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