includes/parser/StripState.php

   1 <?php
   2 /**
   3  * Holder for stripped items when parsing wiki markup.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Parser
  22  */
  23
  24 /**
  25  * @todo document, briefly.
  26  * @ingroup Parser
  27  */
  28 class StripState {
  29         protected $prefix;
  30         protected $data;
  31         protected $regex;
  32
  33         protected $tempType, $tempMergePrefix;
  34         protected $circularRefGuard;
  35         protected $recursionLevel = 0;
  36
  37         const UNSTRIP_RECURSION_LIMIT = 20;
  38
  39         /**
  40          * @param string|null $prefix
  41          * @since 1.26 The prefix argument should be omitted, as the strip marker
  42          *  prefix string is now a constant.
  43          */
  44         public function __construct( $prefix = null ) {
  45                 if ( $prefix !== null ) {
  46                         wfDeprecated( __METHOD__ . ' with called with $prefix argument' .
  47                                 ' (call with no arguments instead)', '1.26' );
  48                 }
  49                 $this->data = array(
  50                         'nowiki' => array(),
  51                         'general' => array()
  52                 );
  53                 $this->regex = '/' . Parser::MARKER_PREFIX . "([^\x7f]+)" . Parser::MARKER_SUFFIX . '/';
  54                 $this->circularRefGuard = array();
  55         }
  56
  57         /**
  58          * Add a nowiki strip item
  59          * @param string $marker
  60          * @param string $value
  61          */
  62         public function addNoWiki( $marker, $value ) {
  63                 $this->addItem( 'nowiki', $marker, $value );
  64         }
  65
  66         /**
  67          * @param string $marker
  68          * @param string $value
  69          */
  70         public function addGeneral( $marker, $value ) {
  71                 $this->addItem( 'general', $marker, $value );
  72         }
  73
  74         /**
  75          * @throws MWException
  76          * @param string $type
  77          * @param string $marker
  78          * @param string $value
  79          */
  80         protected function addItem( $type, $marker, $value ) {
  81                 if ( !preg_match( $this->regex, $marker, $m ) ) {
  82                         throw new MWException( "Invalid marker: $marker" );
  83                 }
  84
  85                 $this->data[$type][$m[1]] = $value;
  86         }
  87
  88         /**
  89          * @param string $text
  90          * @return mixed
  91          */
  92         public function unstripGeneral( $text ) {
  93                 return $this->unstripType( 'general', $text );
  94         }
  95
  96         /**
  97          * @param string $text
  98          * @return mixed
  99          */
 100         public function unstripNoWiki( $text ) {
 101                 return $this->unstripType( 'nowiki', $text );
 102         }
 103
 104         /**
 105          * @param string $text
 106          * @return mixed
 107          */
 108         public function unstripBoth( $text ) {
 109                 $text = $this->unstripType( 'general', $text );
 110                 $text = $this->unstripType( 'nowiki', $text );
 111                 return $text;
 112         }
 113
 114         /**
 115          * @param string $type
 116          * @param string $text
 117          * @return mixed
 118          */
 119         protected function unstripType( $type, $text ) {
 120                 // Shortcut
 121                 if ( !count( $this->data[$type] ) ) {
 122                         return $text;
 123                 }
 124
 125                 $oldType = $this->tempType;
 126                 $this->tempType = $type;
 127                 $text = preg_replace_callback( $this->regex, array( $this, 'unstripCallback' ), $text );
 128                 $this->tempType = $oldType;
 129                 return $text;
 130         }
 131
 132         /**
 133          * @param array $m
 134          * @return array
 135          */
 136         protected function unstripCallback( $m ) {
 137                 $marker = $m[1];
 138                 if ( isset( $this->data[$this->tempType][$marker] ) ) {
 139                         if ( isset( $this->circularRefGuard[$marker] ) ) {
 140                                 return '<span class="error">'
 141                                         . wfMessage( 'parser-unstrip-loop-warning' )->inContentLanguage()->text()
 142                                         . '</span>';
 143                         }
 144                         if ( $this->recursionLevel >= self::UNSTRIP_RECURSION_LIMIT ) {
 145                                 return '<span class="error">' .
