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 define('MAG_GRAMMAR', 34);
  51
  52 $wgVariableIDs = array(
  53         MAG_CURRENTMONTH,
  54         MAG_CURRENTMONTHNAME,
  55         MAG_CURRENTDAY,
  56         MAG_CURRENTDAYNAME,
  57         MAG_CURRENTYEAR,
  58         MAG_CURRENTTIME,
  59         MAG_NUMBEROFARTICLES,
  60         MAG_CURRENTMONTHNAMEGEN,
  61         MAG_SITENAME,
  62         MAG_SERVER,
  63         MAG_PAGENAME,
  64         MAG_PAGENAMEE,
  65         MAG_NAMESPACE
  66 );
  67
  68 class MagicWord {
  69         /*private*/ var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  70         /*private*/ var $mRegexStart, $mBaseRegex, $mVariableRegex;
  71         /*private*/ var $mModified;
  72
  73         function MagicWord($id = 0, $syn = '', $cs = false) {
  74                 $this->mId = $id;
  75                 $this->mSynonyms = (array)$syn;
  76                 $this->mCaseSensitive = $cs;
  77                 $this->mRegex = '';
  78                 $this->mRegexStart = '';
  79                 $this->mVariableRegex = '';
  80                 $this->mVariableStartToEndRegex = '';
  81                 $this->mModified = false;
  82         }
  83
  84         # Factory: creates an object representing an ID
  85         /*static*/ function &get( $id ) {
  86                 global $wgMagicWords;
  87
  88                 if ( !is_array( $wgMagicWords ) ) {
  89                         wfDebugDieBacktrace( "Incorrect initialisation order, \$wgMagicWords does not exist\n" );
  90                 }
  91                 if (!array_key_exists( $id, $wgMagicWords ) ) {
  92                         $mw = new MagicWord();
  93                         $mw->load( $id );
  94                         $wgMagicWords[$id] = $mw;
  95                 }
  96                 return $wgMagicWords[$id];
  97         }
  98
  99         # Initialises this object with an ID
 100         function load( $id ) {
 101                 global $wgLang;
 102                 $this->mId = $id;
 103                 $wgLang->getMagic( $this );
 104         }
 105
 106         # Preliminary initialisation
 107         /* private */ function initRegex() {
 108                 $variableClass = Title::legalChars();
 109                 $escSyn = array_map( 'preg_quote', $this->mSynonyms );
 110                 $this->mBaseRegex = implode( '|', $escSyn );
 111                 $case = $this->mCaseSensitive ? '' : 'i';
 112                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 113                 $this->mRegexStart = "/^{$this->mBaseRegex}/{$case}";
 114                 $this->mVariableRegex = str_replace( "\\$1", "([$variableClass]*?)", $this->mRegex );
 115                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "([$variableClass]*?)",
 116                         "/^({$this->mBaseRegex})$/{$case}" );
 117         }
 118
 119         # Gets a regex representing matching the word
 120         function getRegex() {
 121                 if ($this->mRegex == '' ) {
 122                         $this->initRegex();
 123                 }
 124                 return $this->mRegex;
 125         }
 126
 127         # Gets a regex matching the word, if it is at the
 128         # string start
 129         function getRegexStart() {
 130                 if ($this->mRegex == '' ) {
 131                         $this->initRegex();
 132                 }
 133                 return $this->mRegexStart;
 134         }
 135
 136         # regex without the slashes and what not
 137         function getBaseRegex() {
 138                 if ($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                 return preg_match( $this->getRegexStart(), $text );
 152         }
 153
 154         # Returns NULL if there's no match, the value of $1 otherwise
 155         # The return code is the matched string, if there's no variable
 156         # part in the regex and the matched variable part ($1) if there
 157         # is one.
 158         function matchVariableStartToEnd( $text ) {
 159                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 160                 if ( $matchcount == 0 ) {
 161                         return NULL;
 162                 } elseif ( count($matches) == 1 ) {
 163                         return $matches[0];
 164                 } else {
 165                         return $matches[1];
 166                 }
 167         }
 168
 169
 170         # Returns true if the text matches the word, and alters the
 171         # input string, removing all instances of the word
 172         function matchAndRemove( &$text ) {
 173                 global $wgMagicFound;
 174                 $wgMagicFound = false;
 175                 $text = preg_replace_callback( $this->getRegex(), 'pregRemoveAndRecord', $text );
 176                 return $wgMagicFound;
 177         }
 178
 179         function matchStartAndRemove( &$text ) {
 180                 global $wgMagicFound;
 181                 $wgMagicFound = false;
 182                 $text = preg_replace_callback( $this->getRegexStart(), 'pregRemoveAndRecord', $text );
 183                 return $wgMagicFound;
 184         }
 185
 186
 187         # Replaces the word with something else
 188         function replace( $replacement, $subject ) {
 189                 $res = preg_replace( $this->getRegex(), $replacement, $subject );
 190                 $this->mModified = !($res === $subject);
 191                 return $res;
 192         }
 193
 194         # Variable handling: {{SUBST:xxx}} style words
 195         # Calls back a function to determine what to replace xxx with
 196         # Input word must contain $1
 197         function substituteCallback( $text, $callback ) {
 198                 $regex = $this->getVariableRegex();
 199                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 200                 $this->mModified = !($res === $text);
 201                 return $res;
 202         }
 203
 204         # Matches the word, where $1 is a wildcard
 205         function getVariableRegex()     {
 206                 if ( $this->mVariableRegex == '' ) {
 207                         $this->initRegex();
 208                 }
 209                 return $this->mVariableRegex;
 210         }
 211
 212         # Matches the entire string, where $1 is a wildcard
 213         function getVariableStartToEndRegex() {
 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                 foreach ( $this->mSynonyms as $syn ) {
 252                         $array[$syn] = $value;
 253                 }
 254         }
 255 }
 256
 257 # Used in matchAndRemove()
 258 /*private*/ function pregRemoveAndRecord( $match ) {
 259         global $wgMagicFound;
 260         $wgMagicFound = true;
 261         return '';
 262 }
 263
 264 ?>