array[$syn] = $value;
 263                 }
 264         }
 265 }
 266
 267 # Used in matchAndRemove()
 268 /*private*/ function pregRemoveAndRecord( $match )
 269 {
 270         global $wgMagicFound;
 271         $wgMagicFound = true;
 272         return "";
 273 }
 274
 275 ?>