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_PAGENAMEE", 31);
  48 define("MAG_NAMESPACE", 32);
  49 define("MAG_TOC", 33);
  50
  51 $wgVariableIDs = array(
  52         MAG_CURRENTMONTH,
  53         MAG_CURRENTMONTHNAME,
  54         MAG_CURRENTDAY,
  55         MAG_CURRENTDAYNAME,
  56         MAG_CURRENTYEAR,
  57         MAG_CURRENTTIME,
  58         MAG_NUMBEROFARTICLES,
  59         MAG_CURRENTMONTHNAMEGEN,
  60         MAG_SITENAME,
  61         MAG_SERVER,
  62         MAG_PAGENAME,
  63         MAG_PAGENAMEE,
  64         MAG_NAMESPACE
  65 );
  66
  67 class MagicWord {
  68         /*private*/ var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  69         /*private*/ var $mRegexStart, $mBaseRegex, $mVariableRegex;
  70         /*private*/ var $mModified;
  71
  72         function MagicWord($id = 0, $syn = '', $cs = false) {
  73                 $this->mId = $id;
  74                 $this->mSynonyms = (array)$syn;
  75                 $this->mCaseSensitive = $cs;
  76                 $this->mRegex = '';
  77                 $this->mRegexStart = '';
  78                 $this->mVariableRegex = '';
  79                 $this->mVariableStartToEndRegex = '';
  80                 $this->mModified = false;
  81         }
  82
  83         # Factory: creates an object representing an ID
  84         /*static*/ function &get( $id ) {
  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                 global $wgLang;
 101                 $this->mId = $id;
 102                 $wgLang->getMagic( $this );
 103         }
 104
 105         # Preliminary initialisation
 106         /* private */ function initRegex() {
 107                 $variableClass = Title::legalChars();
 108                 $escSyn = array_map( "preg_quote", $this->mSynonyms );
 109                 $this->mBaseRegex = implode( "|", $escSyn );
 110                 $case = $this->mCaseSensitive ? "" : "i";
 111                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 112                 $this->mRegexStart = "/^{$this->mBaseRegex}/{$case}";
 113                 $this->mVariableRegex = str_replace( "\\$1", "([$variableClass]*?)", $this->mRegex );
 114                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "([$variableClass]*?)",
 115                         "/^({$this->mBaseRegex})$/{$case}" );
 116         }
 117
 118         # Gets a regex representing matching the word
 119         function getRegex() {
 120                 if ($this->mRegex == '' ) {
 121                         $this->initRegex();
 122                 }
 123                 return $this->mRegex;
 124         }
 125
 126         # Gets a regex matching the word, if it is at the
 127         # string start
 128         function getRegexStart() {
 129                 if ($this->mRegex == '' ) {
 130                         $this->initRegex();
 131                 }
 132                 return $this->mRegexStart;
 133         }
 134
 135         # regex without the slashes and what not
 136         function getBaseRegex() {
 137                 if ($this->mRegex == '') {
 138                         $this->initRegex();
 139                 }
 140                 return $this->mBaseRegex;
 141         }
 142
 143         # Returns true if the text contains the word
 144         function match( $text ) {
 145                 return preg_match( $this->getRegex(), $text );
 146         }
 147
 148         # Returns true if the text starts with the word
 149         function matchStart( $text ) {
 150                 return preg_match( $this->getRegexStart(), $text );
 151         }
 152
 153         # Returns NULL if there's no match, the value of $1 otherwise
 154         # The return code is the matched string, if there's no variable
 155         # part in the regex and the matched variable part ($1) if there
 156         # is one.
 157         function matchVariableStartToEnd( $text ) {
 158                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 159                 if ( $matchcount == 0 ) {
 160                         return NULL;
 161                 } elseif ( count($matches) == 1 ) {
 162                         return $matches[0];
 163                 } else {
 164                         return $matches[1];
 165                 }
 166         }
 167
 168
 169         # Returns true if the text matches the word, and alters the
 170         # input string, removing all instances of the word
 171         function matchAndRemove( &$text ) {
 172                 global $wgMagicFound;
 173                 $wgMagicFound = false;
 174                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 175                 return $wgMagicFound;
 176         }
 177
 178         function matchStartAndRemove( &$text ) {
 179                 global $wgMagicFound;
 180                 $wgMagicFound = false;
 181                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 182                 return $wgMagicFound;
 183         }
 184
 185
 186         # Replaces the word with something else
 187         function replace( $replacement, $subject ) {
 188                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 189                 $this->mModified = !($res === $subject);
 190                 return $res;
 191         }
 192
 193         # Variable handling: {{SUBST:xxx}} style words
 194         # Calls back a function to determine what to replace xxx with
 195         # Input word must contain $1
 196         function substituteCallback( $text, $callback ) {
 197                 $regex = $this->getVariableRegex();
 198                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 199                 $this->mModified = !($res === $text);
 200                 return $res;
 201         }
 202
 203         # Matches the word, where $1 is a wildcard
 204         function getVariableRegex()     {
 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                 if ( $this->mVariableStartToEndRegex == '' ) {
 214                         $this->initRegex();
 215                 }
 216                 return $this->mVariableStartToEndRegex;
 217         }
 218
 219         # Accesses the synonym list directly
 220         function getSynonym( $i ) {
 221                 return $this->mSynonyms[$i];
 222         }
 223
 224         # Returns true if the last call to replace() or substituteCallback()
 225         # returned a modified text, otherwise false.
 226         function getWasModified(){
 227                 return $this->mModified;
 228         }
 229
 230         # $magicarr is an associative array of (magic word ID => replacement)
 231         # This method uses the php feature to do several replacements at the same time,
 232         # thereby gaining some efficiency. The result is placed in the out variable
 233         # $result. The return value is true if something was replaced.
 234
 235         /* static */ function replaceMultiple( $magicarr, $subject, &$result ){
 236                 $search = array();
 237                 $replace = array();
 238                 foreach( $magicarr as $id => $replacement ){
 239                         $mw = MagicWord::get( $id );
 240                         $search[] = $mw->getRegex();
 241                         $replace[] = $replacement;
 242                 }
 243
 244                 $result = preg_replace( $search, $replace, $subject );
 245                 return !($result === $subject);
 246         }
 247
 248         # Adds all the synonyms of this MagicWord to an array, to allow quick lookup in a list of magic words
 249         function addToArray( &$array, $value ) {
 250                 foreach ( $this->mSynonyms as $syn ) {
 251                         $array[$syn] = $value;
 252                 }
 253         }
 254 }
 255
 256 # Used in matchAndRemove()
 257 /*private*/ function pregRemoveAndRecord( $match ) {
 258         global $wgMagicFound;
 259         $wgMagicFound = true;
 260         return '';
 261 }
 262
 263 ?>