 146                                         wfMessage( 'parser-unstrip-recursion-limit' )
 147                                                 ->numParams( self::UNSTRIP_RECURSION_LIMIT )->inContentLanguage()->text() .
 148                                         '</span>';
 149                         }
 150                         $this->circularRefGuard[$marker] = true;
 151                         $this->recursionLevel++;
 152                         $value = $this->data[$this->tempType][$marker];
 153                         if ( $value instanceof Closure ) {
 154                                 $value = $value();
 155                         }
 156                         $ret = $this->unstripType( $this->tempType, $value );
 157                         $this->recursionLevel--;
 158                         unset( $this->circularRefGuard[$marker] );
 159                         return $ret;
 160                 } else {
 161                         return $m[0];
 162                 }
 163         }
 164
 165         /**
 166          * Get a StripState object which is sufficient to unstrip the given text.
 167          * It will contain the minimum subset of strip items necessary.
 168          *
 169          * @param string $text
 170          *
 171          * @return StripState
 172          */
 173         public function getSubState( $text ) {
 174                 $subState = new StripState();
 175                 $pos = 0;
 176                 while ( true ) {
 177                         $startPos = strpos( $text, Parser::MARKER_PREFIX, $pos );
 178                         $endPos = strpos( $text, Parser::MARKER_SUFFIX, $pos );
 179                         if ( $startPos === false || $endPos === false ) {
 180                                 break;
 181                         }
 182
 183                         $endPos += strlen( Parser::MARKER_SUFFIX );
 184                         $marker = substr( $text, $startPos, $endPos - $startPos );
 185                         if ( !preg_match( $this->regex, $marker, $m ) ) {
 186                                 continue;
 187                         }
 188
 189                         $key = $m[1];
 190                         if ( isset( $this->data['nowiki'][$key] ) ) {
 191                                 $subState->data['nowiki'][$key] = $this->data['nowiki'][$key];
 192                         } elseif ( isset( $this->data['general'][$key] ) ) {
 193                                 $subState->data['general'][$key] = $this->data['general'][$key];
 194                         }
 195                         $pos = $endPos;
 196                 }
 197                 return $subState;
 198         }
 199
 200         /**
 201          * Merge another StripState object into this one. The strip marker keys
 202          * will not be preserved. The strings in the $texts array will have their
 203          * strip markers rewritten, the resulting array of strings will be returned.
 204          *
 205          * @param StripState $otherState
 206          * @param array $texts
 207          * @return array
 208          */
 209         public function merge( $otherState, $texts ) {
 210                 $mergePrefix = wfRandomString( 16 );
 211
 212                 foreach ( $otherState->data as $type => $items ) {
 213                         foreach ( $items as $key => $value ) {
 214                                 $this->data[$type]["$mergePrefix-$key"] = $value;
 215                         }
 216                 }
 217
 218                 $this->tempMergePrefix = $mergePrefix;
 219                 $texts = preg_replace_callback( $otherState->regex, array( $this, 'mergeCallback' ), $texts );
 220                 $this->tempMergePrefix = null;
 221                 return $texts;
 222         }
 223
 224         /**
 225          * @param array $m
 226          * @return string
 227          */
 228         protected function mergeCallback( $m ) {
 229                 $key = $m[1];
 230                 return Parser::MARKER_PREFIX . $this->tempMergePrefix . '-' . $key . Parser::MARKER_SUFFIX;
 231         }
 232
 233         /**
 234          * Remove any strip markers found in the given text.
 235          *
 236          * @param string $text Input string
 237          * @return string
 238          */
 239         public function killMarkers( $text ) {
 240                 return preg_replace( $this->regex, '', $text );
 241         }
 242 }